Computer IEEE

CS PRESIDENT’S MESSAGE, P. 6

EIC’S MESSAGE, P. 8

COMPUTING CONVERSATIONS: VAN JACOBSON, P. 11

JA

NU

AR

Y 2

013

http

://w

ww

.com

pute

r.or

g

Contents | Zoom in | Zoom out Search Issue | Next PageFor navigation instructions please click here

Contents | Zoom in | Zoom out Search Issue | Next PageFor navigation instructions please click here

________________________

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org&id=17866&adid=PCOVER 1E1

Build for the new Windows Store.Open for business at windowsstore.com

Cook up thenext great app.Opportunity doesn’t just knock. In the Windows Store it swipes, taps and clicks, too. See how Allrecipes and others are building immersive apps for the new Windows experience and learn how you can put your app in the hands of new users everywhere.

Previous Page | Contents | Zoom in | Zoom out | Front Cover | Search Issue | Next PageComputerComputer


qqM

Mq

qM

MqM

Qmags®THE WORLD’S NEWSSTAND

qqM

Mq

qM

MqM


________________

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org&id=17866&adid=logo

http://www.qmags.com/clickthrough.asp?url=www.qmags.com&id=17866&adid=logo



http://www.qmags.com/clickthrough.asp?url=www.windowsstore.com&id=17866&adid=PCOVER 2A1

JANUARY 2013 1

Editorial StaffJudith ProwManaging [email protected] NelsonSenior Editor

Contributing EditorsCamber AgreliusChristine AnthonyLee GarberBob WardStaff Multimedia EditorsBrian BrannonBen Jones

Design and ProductionCamber Agrelius Larry BauerDesignOlga D’AstoliCover DesignDavid AngelKate WojogbeJennie Zhu-Mai

Administrative StaffProducts and Services DirectorEvan ButterfieldSenior Manager, Editorial Services Lars Jentsch

Manager, Editorial ServicesJennifer StoutSenior Business Development ManagerSandy BrownSenior Advertising CoordinatorMarian Anderson

Editor in ChiefRon VetterUniversity of North Carolina [email protected]

Associate Editor in ChiefSumi Helal University of [email protected]

Associate Editor in Chief,Research FeaturesKathleen Swigger University of North [email protected]

Associate Editor in Chief,Special IssuesBill N. [email protected]

Computing PracticesRohit [email protected]

PerspectivesBob [email protected]

Multimedia EditorCharles R. Severance [email protected]

2013 IEEE Computer Society PresidentDavid Alan [email protected]

Area EditorsComputer ArchitecturesDavid H. AlbonesiCornell UniversityGreg ByrdNorth Carolina State UniversityGraphics and MultimediaOliver BimberJohannes Kepler University LinzHigh-Performance ComputingVladimir GetovUniversity of WestminsterInformation andData ManagementNaren RamakrishnanVirginia Tech Internet ComputingSimon ShimSan Jose State University MultimediaSavitha SrinivasanIBM Almaden Research Center Networking Ahmed HelmyUniversity of Florida Ying-Dar LinNational Chiao Tung UniversitySecurity and PrivacyRolf OppligereSECURITY TechnologiesSoftwareRenée BryceUniversity of North TexasJean-Marc JézéquelUniversity of RennesDavid M. WeissIowa State University

Column EditorsCloud CoverSan MurugesanBRITE Professional ResourcesComputing ConversationsCharles R. SeveranceUniversity of MichiganDiscovery AnalyticsNaren RamakrishnanVirginia TechEducationAnn E.K. SobelMiami UniversityEntertainment ComputingKelvin SungUniversity of Washington, BothellThe Errant HashtagDavid Alan GrierGeorge Washington UniversityGreen ITKirk W. CameronVirginia TechIdentity SciencesKarl RicanekUniversity of North Carolina WilmingtonIn DevelopmentChris HuntleyFairfield UniversityInvisible ComputingAlbrecht SchmidtUniversity of StuttgartOut of BandHal BerghelUniversity of Nevada, Las VegasScience Fiction PrototypingBrian David JohnsonIntel

SecurityJeffrey M. VoasNISTSocial ComputingJohn RiedlUniversity of MinnesotaSoftware TechnologiesMike HincheyLero—the Irish Software Engineering Research Centre32 & 16 Years AgoNeville HolmesUniversity of Tasmania

Advisory PanelCarl K. ChangEditor in Chief EmeritusIowa State UniversityJean BaconUniversity of CambridgeHal BerghelUniversity of Nevada, Las VegasDoris L. CarverLouisiana State UniversityNaren RamakrishnanVirginia TechTheresa-Marie RhyneConsultantAlf WeaverUniversity of Virginia

2013 Publications BoardThomas M. Conte (chair), Alain April, David Bader, Angela R. Burgess, Greg Byrd, Koen DeBosschere, Frank Ferrante, Paolo Montuschi, Linda I. Shafer, Per Stenström

2013 Magazine Operations CommitteePaolo Montuschi (chair), Erik R. Altman, Nigel Davies, Lars Heide, Simon Liu, Cecelia Metra, Shari Lawrence Pfleeger, Michael Rabinovich, Forrest Shull, John Smith, Gabriel Taubin, George K. Thiruvathukal, Ron Vetter, Daniel Zeng

Circulation: Computer (ISSN 0018-9162) is published monthly by the IEEE Computer Society. IEEE Headquarters, Three Park Avenue, 17th Floor, New York, NY 10016-5997; IEEE Computer Society Publications Office, 10662 Los Vaqueros Circle, Los Alamitos, CA 90720; voice +1 714 821 8380; fax +1 714 821 4010; IEEE Computer Society Headquarters, 2001 L Street NW, Suite 700, Washington, DC 20036. IEEE Computer Society membership includes $19 for a subscription to Computer magazine. Nonmember subscription rate available upon request. Single-copy prices: members $20; nonmembers $175. Postmaster: Send undelivered copies and address changes to Computer, IEEE Membership Processing Dept., 445 Hoes Lane, Piscataway, NJ 08855. Periodicals Postage Paid at New York, New York, and at additional mailing offices. Canadian GST #125634188. Canada Post Corporation (Canadian distribution) publications mail agreement number 40013885. Return undeliverable Canadian addresses to PO Box 122, Niagara Falls, ON L2E 6S8 Canada. Printed in USA.Editorial: Unless otherwise stated, bylined articles, as well as product and service descriptions, reflect the author’s or firm’s opinion. Inclusion in Computerdoes not necessarily constitute endorsement by the IEEE or the Computer Society. All submissions are subject to editing for style, clarity, and space.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


__________

__________ ____________

_________________________

_______________

_________

___________

___________

mailto:[email protected]





mialto:[email protected]








CO

NT

EN

TS

COVER FEATURES28 Mobile Health: Revolutioniz-

ing Healthcare through Transdisciplinary Research

Santosh Kumar, Wendy Nilsen, Misha Pavel, and Mani SrivastavaMobile health (mHealth) seeks to improve individuals’ health and well-being by continuously monitoring their status, rapidly diagnosing medical conditions, recognizing behaviors, and delivering just-in-time interventions, all in the user’s natural mobile environment.

36 The Future of Human-in-the- Loop Cyber-Physical Systems

Gunar Schirner, Deniz Erdogmus, Kaushik Chowdhury, and Taskin Padir A prototyping platform and a design framework for rapid exploration of a novel human-in-the-loop application serves as an accelerator for new research into a broad class of systems that augment human interaction with the physical world.

46 Experiential Media and Digital Culture

Thanassis Rikakis, Aisling Kelliher, and Nicole LehrerMultidisciplinary value structures and a design approach focusing on combining efficiency, reflection, and quality of experience will foster the true hybrid physical-digital culture that is foundational to solving complex societal problems.

ABOUT THIS ISSUE

Topics included in this year’s outlook issue include wearable wireless sensor systems for

continuous assessment of human health (mHealth); cyber-physical systems that augment human inter-action with the physical world; experiential media technologies and how they incorporate integrative approaches across the arts, design, science, and engineering; and an ethical assessment of robot surgery.

56 Robotic Surgery: On the Cutting Edge of Ethics

Noel Sharkey and Amanda SharkeyRobotic surgery is a relatively new technology that promises unparalleled innovation with minimization of pain and time spent in recovery, but surgeons must follow realistic ethical guidelines if this treatment option is to thrive.

PERSPECTIVES65 Management of Hidden Risks

Kjell Jørgen HoleThe ability to recover quickly from large-impact, hard-to-predict, and rare incidents without incurring sizable permanent damage is vital to major stakeholders in ICT infrastructures of national importance.

RESEARCH FEATURE71 A Smart Approach to

Medication Management Upkar VarshneySmart medication management systems can remotely monitor and regulate patients’ prescription drug use, support communication with caregivers to improve scheduling and increase motivation, and generate context-aware reminders.

w w w. c o m p u t e r. o r g /c o m p u t e r

For more information on computing topics, visit the Computer Society Digital Library at www.computer.org/csdl.

See www.computer.org/computer-multimedia for multimedia content related to the features in this issue.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____

MU

LTIM

ED

IA

http://www.qmags.com/clickthrough.asp?url=www.computer.org/csdl&id=17866&adid=P2E3

http://www.qmags.com/clickthrough.asp?url=www.computer.org/computer-multimedia&id=17866&adid=P2E2

http://www.qmags.com/clickthrough.asp?url=www.computer.org/computer&id=17866&adid=P2E1





We welcome your letters. Send them to [email protected]. Letters are subject to editing for style, clarity, and length.

Reuse Rights and Reprint Permissions: Educational or personal use of this material is permitted without fee, provided such use: 1) is not made for profit; 2) includes this notice and a full citation to the original work on the first page of the copy;

and 3) does not imply IEEE endorsement of any third-party products or services. Authors and their companies are permitted to post the accepted version of their IEEE-copyrighted material on their own Web servers without permission, provided that the IEEE copyright notice and a full citation to the original work appear on the first screen of the posted copy. An accepted manuscript is a version which has been revised by the author to incorporate review suggestions, but not the published version with copyediting, proofreading and formatting added by IEEE. For more information, please go to: http://www.ieee.org/publications_standards/publications/rights/paperversionpolicy.html.

Permission to reprint/republish this material for commercial, advertising, or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by writing to the IEEE Intellectual Property Rights Office, 445 Hoes Lane, Piscataway, NJ 08854-4141 or [email protected]. Copyright © 2013 IEEE. All rights reserved.

Abstracting and Library Use: Abstracting is permitted with credit to the source. Libraries are permitted to photocopy for private use of patrons, provided the per-copy fee indicated in the code at the bottom of the first page is paid through the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923.

IEEE prohibits discrimination, harassment, and bullying. For more information, visit www.ieee.org/web/aboutus/whatis/policies/p9-26.html.

COLUMNS11 Computing Conversations

Van Jacobson: Content-Centric NetworkingCharles Severance

15 Computing and the LawProtecting Bioinformatics as Intellectual PropertyBrian M. Gaff, Ralph A. Loren, and Gareth Dickson

18 32 & 16 Years AgoComputer, January 1981 and 1997Neville Holmes

77 Cloud CoverCloud Computing: The New Normal?San Murugesan

80 Science Fiction PrototypingWhen Science Fiction and Science Fact MeetBrian David Johnson

83 Green ITEnergy Oddities, Part 1: Why the Energy World Is OddKirk W. Cameron

85 Out of BandRFIDiocy: It’s Déjà Vu All over AgainHal Berghel

89 Social ComputingCrowdsourcing Medical ResearchJohn Riedl and Eric Riedl

January 2013, Volume 46, Number 1

IEEE Computer Society: http://computer.orgComputer: http://computer.org/[email protected] IEEE Computer Society Publications Office: +1 714 821 8380

F l a g s h i p P u b l i c a t i o n o f t h e I E E E C o m p u t e r S o c i e t y

108 Errant HashtagThe Here and NowDavid Alan Grier

NEWS20 Technology News

Bringing Big Analytics to the MassesNeal Leavitt

24 News BriefsLee Garber

MEMBERSHIP NEWS6 CS President’s Message

8 EIC’s Message

103 IEEE Computer Society Connection

105 Call and Calendar

DEPARTMENTS 4 Elsewhere in the CS

14 Computer Society Information94 Career Opportunities



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___________

____________

_______________

________________________________________________

__________________________

http://www.qmags.com/clickthrough.asp?url=www.ieee.org/web/aboutus/whatis/policies/p9-26.html&id=17866&adid=P3E4

http://www.qmags.com/clickthrough.asp?url=www.ieee.org/publications_standards/publications/rights/paperversionpolicy.html&id=17866&adid=P3E3

http://www.qmags.com/clickthrough.asp?url=http://computer.org/computer&id=17866&adid=P3E2

http://www.qmags.com/clickthrough.asp?url=http://computer.org&id=17866&adid=P3E1




http://www.qmags.com/clickthrough.asp?url=www.sfiprogram.org&id=17866&adid=P3E5





4 COMPUTER Published by the IEEE Computer Society 0018-9162/13/$31.00 © 2013 IEEE

T he IEEE Computer Society’s lineup of 12 peer-reviewed technical magazines cover cutting-edge topics in computing, including scientific applications, Internet computing,

machine intelligence, pervasive computing, security and privacy, digital graphics, and computer history. Select arti-cles from recent issues of Computer Society magazines are highlighted below.

Software’s November/December theme is technical debt, a term coined by Ward Cunningham in 1992 to describe “not quite right code which we postpone making right.” In their introduction, “Technical Debt: From Metaphor to Theory and Practice,” guest editors Philippe Kruchten of the University of British Columbia, Vancouver, and Robert L. Nord and Ipek Ozkaya of the Software Engineering Institute argue that the term has become diluted over the years. They propose some structures to establish a theoretical foundation for defining and applying technical debt to software development and introduce four feature articles addressing various aspects of the concept.

S&P’s November/December special issue, titled “Lost Treasures,” is guest edited by Dan Thomsen of Smart Information Flow Technologies and Jeremy Epstein and Peter G. Neumann of SRI International. They present five articles focused on key moments in the history of secure systems—not for posterity’s sake, but to highlight shifts in thinking that new security engineers must experience to build on the past rather than reinventing it as the latest fad.

In their introduction to IC’s November/December special issue on future Internet protocols, editorial board members and guest editors Charles Petrie and Oliver Spatscheck review some false predictions of the Internet’s demise as

well as real challenges such as the exhaustion of IPv4’s address space. “Given how long it took to adopt IPv6,” they write, “it’s clearly not too early to start considering the issues and alternatives.” They introduce three articles that address the challenges of scalability in the routing domain, spam in IPv6, and design principles for evolving the transport protocol architecture.

In “Using Social Media to Enhance Emergency Situation Awareness,” from IS’s November/December issue, researchers from CSIRO ICT Centre and Palantir Technologies describe a system using natural language processing and data mining techniques to extract situation awareness information from Twitter messages generated during various disasters and crises. This type of on-the-ground information can help establish timely situation awareness and response times for a range of crisis types.

IT Pro features two articles on social computing in its November/December issue. In “E-Government Meets Social Media: Realities and Risks,” Rhoda C. Joseph of Pennsylvania State University analyzes 100 US e-government websites and relevant social media accounts to assess how state government portals and state governors are using social media. In “Internet Diffusion in China: Economic and Social Implications,” Linda S.L. Lai and Wai-Ming To of the Macao Polytechnic Institute explore the IT implications of the expected surge in e-commerce applications over the next three years.

Performance and total cost of ownership (TCO) are key optimization metrics in large-scale datacenters, and servers present a prime optimization target in the quest for more efficient operations. “Optimizing Data-Center TCO with Scale-Out Processors” is a Micro September/October

Computer Highlights Society Magazines

COMPUTER

ELSEWHERE IN THE CS



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 5

special issue article on energy-aware computing. European researchers describe a methodology they developed for designing performance-density-optimal server chips called scale-out-processors (SOPs). They present evaluation results showing the performance and TCO advantages of SOPs in fully leveraging existing software stacks in a nondisruptive manner.

“Presence” is the sensation of being in a virtual place, operationalized by responding to the virtual environment as if it were real. “Visual Realism Enhances Realistic Response in an Immersive Virtual Environment—Part 2,” a feature article in CG&A’s November/December issue, extends an experiment reported earlier in CG&A that compared presence in a virtual environment illuminated by either ray casting or real-time ray tracing. Part 1 concluded that ray tracing was the more effective technique. However, ray casting doesn’t produce dynamic changes such as shadows and refl ections, so Part 2 isolates the relative contributions of realistic lighting and dynamic effects to presence.

“The Computational Materials Repository” in CiSE’s November/December issue describes an eponymous soft-ware infrastructure for use in addressing the challenges of calculations based on quantum physics in the design of new materials. The CMR implements a modular framework in Python that provides tools for collecting, storing, group-ing, searching, retrieving, and analyzing the huge amounts of data that modern electronic-structure simulators gen-erate. The result of a collaboration under the Quantum Materials Informatics Project (www.qmip.org), the CMR is available under an open source license to any group or individual who might fi nd it useful.

Surveillance systems record and archive multiple streams of video 24 hours a day, every day. In “Posterity Logging of Face Imagery for Video Surveillance” from MultiMedia’s October-December issue, Italian academic and industry researchers report their work developing a robust face-logging system that associates an “identity” in the form of a high-resolution face image with each individual entering a surveillance area. Evaluation results show the system outperforming other face-logging methods described in the literature.

In PvC’s October-December Smartphones department, researchers from Microsoft, the University of California

at Santa Barbara, and Purdue University present “Swordfi ght: Exploring Phone-to-Phone Motion Games.” The authors describe this novel class of games, drawing inspiration from the Nintendo Wii and Xbox Kinect, and illustrate the challenges of utilizing localization techniques and existing hardware to facilitate a new class of mobile-to-mobile interaction.

Annals’ October-December issue is the fi fth in a special series on software history edited by Burton Grad and Luanne Johnson, cofounders of the Computer History Museum’s Software Industry Special Interest Group. “Relational Database Management Systems: The Formative Years” features six articles: an overview, two histories by eminent computer historians, and three recollections by industry pioneers about the companies they worked for—IBM, Oracle, and Ingres. Additionally, the Anecdotes department features “The Early History of SQL” by Donald Chamberlin, who received ACM’s SIGMOD Innovation Award for his work at IBM on SQL and System R.

Online Only

View Computer’s 2012 Annual Index:www.computer.org/computer/12index

Computer thanks our 2012 reviewers. View the list here:www.computer.org/computer/12reviewers

NEXT ISSUE

THE INTERNET OF THINGS



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


http://www.qmags.com/clickthrough.asp?url=www.qmip.org&id=17866&adid=P5E3

http://www.qmags.com/clickthrough.asp?url=www.computer.org/computer/12reviewers&id=17866&adid=P5E2

http://www.qmags.com/clickthrough.asp?url=www.computer.org/computer/12index&id=17866&adid=P5E1





The Unseen and the Seen

David Alan GrierIEEE Computer Society 2013 President

Our progress is measured by the accumulated effects of all the ideas, seen and unseen, that appear, only to vanish and then perhaps return at the proper time.

PRESIDENT’S MESSAGE


A presidential transition represents that moment when the invisible and the

visible switch places. A new president steps from the wings to head the organization. The outgoing leader takes a bow for work well done and returns to serve the Society in other ways.

The event tells us more about the organization than about any one individual who holds the office of president. It points to the constant renewal of the Society by its members, as they look to new tasks and new opportunities. Yet it is still useful to consider how the seen and the unseen switch places, how the organization’s accomplishments slide into the background and new problems take their place.

LOOKING BACKIn preparing for my 2013 presi-

dential term, I’ve been reading the messages of the individuals who held the presidency during the past four decades. I’ve seen how they celebrated the accomplishments of the group and prepared it for an unknown future. I recently read the message of Sam Levine, who served as chair of the Computer Working

Group in 1966 and 1967 and was one of the individuals who helped transform that working group into the modern IEEE Computer Society.

From our perspective, the switch that occurred at the end of Levine’s term is almost obvious. In the mid-1960s, the members of the Computer Working Group were devoting a lot of attention to the reliability of computing components. Their accomplishments in this field were substantial.

In Levine’s time, computing hardware was not sufficiently reliable. Many manufacturers admitted that their machines would run for only a couple dozen hours without failure and would need preventive maintenance weekly. None had the reliability that we currently demand from smartphones, tablets, and cloud systems.

Yet, within three or four years of Levine’s chairmanship, reliability began to slip into the background. Although it once had dominated the pages of the Computer Work-ing Group News, Computer’s predecessor, reliability slowly faded into the background. At the same time, another issue gradually became prominent: large software systems and software engineering.

The year that followed Levine’s presidency would see the two events that inaugurated the software age: the Garmish NATO Conference on Software Engineering, and IBM’s decision to separate software from hardware and allow other firms to create software for its machines. In 1970, when the Computer Working Group became the IEEE Computer Society, the software industry was already beginning to grow, and software engineering began to be a subject for academic study.

The growth of software was invisible to the Working Group leaders of the late 1960s, not because they couldn’t see it but because they couldn’t identify the most important issues among the many they faced. Software was just one topic on a list that included computer architectures, computer communication systems, solid-state circuits, memory and storage systems, analog computers, and simulation. However, Levine did indeed identify software as one of the many subjects that might be of interest to a new Computer Society. “The close relationship of software to hardware in the design of modern computer systems is evident,” he wrote.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


VI

DE

O





http://online.qmags.com/media/CMG/2013/January/video_cmg0113_06v1.swf

JANUARY 2013 7

In looking at Levine’s plans for the Society, it’s evident that the leadership intended to create an organization that would catch the important trends of the future when those trends were identified as visible entities. Levine wrote about technical committees and conferences, the need for new publications, and the value of standards. He argued for broadening the membership of the working group beyond “electrical engineers and physicists.”

In planning for the future, Levine argued for one activity that seems a little quaint in this age of the Internet and social media. He wanted to see a distinguished visitors program for universities. These visitors would be the ambassadors for the avant garde. They would spread the unseen ideas around the country.

A NEW ERAWe live in a different era than

Sam Levine and his peers. Our field is strong and mature. We’ve offered contributions to the world for more than 60 years and have seen our ideas move into fields quite far from our own. Indeed, we’ve pushed many IEEE entities toward engineering practices that more closely resemble those of software engineering than those that are used to develop circuits and power supplies.

Yet, at this point, we probably still need an avant garde of the unseen. We still need something to help us identify the ideas and problems that are not well understood but are about to become important. In a stable organization, it’s too easy to believe that something is of value because it supports our function or gives us status.

Time and again over the past six months, I’ve seen a new idea struggle toward visibility, temporarily blinding those who were witnessing its birth. A paper on market mechanisms in algorithms

caused one computer scientist to ask if this idea marked the end of traditional space/time complexity analysis. It seemed to the questioner that the research results implied that we need to worry about efficiency.

A second paper, closely following the first, claimed that we need to be concerned not only with space and time complexity but also with energy complexity. If energy eventually becomes the scarce item on the planet, do we need to make

sure that our programs produce the most calculations per watt?

In some cases, new ideas pass by without generating much real interest. In others, they produce real emotion. For example, at one point, I listened to a truly emotional speech by an older computer scientist who said he felt that computer science, the field to which he had devoted his life, was slipping away.

The world turns upon its axis and, to some extent, so does our field. New ideas appear, only to vanish and then return at the proper time. Our progress is measured by the accumulated effects of all these ideas, currently seen and unseen.

LEADERSHIP TASKSAs I look forward to this year,

I see three principle tasks for the IEEE-CS leadership.

The first task is to look into the future and grasp the ideas that are struggling to move into the light. Of course, the leadership isn’t always in the best position to identify new ideas. However, we can help the members understand the concepts that will dominate the field in the next decade in the way that software development dominated computer science in the decade that followed

Sam Levine’s presidency in 1966 and 1967.

Next, we need to understand our current position, the ideas we’ve mastered, and the tools we bring to the field. As I have begun to see more of the IEEE-CS, I have acquired a new appreciation of the depth of our expertise and the extent to which we dominate fields such as computer security. If we don’t fully understand our own strengths, we won’t be able to apply those ideas to

the next generation of problems. Finally, we need to ensure

that our accomplishments are well understood by those outside the Society, including other IEEE members, the people who work in the technical community, and the world at large. Our task is to make the unseen visible, so that those ideas can be well used.

In recognizing that societies such as ours involve a dialog between the accomplishments

of the past and the hopes of the future, I need to thank the leaders who are rotating out of their current IEEE-CS positions. In particular, I thank John Walz, who served as 2012 president, and Sorel Reisman, who’s finishing his term as past president. I also thank the vice presidents, who’ve served nobly, as well as the members of the Board of Governors.

David Alan Grier is an associate professor of international science and technology policy at George Washington University. He received a PhD in mathematical statistics from the University of Washington. Grier, an IEEE Fellow, is the author of several books. Contact him at [email protected].

We need to understand our current position, the ideas we’ve mastered, and the tools we bring to the field.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________






Embracing Digital Publishing

Ron VetterUniversity of North Carolina Wilmington

Efforts undertaken during the past two years have resulted in evolving Computer’s content for mobile applications so that readers can access information in a form that’s convenient for them.

EIC’S MESSAGE


H appy New Year! In many ways, we can say that in 2013, the future of publishing

is here.As I highlighted in previous

EIC messages (Jan. 2011, pp. 9-11; Jan. 2012, pp. 8-9), a primary goal for Computer has been to focus on improving mobile access to digital content. This goal has been achieved by augmenting traditional print content with multimedia to produce an enhanced digital version of Computer (www.computer.org/digitalcomputer).

At the same time, efforts were undertaken to produce a more interactive multimedia experience through the development of “tablet apps” for both the iOS and Android platforms. Computer’s iPad app (https://itunes.apple.com/us/app/ieee-computer/id492519269?mt=8), released in early 2012, had nearly 5,000 downloads by mid-December. The Android version just became available in December (https://play.google.com/store/apps/details?id=org.qmags.com.computer).

I am pleased to report that approximately 30 percent of

Computer Society members have switched to Computer’s digital version and that more than 4,500 copies of the iPad app were downloaded in 2012. During 2013, we plan to augment more articles with embedded multimedia content, including not only slide show presentations, audio, and video, but also interactive content such as animations and visualizations. In addition to enhancing the printed text, this new media content will also improve our readers’ experience with Computer.

Another exciting development in 2012 was the signing of an agreement with Borger, a publisher in China, to produce a print-only Chinese version of Computer. The goal is to launch this translated version in July 2013. Borger also produces Chinese versions of MIT Technology Review, The New York Times Science, and IEEE Spectrum.The agreement also allows the Computer Society to redistribute Computer’s translated content on the Computing Now Web portal, which will help to broaden Computer’s reach. I am excited about the potential for this new and unique distribution channel.

EDITORIAL BOARD CHANGES

Computer is fortunate to have so many dedicated volunteers who serve as members of the editorial board in a variety of capacities—whether as members of the advisory panel or as area or column editors. The masthead (see p. 1 of this issue) lists the people who help to make Computer possible each month. These volunteers and staff members are critical to providing the expertise that is required to produce such a high-quality monthly publication.

Each January, this EIC message offers an opportunity to thank those individuals have completed their terms of service to the publication and to welcome new members of the editorial board. Board members who retired at the end of 2012 include advisory panel member Richard G. Mathieu, computer architecture editors Tom Conte and Steven Reinhardt, software editor Robert France, and Forward Slash column coauthor Erin Dian Dumbacher. Each of these individuals has given many hours of time to Computer,and it is a better publication because of their contributions. I am greatly



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


__________

______________________

______

AU

DI

O

http://www.qmags.com/clickthrough.asp?url=https://play.google.com/store/apps/details?id=org.qmags.com.computer&id=17866&adid=P8E3

http://www.qmags.com/clickthrough.asp?url=https://itunes.apple.com/us/app/ieee-computer/id492519269?mt=8&id=17866&adid=P8E2

http://www.qmags.com/clickthrough.asp?url=www.computer.org/digitalcomputer&id=17866&adid=P8E1





JANUARY 2013 9

appreciative of their efforts—their advice and support will be greatly missed.

Computer is pleased to have welcomed several additions to the editorial board during 2012. Reneé Bryce from the University of North Texas is an area editor for software, and Greg Byrd from North Carolina State University and David Albonesi from Cornell University are area editors for computer architectures.

Two new column editors are beginning their editorial board tenure in 2013.

San Murugesan, of BRITE Professional Services, is the editor of Cloud Cover, a column that covers all aspects of cloud computing and its adoption and applications, including new developments, best practices, challenges, strategies, standards, regulations, trends, and perspectives. The column is intended to feature contributions from researchers, developers, cloud services providers and users, and executives.

Brian David Johnson from Intel is the editor of Science Fiction Prototyping, a column written as science fiction narrative to capture ethical, behavioral, economic, and other implications to provide insights into new and evolving technologies.

Finally, although David Alan Grier has retired as the coauthor of the Forward Slash column, he is continuing to serve as an editorial board member as the author of The Errant Hashtag. This monthly back-of-the-book column features stories from the front lines of computing technology: the people who create new ideas, the organizations that bring these ideas to the public, and the lessons we learn along the way. A multimedia podcast accompanies each column.

NEW BOARD MEMBER BIOSDavid Albonesi is a professor in the School of Electrical and Computer

Engineering at Cor-nell University. His research interests include adaptive and reconfigurable computer architec-

tures, power- and reliability-aware computing, and high-performance interconnect architectures using sili-con nanophotonics.

Albonesi received a PhD in computer engineering from the University of Massachusetts Amherst. He is an IEEE Fellow and has received the National Science Foundation CAREER Award, three IBM Faculty Awards, three IEEE Micro Top Picks paper awards, and the Michael Tien 72 Excellence in Teaching Award. Albonesi was editor in chief of IEEE Micro from 2007 to 2010 and currently serves on the editorial board of ACM Transactions on Architecture and Code Optimization.

Reneé Bryce is an associate professor in the Department of Computer Science at the University of North Texas in

Denton. Her research interests include software engineering, software testing, applications of combinatorial designs to software testing, software test suite prioritiza-tion, and Web application testing.

Bryce received a PhD in computer science from Arizona State University. In addition to having authored numerous journal articles, peer-reviewed conference and workshop papers, and a book chapter, Bryce has received active funding from the NationalScience Foundation and the National Institute of Standards and Technology to further her research.

Greg Byrd is a professor in the Department of Electrical and Computer Engineering at North Carolina State University. In addi-

tion to having authored numerous journal articles, peer-reviewed conference and workshop papers,

and book chapters in the area of computer architectures, Byrd previ-ously served as the associate editor of ACM Transactions on Embedded Computing Systems and is currently the IEEE Computer Society Publica-tions Board plagiarism chair.

Byrd received a PhD in electrical engineering from Stanford University. He is a senior member of IEEE and the IEEE Computer Society.

Brian David John-son, Intel’s Director of Future Casting, is the company’s first and only futurist. In this position, he is

charged with developing an action-able vision for computing in 2021. Johnson works with silicon platform and software architects to incor-porate key capabilities into future technologies and collaborates with researchers to develop an experi-ence road map that will provide Intel executives, strategic planners, and marketing groups with a clear and substantiated vision for platform direction.

Johnson received a BA in interdisciplinary communications studies from The New School for Social Research. In addition to speaking to representatives of academia, government, and industry about future casting, Johnson has written extensively about future technologies in articles and scientific papers as well as science fiction short stories and novels, and has di-rected two feature films on the topic.

The Tomorrow Project (www.tomorrow-projects.com), an ongoing international endeavor that explores possible futures with scientists, engineers, artists, and the general public, provides an overview of



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___

_______________

http://www.qmags.com/clickthrough.asp?url=www.tomorrow-projects.com&id=17866&adid=P9E1





DIGITAL MAGAZINESKeep up on the latest tech innovations with new digital maga-zines from the IEEE Computer Society. At more than 65% off regular print prices, there has never been a better time to try one. Our industry experts will keep you informed. Digital magazines are:

Easy to Save. Easy to Search.Email notifi cation. Receive an alert as soon as each digi-tal magazine is available. Two formats. Choose the enhanced PDF version OR the web browser-based version.Quick access. Download the full issue in a fl ash.Convenience. Read your digital magazine anytime, any-where—on your laptop, iPad, or other mobile device.Digital archives. Subscribers can access the digital issues archive dating back to January 2007.

Interested? Go to www.computer.org/digitalmagazinesto subscribe and see sample articles.

EIC’S MESSAGE

10 COMPUTER

Johnson’s work. Through the Tomorrow Project, Johnson has collaborated with Arizona State University, the University of Washington, and New Scientist as well as the Society for Science and the Public.

San Murugesan, a corporate trainer and consultant, is the director of BRITE Professional Services and an

adjunct professor at the University of Western Sydney, Australia. His expertise and interests include cloud computing, green IT, and IT for emerging regions.

Murugesan received a PhD in computers and automation from the Indian Institute of Science.

An IEEE Computer Society Distinguished Visitor who currently serves as an associate editor in chief for IT Professional, Murugesan is the author or coauthor of more than 220 publications, including peer-reviewed journal articles and conference papers, book chapters, and technical essays and reports.

GET INVOLVED Opportunities to become

involved with the IEEE Computer Society and Computer include becoming an author, serving as the guest editor of a special issue, or volunteering as a reviewer. Visit http://www.computer.org/portal/web/volunteercenter/getinvolved to find detailed information about these opportunities.

As always, I welcome your com-ments and encourage you to submit suggestions for topics to be covered in future issues of Computer.

Ron Vetter is the Interim Associate Provost for Research and Dean of the Graduate School at the University of North Carolina Wilmington and cofounder of Mobile Education LLC, a technology company that specializes in developing interactive short mes-sage service applications. Vetter has served on numerous journal editorial boards and conference committees, including his current appointment as an associate editor for Computing Now. Vetter received a PhD in com-puter science from the University of Minnesota, Minneapolis. Contact him at [email protected].



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________

_____________________

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org/portal/web/volunteercenter/getinvolved&id=17866&adid=P10E1

http://www.qmags.com/clickthrough.asp?url=www.computer.org/digitalmagazines&id=17866&adid=P10A1






COMPUTING CONVERSATIONS

0018-9162/13/$31.00 © 2013 IEEE Published by the IEEE Computer Society JANUARY 2013 11

Van Jacobson: Content-Centric Networking

A s engineers, every once in a while, we need to seriously revisit our underlying assumptions and make sure

they still hold true. In terms of the best architecture for the Internet, the four-layer model based on TCP/IP is pretty much accepted as absolute and unquestionable truth.

I recently interviewed Van Jacobson of PARC, and we talked of a major re-architecting of the Internet to deal with the fact that it is increasingly a global content distribution system layered atop a communication model, with computers making virtual “long-distance phone calls” to each other. What if they took a more content-centric approach?

Visit www.computer.org/computingconversations to view a video of this interview.

THE TELEPHONE MODELTo delve into a possible alternate

future, we must first let go of the notion that the current state-of-the-Internet architecture is “right” simply because we’re using it and it seems to work:

If you look at how the Internet has

evolved, it started as a telephone

Although the ARPANET and the Internet were very different from the telephone network in their implementation and use of physical wires, they were similar in that the ultimate goal of TCP/IP was to allow two distinct applications to “call” each other, get a connection, and let those applications have a conversation. This conversational model between applications was very general and allowed rich research into many different kinds of uses for the Internet. It kept the four-layer architecture pristine and avoided embedding application-specific understanding in routers:

This really changed the world, but

the bulk of that change didn’t happen

in the 1970s, 80s, or early 90s, when

the Net was first growing out—it

happened in the late 1990s and

2000s, when the Web took off. The

Web had nothing whatsoever to do

with computers having a conversation

model. It had to do with people

creating and consuming content.

The Web showed us for the first time

what happens if we leave behind this

18th century model of telephony

between applications, stop looking

at the wires, and instead focus on the

information in the wires.

Content-centric networking is much more than caching of content—it also works well for live streams of popular data.

Charles Severance

system for computers. People wanted

their computers to exchange data, and

the model we had for communication

for 140 years had been the

phone system. We said, “Okay,

communication is conversation

over long distances,” so let’s make

protocols and infrastructure

that allow computers to have a

conversation. The first cut of that

was the ARPANET—a network that

would handle different bandwidths

and didn’t require the global clock

distribution of a telephony network.

Instead, it substituted buffering.

The earliest telephone systems used the dialing of a phone number to physically configure relays and create a temporary “physical” wire that could transmit amplified analog audio signals over long distances:

A crucial thing coming out of Paul

Baran was to not emulate the phone

system, where communication was all

about building a wire hop-by-hop or

link-by-link between two end points,

basically instructing the switching

system how to make one long wire.

Instead, Paul said, “Just identify the

end points and let the network take

care of getting the data there.”



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


________________

VI

DE

OA

UD

IO

http://www.qmags.com/clickthrough.asp?url=www.computer.org/computingconversations&id=17866&adid=P11E1






COMPUTING CONVERSATIONS

12 COMPUTER

TCP/IP is architected around the notion that the network number portion of an IP address connects to one and exactly one router in the global Internet. To support world-scale applications, it was increasingly necessary to “lie” to TCP/IP about what’s really happening:

If you look at YouTube, Google,

Amazon, Facebook, Twitter, all of

these very heavily used services that

manifest themselves to the Net as

an IP address that looks like a single

location, if they’re a single location

with hundreds of millions of users,

the traffic in a conversational model

always scales like the popularity, it

scales like the number of consumers.

If you’re doing a Twitter update or

creating a video that will be seen by

millions, you can’t deploy it in a pure

conversational model, and so you’re

forced to spend all your time fooling

the Net.

In a world-scale Web application like Google, the destination IP address in a packet has little to do with your request’s ultimate destination. The address is simply the quickest way to get your request into a nearby datacenter, where Google looks more deeply into the request to figure out what you want so it can virtually route it to the closest copy of the requested information:

Information that’s qualitatively the

same is spread randomly across

the whole packet. We have source

Content-centric networking is much more than caching content—it also works well for live streams of popular data.

The Web gave us a way to name information, and representational state transfer (REST)-style Web services gave us a namespace for data as content. Today, we enter a URL into our browser to indicate “what we want” instead of “how to get it.” We layer the HTTP protocol atop a conversational model, but as moving data and content becomes the dominant use of Internet infrastructure, perhaps the conversational model is becoming our limiting factor:

We’re having massive scaling

problems today trying to join

together the very information-

centric Web model with the very

host-centric TCP/IP model.

Look at what it takes to build

something like YouTube. You can

create videos and put them on your

own website, but you have to pray

that they never become popular.

If they get popular, your ISP will

almost immediately shut you down

because your link will be completely

saturated, in what we call the

“Slashdot effect.”

That problem is intrinsic to the

conversational model: we don’t do

broadcast television by making

phone calls to the television

station because there is no way to

scale that. We broadcast it out to

everybody who wants to listen, and

we don’t know who they are—they

aren’t individually identified.

WORLD SCALEAs world-scale applications and

services became the norm on the Internet, it was impossible to have a single connection from something like Google to the rest of the Internet and route all traffic from around the world in a single network connection to a single server room in Mountain View, California. For Google to function effectively, it needs many facilities around the world and many connections between its facilities and the public Internet.

and destination addresses that we

conventionally think of as IP, so

it’s at the front of the packet to be

used by the network layer. Then we

have ports that are TCP, so they’re

a little bit deeper in because they’re

supposed to be used by the end node

for its demultiplexing to get you

to a particular application. Inside

that, we have sequence numbers

that are used when you get to the

application to reassemble the larger

unit of information, and inside that,

we have URLs, which are used by a

higher-level part of the application for

session meaning and the like.

You just have all this information,

and it’s all fundamentally name

information that indicates “What do

these bits mean?” If you pull together

the source addresses, the ports, the

sequence numbers, and the URLs,

they all give you context. They’re all

the “name” of the information.

What if I just said that packets

have a name on the front and all that

information gets collected to the

name? At any point in the network,

you look at the name to do your job. If

just the front part will work because

all you’re doing is gross-level steering,

just look at the front. If you need

to look at more of it, we don’t have

layers, we have a set of structured

information, and we know that we’ll

be looking at different parts of it for

different reasons.

CONTENT IS KINGOnce we switch to a naming

scheme that uniquely identifies each packet or segment of content, it no longer matters where the data actually comes from. The content segment could come as easily from a nearby router as from the originating source. Packets could then be cached throughout the network in the memory of the routers, and those packets could be reused for popular content.

Content-centric networking is much more than caching of content—it also works well for live



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 13

Corporate Affiliate ProgramIncreases technical training while cutting costs.

For more information, call 1-855-727-3632 or email us at [email protected]

Provides company-wide, employee access to 4,300 technical courses, 600 technical and business books, dozens of Brainbench Exams and free or discounted training webinars and software development certifications.

Antipolis, France, in September 2012, with 29 presentations, nine demonstrations, 16 poster sessions, and attendees from many different organizations. You can find more information on CCN and related projects at www.ccnx.org.

Perhaps the next time you have a little free time while you wait for a YouTube video to buffer, you might let your mind wander and imagine for a moment how you might re-engineer the Internet architecture to better handle our increasingly content-oriented use of the network.

Charles Severance, Computing Conversations column editor and Computer’s multimedia editor, is a clinical associate professor and teaches in the School of Information at the University of Michigan. Follow him on Twitter @drchuck or contact him at [email protected].

into that data, now you have a copy

and now you’re done with your

distribution.

Of course, it’s one thing to pos-tulate that we need a fundamental paradigm shift in the architecture and yet another to move a new approach into broad worldwide production. Like the shift from voice to data communications from the 1960s to the 1990s, shifting data communications from a conversa-tional model to a content-centric one will also take time and require many experiments. And like the engineer-ing of TCP/IP itself, there will likely be many versions as researchers identify new issues and use cases.

T he good news is that an active community is exploring CCN and its applications worldwide.

The CCN community met in Sophia

streams of popular data. One of the many prototype applications built on top of CCN’s early implementations is a multiuser/multipoint video conferencing system. As the popularity of a live video conference or event scales, the likelihood of packet reuse increases dramatically:

If you don’t care where you’re getting

the data—if all that matters to you is

what the data is, not where it comes

from—then all of this memory that

has to be in the network as buffering

in order to manage the multiplexing

suddenly becomes a viable source of

data.

Having the load scale with popu-

larity is strictly a function of the fact

that the data can only originate from

one place. If you just care about

the data, just start going toward

that place, and as soon as you run



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___________

_______________

http://www.qmags.com/clickthrough.asp?url=www.ccnx.org&id=17866&adid=P13E1


mialto:[email protected]





PURPOSE: The IEEE Computer Society is the world’s largest association of computing professionals and is the leading provider of technical information in the field.MEMBERSHIP: Members receive the monthly magazine Computer, discounts, and opportunities to serve (all activities are led by volunteer members). Membership is open to all IEEE members, affiliate society members, and others interested in the computer field.COMPUTER SOCIETY WEBSITE: www.computer.orgOMBUDSMAN: To check membership status or report a change of address, call the IEEE Member Services toll-free number, +1 800 678 4333 (US) or +1 732 981 0060 (international). Direct all other Computer Society-related questions—magazine delivery or unresolved complaints—to [email protected]: Regular and student chapters worldwide provide the opportunity to interact with colleagues, hear technical experts, and serve the local professional community.AVAILABLE INFORMATION: To obtain more information on anyof the following, contact Customer Service at +1 714 821 8380 or+1 800 272 6657:

Membership applicationsPublications catalogDraft standards and order formsTechnical committee listTechnical committee applicationChapter start-up proceduresStudent scholarship informationVolunteer leaders/staff directoryIEEE senior member grade application (requires 10 yearspractice and significant performance in five of those 10)

PUBLICATIONS AND ACTIVITIESComputer: The flagship publication of the IEEE Computer Society, Computer, publishes peer-reviewed technical content that covers all aspects of computer science, computer engineering, technology, and applications.Periodicals: The society publishes 12 magazines, 15 transactions, and one letters. Refer to membership application or request information as noted above.Conference Proceedings & Books: Conference Publishing Services publishes more than 175 titles every year. CS Press publishes books in partnership with John Wiley & Sons.Standards Working Groups: More than 150 groups produce IEEE standards used throughout the world.Technical Committees: TCs provide professional interaction in more than 45 technical areas and directly influence computer engineering conferences and publications.Conferences/Education: The society holds about 200 conferences each year and sponsors many educational activities, including computing science accreditation.Certifications: The society offers two software developer credentials. For more information, visit www.computer.org/certification.

NEXT BOARD MEETING6–8 Feb., 2013, Long Beach, CA, USA

EXECUTIVE COMMITTEEPresident: David Alan GrierPresident-Elect: Dejan S. MilojicicPast President: John W. WalzVP, Standards Activities: Charlene (“Chuck”) J. WalradSecretary: David S. EbertTreasurer: Paul K. JoannouVP, Educational Activities: Jean-Luc GaudiotVP, Member & Geographic Activities: Elizabeth L. Burd (2nd VP)VP, Publications: Tom M. Conte (1st VP)VP, Professional Activities: Donald F. ShaferVP, Technical & Conference Activities: Paul R. Croll2013 IEEE Director & Delegate Division VIII: Roger U. Fujii2013 IEEE Director & Delegate Division V: James W. Moore2013 IEEE Director-Elect & Delegate Division V: Susan K. (Kathy) Land

BOARD OF GOVERNORSTerm Expiring 2013: Pierre Bourque, Dennis J. Frailey, Atsuhiro Goto, André Ivanov, Dejan S. Milojicic, Paolo Montuschi, Jane Chu Prey, Charlene (“Chuck”) J. WalradTerm Expiring 2014: Jose Ignacio Castillo Velazquez, David. S. Ebert, Hakan Erdogmus, Gargi Keeni, Fabrizio Lombardi, Hironori Kasahara, Arnold N. PearsTerm Expiring 2015: Ann DeMarle, Cecilia Metra, Nita Patel, Diomidis Spinellis, Phillip Laplante, Jean-Luc Gaudiot, Stefano Zanero

EXECUTIVE STAFFExecutive Director: Angela R. BurgessAssociate Executive Director & Director, Governance: Anne Marie KellyDirector, Finance & Accounting: John MillerDirector, Information Technology & Services: Ray KahnDirector, Membership Development: Violet S. DoanDirector, Products & Services: Evan ButterfieldDirector, Sales & Marketing: Chris Jensen

COMPUTER SOCIETY OFFICESWashington, D.C.: 2001 L St., Ste. 700, Washington, D.C. 20036-4928Phone: Fax: +1 202 728 9614Email: [email protected] Alamitos: 10662 Los Vaqueros Circle, Los Alamitos, CA 90720Phone: +1 714 821 8380Email: [email protected]

MEMBERSHIP & PUBLICATION ORDERSPhone: Fax:Asia/Pacific: Watanabe Building, 1-4-2 Minami-Aoyama, Minato-ku, Tokyo 107-0062, JapanPhone: Fax: +81 3 3408 3553Email: [email protected]

IEEE BOARD OF DIRECTORSPresident: Peter W. StaeckerPresident-Elect: Roberto de MarcaPast President: Gordon W. DaySecretary: Marko DelimarTreasurer: John T. BarrDirector & President, IEEE-USA: Marc T. ApterDirector & President, Standards Association: Karen BartlesonDirector & VP, Educational Activities: Michael R. LightnerDirector & VP, Membership and Geographic Activities: Ralph M. FordDirector & VP, Publication Services and Products: Gianluca SettiDirector & VP, Technical Activities: Robert E. HebnerDirector & Delegate Division V: James W. MooreDirector & Delegate Division VIII: Roger U. Fujii

revised 4 Dec. 2012



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________

_____________

__________

___________

_______

http://www.qmags.com/clickthrough.asp?url=www.computer.org/certification&id=17866&adid=P14E2

http://www.qmags.com/clickthrough.asp?url=www.computer.org&id=17866&adid=P14E1









COMPUTING AND THE L AW


Protecting Bioinformatics as Intellectual Property

S cientists and software devel-opers often identify patent law as the area of intellec-tual property with which

they have at least some familiarity. Sometimes their experience has been positive; for example, a patent holder has been able to use a patent to stop a competitor from misusing, or copying, a product or a process that he has expended significant amounts of time and money devel-oping. For others, the experience has persuaded them that monopoly rights stifle innovation and protect established market participants at the expense of emerging businesses.

However, with regard to protect-ing bioinformatics as intellectual property—using the laws of trade secrets, copyright, and patents, for example—the debate some-times expands to include certain ethical considerations that arise because of the potential therapeu-tic benefits of this technology.

Visit the IEEE Computer Society’s website for the podcast that

With approximately 30,000 genes in the sequence, and around 18,000 megabases (1 million genetic base pairs), the data scale is remarkable.

Without a computer, a researcher could never review even a single complete sequence, much less analyze it for patterns or anomalies, compare one person’s genome to that of another, or compare it to genomes of other organisms. The cost of sequencing has also fallen more quickly than Moore’s law alone would suggest: whereas the cost of sequencing a megabase was just over $5,000 in September 2001, by January 2012, the cost had fallen to just 9 cents. By combining in vitro and in silico research, researchers have accessed databases of previously sequenced genomes and other biological data to analyze and compare that data, and they have done this quickly, accurately, and cost-effectively.

Using bioinformatics has several key consequences for medical research and disease control. For

Although trade secrets and copyright provide some protection to bioinformatics tools, a carefully drafted patent can provide the broadest available protection in an increasingly competitive market.

Brian M. Gaff, Ralph A. Loren, and Gareth DicksonEdwards Wildman Palmer LLP

accompanies this article (www.computer.org/portal/web/computingnow/computing-and-the-law).

WHAT IS BIOINFORMATICS?Broadly speaking, the term

bioinformatics refers to the use of information technology in the analysis and organization of data relating to biology. Researchers have pursued the development of such technology for decades, and its necessity for accessing the rapidly expanding data made available in the postgenomic era is evidenced by information from the Human Genome Project (HGP) itself.

Launched in 1990, the HGP set out to sequence the entire human genome by 2005, at an estimated cost of US$3 billion. It did so more quickly than projected, and under budget, and a working draft of the genome was announced in 2001; the finished sequence—more than 330 times longer than the working draft—was made available in 2003.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________________

_____

AU

DI

O

http://www.qmags.com/clickthrough.asp?url=www.computer.org/portal/web/computingnow/computing-and-the-law&id=17866&adid=P15E1





COMPUTING AND THE L AW

16 COMPUTER

courts in the US and Europe indicate that data formats can’t get their own protection under copyright law.

As with other intellectual property rights, rights in databases and the information they contain grant their owners the ability—although not the obligation—to exclude others from using their works without permission. An HGP operating principle is that the data and resources the project generates should be made available to researchers. Nonetheless, the right to exclude people from performing particular acts arises automatically, irrespective of whether the owner intends to exercise it.

Software toolsThe bioinformatics tools

themselves can typically be protected like software—by using trade secrets, copyright, and patents. Trade secrets are not well-suited to business models that distribute to users any product that embodies or incorporates the secret, rather than licensing that product’s use as a service through the cloud.

Copyright law also has limitations when it comes to software. Subject to very limited exceptions, copyright doesn’t protect a program’s functionality or design; because it covers only original expression, only the GUI and the source code can be protected. This means that if code has been rewritten such that it’s no longer substantially similar to the original code, a copyright claim will likely fail. Similarly, anyone seeking

Patents offer the broadest protection since they are the only true monopoly over the patented work and don’t need to be copied to be infringed.

example, researchers can use an algorithm to identify a similarity between one gene sequence for which the function is known and another gene sequence for which the function is being investigated. Using this information, researchers can infer that the second sequence might have a function similar to the first. In the study of protein and channel structures, the understanding of which is fundamental to the targeting of drugs, homology dictates that it’s possible to predict a protein’s structure, and therefore the effect of a particular drug on it, using a combination of experimental data and a homologous protein’s known structure.

AVAILABLE PROTECTIONBioinformatics represents a

significant force in education, personalized medicine (or biological warfare), and commerce. As bioinformaticians avail themselves of the ability to tap into unused CPU capacity from anywhere in the world through online processor exchanges, bioinformatics will do more at a lower cost, and will become increasingly competitive as a business model. Yet the availability of bioinformatics protection is complex.

DatabasesDatabases such as the US

National Institutes of Health’s GenBank, which contains electronic records of the sequence data that the HGP generated, can be protected in their own right. In the US, they can be protected under copyright law as compilations. In Europe, they can be protected by copyright or by a separate database right that prevents the unauthorized extraction or reutilization of the whole or a substantial part of the database. The stored data format can be more of a practical issue than a legal one: recent decisions from

to enforce a copyright must also demonstrate copying of the work: if the defendant can show that he created his work independently, then the copyright claim will fail.

Of these three rights, patents offer the broadest protection since they are the only true monopoly over the patented work and don’t need to be copied to be infringed. As such, they’re attractive to those seeking the right to prevent others from using an invention without permission. Software patents, however, are the most difficult to obtain.

The US and the European Union, as well as individual EU members such as the UK, France, and Germany, are all signatories to the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), which aims to harmonize the fundamentals of various IP rights, including patent rights. TRIPS provides that patent protection of about 20 years shall be available for any inventions, whether products or processes, that are new, involve an inventive step (that is, are not obvious), and have industrial applications. These same criteria are to apply “in all fields of technology.”

Notwithstanding this, Europe treats software inventions differently than the US. In Europe, inventions that are computer programs “as such” are excluded from patentability. An applicant must instead claim more than just a program for a computer: at the European Patent Office, it’s likely sufficient that the invention be tied to hardware; but in the UK, an application must describe to a person skilled in the art how the computer-implemented invention achieves a “technical effect” before it can proceed to be examined for novelty, inventive step, and industrial application.

Where bioinformatics programs are concerned, a particularly efficient method of identifying



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






IEEE Computer Graphics and Applications is indispensable reading for people who want to

stay current on the latest tools and applications,gain invaluable practical and research knowledge, andread objective and trustworthy content.

www.computer.org/cga

JANUARY 2013 17

intellectual property protection enables these institutes to protect their products and processes, but it doesn’t oblige them to do so.

Brian M. Gaff is a senior member of IEEE and a partner at the Edwards Wildman Palmer LLPlaw firm. Contact him at [email protected].

Ralph A. Loren is a partner at the Edwards Wildman Palmer LLP law firm. Contact him at [email protected].

Gareth Dickson is an associate at the Edwards Wildman Palmer LLP law firm. Contact him at [email protected].

to note that TRIPS only provides for signatories to exclude patent protection for inventions for which commercial exploitation would run counter to public policy or morality. There’s no public policy or moral objection to patentability where a patent-holder refuses to allow an invention to be commercially exploited.

A lthough trade secrets and copyright provide some protection to bioinformat-

ics tools, a carefully drafted patent can provide the broadest available protection in an increas-ingly competitive market. Many research institutes, however, have no desire to exclude others from using tools that are essential in furthering their shared aims of improving health and well-being.

Many, such as those involved in the HGP, have made it a priority to ensure that the ethical, legal, and social implications of their work are not ignored. The challenge, however, is to ensure that those who invest time and money in developing these tools can continue to do so, without hindering the essential work they and their competitors do. Securing

patterns in a sequence might be patentable, but the mere use of a computer to perform a mental calculation is clearly not.

In the US, the Supreme Court’s 2010 decision in Bilski v. Kapposdidn’t recast software such as bioinformatics as subject matter that’s unpatentable (www.supremecourt.gov/opinions/09pdf/08-964.pdf). Indeed, the Supreme Court broadened the test for what’s patentable subject matter by stepping back from the so-called machine or transformation test, which could have limited the patentability of bioinformatics in the US. The Supreme Court stressed that additional criteria would be used to test for patentability and, at this point, those criteria have not presented any significant roadblocks to patenting bioinformatics.

Assuming that an invention is not excluded from patent protection, a practical challenge for the bioinformatics industry is that the examiner must review the application from the point of view of a person skilled in the art. It’s unlikely that large numbers of patent examiners are themselves skilled in the art of bioinformatics. Therefore, applicants in both the US and Europe would do well to make their applications as clear as possible, with both jurisdictions guarding against patents for abstract ideas by requiring applicants to disclose “specific, substantial, and credible” uses for the claimed invention.

ETHICAL ISSUES UNDER TRIPS

One issue concerns the ethical objection to governments granting entities the right to prohibit other entities from using inventions that have the potential to save lives and alleviate suffering.

Leaving to one side the issue of whether compulsory licensing can meet this objection, it’s interesting

The content of this article is intended to provide accurate and authoritative information with regard to the subject matter cov-ered. It is offered with the understanding that neither IEEE nor the IEEE Computer Society is engaged in ren-dering legal, accounting, or other professional services or advice. If legal advice or other expert assistance is required, the services of a competent professional person should be sought.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________________

___________________

______________

_____

_________________

http://www.qmags.com/clickthrough.asp?url=www.computer.org/cga&id=17866&adid=P17A1

http://www.qmags.com/clickthrough.asp?url=www.supremecourt.gov/opinions/09pdf/08-964.pdf&id=17866&adid=P17E1








32 & 16 YE ARS AGO


JANUARY 1981www.computer.org/csdl/mags/co/1981/01/index.html

PRESIDENT’S MESSAGE (p. 5) “… A few of the more significant achievements [of outgoing president, Tse-yun Feng] include the initiation of two new magazines aimed at a broad spectrum of our membership, purchase of a building to house our West Coast operations, inauguration of Tutorial Weeks to bring the latest technical information in an inexpensive way to our membership, establishment of the Computer Society Press to better support non-periodical publications, and carrying out a much-needed rewriting of our bylaws to reflect the present structure and organizational requirements.”

HANDICAPPED COMPUTING (p. 9) “The United Nations has designated 1981 the International Year for Disabled Persons. It is therefore appropriate that Computer is dedicating its first issue of the year to computing and the handicapped. The evolution of low-cost computing technology promises an unprecedented spectrum of opportunities for the developmentally disabled. From aids to independent living to flexible tools that can greatly increase the variety and quality of job opportunities, the rapidly evolving field of computing is pregnant with possibilities. …”

JOBS FOR THE HANDICAPPED (p. 12) “The National Institute of Handicapped Research has been involved in developing programs and adapting computer technology to solve the problems of handicapped individuals. Our research and training center at Northwestern University, for example, has developed and is testing a sophisticated office environment in which a quadriplegic can effectively use a computer-controlled system to handle the telephone, typewriter, information storage bank, and tape recorder. As the system becomes fully marketable for use in other office settings, the opportunity to compete fairly in the job market will provide a great boost to the working-age handicapped public.”

HANDICAPPED EDUCATION (p. 15) “Traditionally, the public educational system has had the responsibility of preparing citizens for their roles as contributing members of society; the handicapped are now, legally as well as morally, a part of that responsibility. But the educational community cannot perform this monumental task by itself—the pooled, cooperative effort of many disciplines will be required to make life more meaningful for the handicapped. This article addresses the need for a joint effort by the educational and computer communities.”

COMMUNICATION DEVICES (p. 25) “At the Biomedical Engineering Center of Tufts-New England Medical Center … we are trying to address the communication problems

o f d i s a b l e d people through a c o h e r e n t p h i l o s o p h y of electronic commu n ic a-t i o n d e v i c e d e v e l o p m e n t that stresses the personal, portable, and affordable as key concepts. Intertwined with these three concepts, although not as easily defined, is a fourth—speed.”

HANDICAPPED PROGRAMMERS (p. 49) “A lmos t a ll graduates of these [rehabilitation] programs are placed in professional positions and become successful programmers; since 1973 over 300 seriously handicapped persons … have become self-supporting. However, without the availability of recent advances in computer technology and rehabilitation engineering, training of the severely handicapped could not have been considered.”

MICROCOMPUTER AIDS (p. 54) “The potential of the microcomputer as a rehabilitation tool, however, is currently limited by a number of system and configurational constraints. … Removal of these constraints would permit application of microcomputers to a much broader range of rehabilitation problems, including those faced by the severely and multiply handicapped.”

COMPUTING AND UNIVERSITIES (p. 74) “Universities are not as well suited to teaching many topics, including software engineering, computer-aided design/manufacture, and integrated circuit manufacture, as are large, technically proficient companies. During the last five years, major universities have begun tentative efforts in these subjects, but companies such as IBM, Texas Instruments, General Electric, and AT&T have had extensive in-house training programs in these areas for many years. Perhaps the universities would do well to leave such efforts to the large companies.”

BLIND PROGRAMMERS (p. 92) “Five years ago, Arkansas Enterprises for the Blind, in Little Rock, started offering a computer programming course on an individual, informal basis; today, the course is part of its regular curriculum.

“As a result, 17 blind and visually impaired graduates of the course are now employed in programming positions across the country; two members of the most recent class of eight graduates have already found employment, and the others are seeking jobs. Job placement so far has been 100 percent.”



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


http://www.qmags.com/clickthrough.asp?url=www.computer.org/csdl/mags/co/1981/01/index.html&id=17866&adid=P18E1





Editor: Neville Holmes; [email protected]

JANUARY 2013 19

JANUARY 1997www.computer.org/csdl/mags/co/1997/01/index.html

EDITOR IN CHIEF’S MESSAGE (p. 8) “This is also the 51st year for the Computer Society. We thought it would be fun to project ahead a little bit as to what the next 50 years might bring, and so we put together a collection of essays. Here I’d like to take a little more space to project the future of Computer in the near future and to once again ask for your participation.”

THE NEXT 50 YEARS (p. 16) “As the Computer Society begins its 51st year, some computing pioneers, past and present leaders of the Computer Society, and members look ahead to what the next 50 might bring.”

ENCRYPTION AND PRIVACY (p. 28) “Hewlett-Packard has developed the International Cryptography Framework (ICF), the strongest encryption system approved for export by the US government to date. The hardware-software system provides various levels of encryption strength, depending on government regulations in the US and importing countries.”

“However, some say ICF may prove to be a Trojan horse in the war over electronic communications privacy. They contend that millions of people will buy ICF-enabled equipment that has weak encryption and that would permit the US government to decrypt data transmissions easily.”

A FASTER INTERNET (p. 31) “Although data indicates that Net traffic generally moves at a good clip ..., it is clear that many people want faster data-transmission speed now and that many others are worried about Internet speed in the future.”

INDUCTION (p. 36) “The science of creating software is based on deductive methods. But induction, deduction’s ignored sibling, could have a profound effect on the future development of computer science theory and practice.”

NUDGE, NUDGE (p. 42) “With busy users relying increasingly on computers to provide reminders and alerts, the enabling technology is rapidly gaining importance. Surprisingly, many issues have not yet been thoroughly explored.”

THE FUTURE INTERNET (p. 50) “Although it is clear that the Internet will have substantial influence in the larger information infrastructure by the turn of the century, it is unclear how the Internet will realize that influence. Will the Internet continue to grow? Will the Internet’s definition expand to include other internetworked entities? How will protocol changes transform the Internet from the inside out? …”

INTERNET SECURITY (p. 57) “… The Internet continues to grow rapidly, but it is becoming increasingly segmented as companies resort to firewalls and intranets. If this continues, the concept of a global network will give way to many islands of private intranets only partially connected to a global structure. …”

HYPERMEDIA (p. 62) “Many organizations will embrace the World Wide Web as their primary application infrastructure. However, in the rush to acquire and retrofit Web applications, they risk bypassing the Web’s greatest supplemental benefit—hypermedia.”

OVERLOAD (p. 71) “Virtually cost-free publication on the Web has led to information overload. AI, with its roots in knowledge representation, is experiencing a renaissance as new tools emerge to make the Web more tractable.”

THE GORDON BELL PRIZE (p. 80) “New hardware and an increase in the scale of an existing machine enabled contestants to nearly double last year’s results in the performance and the price/performance categories.”

INTERNET STANDARDS (p. 114) “The Internet Engineering Task Force—the primary organization for Internet standards—is experiencing phenomenal growth, which implies a vote of confidence in its approach to standards-making. …

“… One remarkable feature of the IETF standards process is its demand for actual implementations of a standard. In fact, the existence of two independently developed, interoperable implementations of a specification is a firm requirement for the specification to be given draft standard status.”

RELIABLE SOFTWARE (p. 129) “… Design by contract … the principle that interfaces between modules of a software system … should be governed by precise specifications, similar to contracts between humans or companies. The contracts will cover mutual obligations (preconditions), benefits (postconditions), and consistency constraints (invariants).

THE YEAR I SHOOT MY TV (p. 136) “Consumerism is transforming our industry, prying it away from its business and academic roots and propelling it toward consumer electronics. … Computer companies may be influenced more by product managers in the entertainment and consumer market than by hard-core MIS managers.”



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


______________

http://www.qmags.com/clickthrough.asp?url=www.computer.org/csdl/mags/co/1997/01/index.html&id=17866&adid=P19E1






TECHNOLOGY NEWS


Bringing Big Analytics to the MassesNeal Leavitt

Businesses have greatly ben-efited from data analytics. Companies analyze activities such as sales, marketing,

fraud and risk management, inven-tory optimization, and customer support to improve their strategic and tactical business decisions.

However, analytics that is power-ful enough to work with big data has been too expensive, complex, and compute- and resource-intensive for smaller companies to use.

Now, though, vendors and service providers are leveraging advanced data-processing, network, and cloud technologies to make big analytics suitable for use by such organizations. Some vendors provide analytics software, while others offer services via the cloud.

The stakes are enormous. Market research firm IDC said the big data market is growing about 40 percent annually, with revenue expected to rise from $3.2 billion in 2010 to $17.2 billion in 2015.

Deloitte Touche Tohmatsu, a professional services firm, predicts that about 90 percent of Fortune 500 companies will have a big data initiative in place this year.

A 2012 study by IBM and The Economist news magazine found that firms that apply analytics outperform their peers that don’t.

As the “Small Companies and big analytics” sidebar explains, small and mid-sized organiza-tions could benefit from analytics as much as large ones, said Robin Bloor, CEO of the Bloor Group, a market research firm. Such companies are struggling with data volumes growing 62 percent annually, according to market research firm Aberdeen Group.

However, these businesses haven’t been able to benefit from high-powered analytics and therefore can’t make the most out of their information, said Todd Papaioannou, founder and CEO of Continuuity, a big data application vendor.

“The major inhibitors for smaller companies are cost and a lack of skilled personnel,” said Aberdeen research analyst Nathaniel Rowe.

Most of these firms don’t have the necessary server, software, and network infrastructures to sufficiently process big data, added Chris Piedmonte, CEO of analytics start-up Suvola.

Now, though, advances in developing applications that work with mobile and other smaller systems have made it possible for many organizations to use big analytics software for the first time.

“Cloud-based analytics solu-tions provide a more accessible price point for adoption,” noted Ed Abrams, vice president of IBM Midmarket Business.

BIG-TIME ANALYTICSBig data products typically

require a large IT staff to administer. They also use multiple expensive servers and include software that can be complex to set up and maintain. Big data software tends to use specialized programming languages such as Apache HiveQL, R, and Python.

Another issue is that big analytics is most valuable when used to study many different information types and sources, said Aberdeen’s Rowe.

Because organizations need different systems for structured and unstructured data to get the best results, they would have to buy multiple products, noted Leena Joshi, senior director

Big data analytics is valuable to many companies but has been too complex and expensive for smaller businesses. This is beginning to change.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 21

Software-based approachIn the past, only larger

organizations had the money to afford big analytics software, as well as the server infrastructure and skilled personnel to run it.

Now, companies such as Suvola offer integrated systems using hardware and software from selected vendors. This lets smaller organizations buy simpler, more affordable all-in-one systems for which the seller provides maintenance and support.

Also, the desire by many companies to use their smartphones and tablets to run powerful business software is inspiring developers to make big analytics and other applications that can run on such devices, noted Mark Levitt, director of business cloud research with market research firm Strategy Analytics.

Unlike desktop software designed to fill large screens with lots of menu options for experienced power users, developers are designing mobile applications to present the most relevant subset of available information and actions based on the user’s needs.

This has yielded analytics software that can run on smaller server systems, as well as on wireless devices.

Vendors offering analytics applications for smaller organizations include big vendors

of solutions marketing for big analytics vendor Splunk.

Added Robert Ducher, vice president of marketing for analytics vendor InsightsOne, smaller companies have trouble finding employees to work with big analytics, which uses complex and advanced technology, is a relatively new field, and is changing rapidly. People with the necessary skill sets are typically scarce and demand high salaries,

All these factors have made big analytics products suitable only for large organizations.

A common big analytics approach is Apache Hadoop, an open source framework based on Google’s MapReduce program-ming model for breaking large, complex problems into small chunks. Hadoop uses the process-ing power of clusters of ordinary servers to tackle tasks too large for any of the individual machines.

However, the extensive Hadoop storage and analytics infrastructure is costly to acquire and set up, and it requires specialized talent and ongoing support and maintenance.

“Adding data integration, business intelligence, and devel-oper tools results in a large and complex software stack,” said Omer Trajman, vice president of field operations for big data-applications vendor WibiData.

Another barrier for smaller organizations is that the Hadoop APIs are difficult for most programmers to use and there aren’t good tools for deploying, running, and monitoring appli-cations, added Continuuity’s Papaioannou.

Large companies some-times build complex in-house machine-learning tools to con-duct analytics, said Jos Verwoerd, cofounder and vice president of business development for machine-learning vendor BigML.

This is prohibitively expensive for smaller organizations, noted Indiana University assistant professor Chung-chieh Shan.

ANALYTICS FOR ALLHandling big data, noted

InsightsOne’s Ducher, requires a lot of storage and good networks. And analytical techniques such as machine learning use considerable compute power.

However, the cloud has helped make big analytics possible for smaller businesses by letting service providers handle the analytics.

Customers could use today’s inexpensive storage to collect large amounts of data, which they could then send to providers via widely available high-bandwidth networks.

And smaller organizations that want to run analytics software themselves can do so on more powerful servers than in the past because of technical advances in lower-cost multicore processors.

However, cloud-based services have become the more popular route because they’re scalable and relatively inexpensive, and they don’t require a company to have employees with data-analytics-related skills.

“Cloud-based analytics come ready out of the box, with connectivity to existing systems and standard algorithms and reports available,” said WibiData’s Trajman.

COMPANIES AND BIG ANALYTICS

T here are many ways that small companies could use big analytics, even though they don’t have nearly the scope of operations found in bigger organizations.

Currently, smaller companies are using big analytics to target customers by capturing useful information about them, said Martin Hack, CEO and cofounder of Skytree, a machine-learning and predictive-analytics software provider.

Marketing managers for smaller stores who need to increase sales could use analytics to identify more effective product promotions and create offers targeted to specific customers.

Product managers and others concerned with customer retention could utilize analyt-ics to predict the risk of each customer going to other companies and then identify actions likely to keep them loyal.

Small companies could analyze IT expenses to ensure they’re not wasting resources because of unused capacity or unnecessary duplication, noted Mark Levitt, senior analyst with market research firm Strategy Analytics.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






TECHNOLOGY NEWS

22 COMPUTER

that they store information only on internal systems.

Cloud-based analytics work in silos, which enables them to be tightly designed for one domain, according to WibiData’s Trajman. However, he noted, this means they don’t typically share data or results across department or disciplines within a company.

When organizations want to do this, he said, they typically must build their own cross-domain analytics system. External cloud-based services typically won’t be able to integrate closely enough in an organization’s IT architecture, noted University of California, San Diego professor Charles Elkan.

Over time and by reviewing customer input, vendors will figure out ways to make big

analytics less complex and thus easier to use.

For example, said Aberdeen’s Rowe, analytics software eventually will be able to process requests through a GUI or simple SQL commands rather than via approaches requiring specialized programming or hand coding in new and unfamiliar languages.

Aberdeen’s research indicates that self-service integration tools will allow for fast connections to multiple data sources, while interactive reports and dashboards will enable easy visualization and exploration of large datasets.

Products could even emerge that integrate data collection and analysis into daily business activities.

According to WibiData’s Trajman, big analytics tools eventually will be able to integrate multiple datasets from across departments and domains within a company to enable deeper analytical insights.

Many of big analytics’ future benefits will arise from combining information from multiple sources,

director for solutions marketing. Because Storm is cloud-based, she said, it’s easily scalable.

DRAWBACKSBig analytics for small

organizations faces several potential hurdles.

According to the results of an Aberdeen survey in which companies could provide up to three reasons why they haven’t invested in big-data-related technologies, 51 percent said the software and services are too expensive, 40 percent said they lack the necessary IT resources, and 37 percent said they don’t see a business need for it.

Cloud-based approaches also suffer from the same challenges that other cloud technologies face, such as cost and scalability.

In addition, noted Suvola’s Piedmonte, there is a limit as to how much data cloud-based analytics can handle. “When you’re talking about terabytes of data,” he explained, “the bandwidth just isn’t there for that now. If you are generating a lot of on-site data, moving the data to a remote cloud for analytics isn’t going to fly if you need to extract tactical business knowledge in real time.”

Security is another issue, as many organizations are hesitant to push their core data into the public cloud, said Continuuity’s Papaioannou. Some organizations, especially those in the financial and healthcare sectors, have policies requiring

such as IBM, Oracle, SAP, and SAS, as well start-ups like QlickTech, Tableau Software, and Tidemark.

Continuuity offers a suite of products that let Java developers more easily build, test, and deploy big analytics programs. It also provides the Continuuity AppFabric runtime hosting platform, which lets developers deploy, monitor, and manage the applications.

Cloud-based approachCloud-based analytics either

accepts uploaded data from a company or connects directly to its website or back-end service. For example, analytics from business-intelligence provider GoodData communicates directly with a company’s Zendesk customer-support software by using the API and credentials for Zendesk’s back-end services.

Smaller companies haven’t used the cloud-based approach for big analytics until now in part because the cloud-based services marketplace is relatively new, noted BigML’s Verwoerd. And, he added, “Prices are just now dropping to levels affordable for [them].”

It hasn’t always been easy either for smaller companies to transmit large amounts of data over networks or for service providers to return analytics quickly. Technical advances, such as faster networks and processors, have helped with this.

Today’s cloud-based big analytics providers include Amazon, BigML, Cloudability, CloudVertical, Cloudyn, Continu-uity, BigML, InsightsOne, Newvem, Rackspace, RightScale, and Uptime Software.

Splunk Storm is a cloud version of Splunk’s big data analytics software, made possible by the emergence of dependable, cost-effective commercial cloud-computing services, noted Leena Joshi, the company’s senior

Cloud-based analytics work in silos, which enables them to be tightly designed for one domain.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






Register today!

16-20 March 2013 Orlando, FL, USA

IEEE VR 2013 is the premier international conference and exhibition on virtual reality. Find the brightest minds, the most innovative research, the leading companies and the most stimulating

IEEE VR 2013 20th IEEE Virtual Reality

JANUARY 2013 23

com), a Fallbrook, California-based international marketing communications company with affiliate offices in Brazil, France, Germany, Hong Kong, India, and the UK. He writes frequently on technology topics and can be reached at [email protected].

Many users who aren’t technically savvy might not fully understand analytics’ benefits or how to utilize the approach and thus may choose not to employ it.

Nonetheless, said Flavio Villanustre, LexisNexis Risk Solutions’ vice president for information security, “Organizations leveraging analytics will have a greater competitive advantage. Those that don’t will lag behind their peers.”

Neal Leavitt is president of Leavitt Communications (www.leavcom.

both within an organization and across collaborating groups, added the University of California, San Diego’s Elkan.

Of the two approaches, said InsightsOne’s Ducher, cloud-based analytics will emerge as the default for small organizations or for departments within big companies.

One potential issue, noted Rowe, is that as even smaller organizations work with increasing data volumes, there may not be the available network bandwidth for quickly and frequently sending information via the cloud to an analytics provider.

Selected CS articles and columns are available for free at http:// ComputingNow.computer.org.

Editor: Lee Garber, Computer;[email protected]



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________

____________

____________________________________________________

___

________________

___

http://www.qmags.com/clickthrough.asp?url=http://ComputingNow.computer.org&id=17866&adid=P23E2

http://www.qmags.com/clickthrough.asp?url=www.leavcom.com&id=17866&adid=P23E1



http://www.qmags.com/clickthrough.asp?url=www.leavcom.com&id=17866&adid=P23E1

http://www.qmags.com/clickthrough.asp?url=http://conferences.computer.org/vr/2013&id=17866&adid=P23A1





NEWS BRIEFS


A transmission electron microscope image shows a new 3D transistor made of indium gallium arsenide rather than silicon, which promises to enable faster, smaller, energy-efficient chips.

New 3D Transistor Could Enable Faster, Smaller Chips

Researchers from Purdue and Harvard Universities have created a 3D transistor made from a material that could replace silicon someday.

Engineers could use the transistor, which has stacked circuitry in three dimensions—instead of the typical two—to build faster, smaller, energy-efficient chips that would generate low amounts of heat.

Each transistor contains three nanowires made of indium gallium arsenide. “Stacking the circuitry results in more current and much faster operation for high-speed computing,” said Purdue professor Peide Ye. “It’s a preview of things to come in the semiconductor industry.”

Researchers have been trying to make processors faster by using circuitry with smaller feature sizes. However, feature sizes below 22 nanometers don’t work well in 2D chips. This has encouraged development and release of 3D chips. Engineers hope to develop transistors with feature sizes of 14 nanometers by 2015 and 10 nanometers by 2018.

However, electrons flow more slowly in silicon than other materials

now being considered for use in chips, such as indium gallium arsenide, Ye said. Dropping feature sizes below 10 nanometers with better performance is thus unlikely with silicon, he noted.

In addition to using indium gallium arsenide, the new transistor’s nanowires are coated with a 4-nanometer-thick layer of lanthanum aluminate and a half-nanometer-thick layer of aluminum oxide, which is different than the insulating material found in traditional transistors. This coating lets the Purdue and Harvard researchers create indium-gallium-arsenide transistors with 20-nanometer feature sizes, which represents a technical breakthrough, according to Ye.

Current insulating material doesn’t perform properly in transistors with small feature sizes and leaks electrical charge when they shut down.

The US National Science Foundation and Semiconductor Research Corp., a technology research consortium, funded the work, which is led by Purdue doctoral student Jiangjiang Gu and Harvard postdoctoral researcher Xinwei Wang.

Collection of Nanomachines Mimics Muscles

Researchers have developed the first nanomachines that, when grouped together, can coordinate their movements to function like contracting muscles.

This work by France’s Centre National de la Recherche Scienti-fique (National Center for Scientific Research) could be used in robotics, computer nanostorage, medicine, and nanomechanical applications.

The project validates theories on the use of nanomachines to mimic muscles.

In humans, thousands of units of highly complex protein assemblies coordinate motion to control muscle movements. Each assembly moves about only a nanometer, but in large groups, they magnify their effect to enable larger motions. Research-ers have developed such assemblies in the past but have not been able to effectively coordinate their movements.

Now, the CNRS team—led by Université de Strasbourg professor Nicolas Giuseppone and including researchers from the Laboratoire de Matière et Systèmes Complexes (Laboratory of Complex Matter and Systems) at Université Paris Diderot—has overcome this hurdle.

The scientists synthesized long polymer chains incorporating thou-sands of protein assemblies that can each move about one nanometer. Changing the assemblies’ pH balance produces coordinated contractions and expansions up to 10 microm-eters. This could help researchers



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 25

The unidentified 35-year-old sus-pect—who faces felony data-theft charges–—had information from identity cards, as well as tax-identi-fication and license-plate numbers and home addresses, according to police.

Investigators haven’t deter-mined yet how the man obtained the data or what he planned to do with it. They say they began look-ing into the matter when a worker at Greece’s Hellenic Data Protec-tion Authority notified them that someone appeared to have a large number of digital files contain-ing personal information. Police tracked down the suspect via the Internet and then searched his house, seizing computer equipment in the process.

Authorities have not released additional information about the case.

UN Report: Developing Nations Should Foster Domestic Software Production

According to a newly released UN report, developing countries should encourage and expand the local cre-ation of software because it would improve their economies, help with

to enable more human-like robotic movements, give nanomaterials new mechanical properties, or create artificial muscles.

Surveillance System Monitors Operators’ Brainwaves to Detect Threats

Many governments and com-panies are adopting surveillance systems that involve people watch-ing cameras for signs of enemy or criminal activity.

However, these systems depend on humans understanding what they’re seeing and reacting appro-priately. For various reasons, including fatigue, people sometimes don’t consciously recognize what they’re seeing or what it represents.

DARPA—with researchers from HRL Laboratories; Quantum Applied Science and Research; Advanced Brain Monitoring; and the University of California, San Diego—has devel-oped a sensor-based surveillance system that promises to overcome these issues.

The agency has finished testing its Cognitive Technology Threat Warning System (CT2WS), designed primarily to enhance the abilities of individuals to view battlefields and other areas and warn US soldiers about potential problems.

CT2WS uses a 120-megapixel, wide-angle, digital video camera, along with image-processing software and a cap with electroen-cephalogram equipment that the operator wears. The system scans a 120-degree arc with its digital camera and shows the user 10 images per second. It then monitors the operator for P-300 brain activity, which relates to the processing of images and sounds. Users may not be aware of something that causes a jump in P-300 activity, but CT2WS will flag it for review.

By factoring in human reactions, the surveillance system could rec-ognize events that fully automated systems might not see as possibly

troublesome, such as the movement of tall grass in an area where there might be guerrilla activity.

The DARPA scientists have tested CT2WS in desert, tropical, and open settings. Each hour, they used 2,034 “events” to test the image-process-ing software alone, without P-300 monitoring. The system yielded 810 false alarms.

With P-300 monitoring, CT2WS caused only five false alarms per hour and successfully identified 91 percent of the actual threats simulated in testing. Adding a com-mercial portable radar improved the threat-identification rate to 100 percent.

Man Accused of Stealing Personal Data of Most Greeks

Greek police have arrested a man who allegedly had 9 million personal records of Greece’s citi-zens, potentially representing most of the country’s population of about 10.8 million.

Authorities say this is Greece’s biggest breach of private informa-tion. Some of the stolen records were duplicates, so police were not sure exactly how many people’s records had been taken.

New technology reads the brainwaves of people monitoring surveillance cameras to detect potential security issues that they may not be conscious of.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






NEWS BRIEFS

26 COMPUTER

social development, and meet local technology needs.

The study by the UN Conference on Trade and Development—titled “Information Economy Report 2012: The Software Industry and Develop-ing Countries” (http://unctad.org/en/PublicationsLibrary/ier2012_en.pdf)—said governments should adopt policies that would increase domestic software production.

The UNCTAD report stated that locally developed software is more likely to fit the needs, con-text, culture, and language of the people who will use it. This could be particularly important in fields such as healthcare, education, and government.

Increased domestic creation of software—which doesn’t require a lot of capital investment but could generate significant revenue—might help transform, modernize, and expand developing countries’ economies and generate jobs in the process, the study noted.

Governments should actively help increase software develop-ment by buying more locally created applications, encouraging the teaching of software devel-opment in schools, upgrading the technology infrastructure, adopting favorable legal and regu-latory frameworks, and working with domestic universities and companies.

The UN report said that increased global access to technology, includ-ing broadband networks and high-speed wireless service, makes it easier for citizens of developing countries to get involved in software development. The study cited initia-tives in countries such as Argentina, Bangladesh, Brazil, India, Malay-sia, Nigeria, and South Korea as examples of governments fostering domestic software development.

Editor: Lee Garber, Computer;[email protected]

CONSERVATIONISTS RESTORE EARLY UKDIGITAL COMPUTER

U K conservationists have restored the world’s oldest original working digital computer—often referred to as the WITCH—and have placed it on display at a museum, more than 60

years after it first began operation.The relay-based computer—known as the Harwell Dekatron or Harwell computer—

was originally used by the UK’s atomic-energy research program in the 1950s.The machine, which sat in a storeroom for 15 years, will now be an exhibit at the National

Museum of Computing at Bletchley Park in Buckinghamshire. Engineers began designing the 2.5-ton computer in 1949 to help officials at the UK’s

Atomic Energy Research Establishment (AERE) perform calculations that they used to do with adding machines.

The Harwell Dekatron debuted in 1951 and took up to 10 seconds to multiply two num-bers. However, it was useful because it was reliable and ran automatically for long periods of time.

The computer had 480 relays and 828 dekatron tubes for volatile memory. Dekatrons were gas-filled decimal counting tubes that were popular in the 1950s and 1960s. The system used paper tape for input and storage.

In 1957, with the advent of faster, smaller machines, AERE stopped using the Harwell and gave it to the Wolverhampton and Staffordshire Technical College (now the University of Wolverhampton) for use in teaching programming. At that point, it got the name Wolverhampton Instrument for Teaching Computing from Harwell (WITCH).

Conservationists The university donated the Harwell in 1973 to Birmingham’s Museum of Science and Industry, where it stayed on exhibit until 1997, when the museum closed and the machine was placed in storage.

Several years ago, a trustee at the National Museum of Computing saw a photograph of the WITCH in storage and suggested restoration. The computer was then sent to the museum.

The restoration effort was led by Delwyn Holroyd, cofounder and director of Codex Dig-ital Ltd., which designs and makes digital equipment for movie and TV production. The focus was to keep as much original equipment as possible.

More than 60 years after it debuted, the WITCH, originally known as the Harwell Dekatron computer, has been restored and is now on exhibit at the UK’s National Museum of Computing.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________

_____________________

____

http://www.qmags.com/clickthrough.asp?url=http://unctad.org/en/PublicationsLibrary/ier2012_en.pdf&id=17866&adid=P26E1






How well do you know the software

development process? Rise to the

challenge by taking the

CSDA or CSDP Examination.

With more and more employers

seeking credential holders,

it’s a great time to add this

unique credential to your resume.

Think You Know Software?

PROVE IT!

For exam dates, locations, pricing, and more information:

Visit: www.computer.org/certifi cation

Email: certifi [email protected]



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


________________________________

http://www.qmags.com/clickthrough.asp?url=www.computer.org/certification&id=17866&adid=P27A1






COVER FE ATURE


This transition from acute to chronic treatment, paired with an extended life span, resulted in a healthcare system that is growing with unsustainable levels of cost (percent-age of GDP). While this problem is evident in the US, it is also becoming an issue in developing countries.1

In tandem with these changes, high-quality, user-friendly wireless consumer devices, such as mobile phones, have emerged that accompany users most of the time. These devices provide not only mobile communica-tion, but also sensing, analytic, and visual capabilities, as well as access to the cloud. Sensors embedded in a mobile phone, complemented by sensors on and in a body, can provide an unprecedented view of the person’s health status and behavior patterns.

mHealth builds upon earlier work in telehealth, mobile computing, and persuasive technology in healthcare set-tings.2 It has the potential to turn mobile devices into personal labs that continuously assess a person’s physiol-ogy, behavior, social context, and environmental exposure. For example, a personal therapist application on a mobile device could mine the Internet for information about the latest health research and apply it while continuously col-lecting personal health data to make inferences about the user’s health, and then share these results with caregivers so they can provide appropriate treatments. Persuasive user interfaces on the mobile device could facilitate com-pliance with the prescribed treatment protocol by applying just-in-time intervention. In addition to directly improving healthcare, mHealth could also accelerate health research and inform the formulation of public health policies.

R ecent advances in mobile technology have opened up enormous opportunities to improve patients’ health and well-being. mHealth tech-nologies offer real-time monitoring and detection

of changes in health status, support the adoption and maintenance of a healthy lifestyle, provide rapid diagnosis of health conditions, and facilitate the implementation of interventions ranging from promoting patient self-care to providing remote healthcare services.

Although mHealth is a new area of scientific develop-ment, researchers have been laying the groundwork over the past four decades. Medical practice and healthcare originated as a system to treat infectious diseases (such as smallpox) and traumatic injuries. As life expectancy increased, by the mid-1900s attention had shifted to manag-ing chronic illnesses such as diabetes and heart disease. By definition, chronic illnesses are not expected to be resolved through treatment.

Mobile health (mHealth) seeks to improve individuals’ health and well-being by con-tinuously monitoring their status, rapidly diagnosing medical conditions, recogniz-ing behaviors, and delivering just-in-time interventions, all in the user’s natural mo-bile environment.

Santosh Kumar, University of Memphis

Wendy Nilsen, US National Institutes of Health

Misha Pavel, Oregon Health and Science University

Mani Srivastava, University of California, Los Angeles

Mobile Health: Revolutionizing Healthcare Through Trans-disciplinary Research



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


AU

DI

O





JANUARY 2013 29

This vision of mHealth can be realized by building on recent advances in mobile technology, but the success of mHealth will require considerable innovation in funda-mental science and engineering, as well as integration of the technology with healthcare systems.

MOBILE HEALTH SYSTEMS RESEARCHAs Figure 1 shows, mobile technology can simulta-

neously acquire information, process the data, make inferences, mediate a range of interventions, and provide communications with other devices and systems.

Sensing in mHealthTo reach its potential, mHealth technology must be able

to capture diverse personal and environmental signals relevant to the health of both individuals and commu-nities. Researchers are repurposing a variety of sensors already included in mobile phones, such as accelerom-eters and GPS technology, via sophisticated algorithms for

use in mHealth applications, resulting in many innovative health-, wellness-, and fitness-related applications. How-ever, the measured signals from these sensors often lack clinical relevance and fall short of the specificity needed to allow definitive diagnosis and treatment of complex health conditions.

As interest in mHealth grows, we anticipate the increasing availability of sensors that are specifically targeted at and optimized for mHealth. In the short term, factors such as cost, market size, mass, volume, and placement constraints might necessitate using external sensors that are wirelessly connected to a smartphone. Over the long term, however, the need for a better user experience is likely to require in-tegrating such sensing into smartphones or other emergent cellular-connected wearable devices such as eyeglasses.

The core challenge lies in developing new sensors that are compatible with incorporation in a smartphone from a cost and size perspective, can be used for continuous real-time sensing without much burden on the user, and

On-b

ody

sens

ors

Smar

tpho

neBa

ck-e

nd se

rver

s

Data logger

Offline modellearning

Sensor data collector

Personal inference(lightweight, real time)

Healthstatus,

behaviors,contexts

Datafault

diagnosis

EMA data collector

Feedback andintervention

Data quality assessment

mod

els @

phon

emodels @ backend

prompt

response

instructionsincentives

Personal and population inference

CaregiversResearchersPolicymakers

Population

Subject

Embeddedphone

sensors

Online modellearning

Healthstatus,

behaviors,contexts

Datafault

diagnosis

Figure 1. An overview of mobile health systems. Sensors at the lowest layer collect raw data that is processed to make infer-ences about individuals. These inferences can then be used to inform the design or delivery of just‐in‐time interventions and to make health inferences at the population level to inform health research, practice, delivery, and policy formulation.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






30 COMPUTER

COVER FE ATURE

enable various tests that can only be done in a clinical set-ting today. Biomolecular sensing, imaging, and bioelectric sensing are particularly important.

Biomolecular sensing. With their high selectivity, ultra sensitivity, and energy efficiency, solid-state sensors can turn a common smartphone into a powerful and easy-to-use diagnostic tool.

Assessing biomarkers and pathogens in body fluids and human breath to detect diseases, their progression, and therapy effectiveness is essential in healthcare. Such sensing typically relies on sample preparation and labora-tory analysis, which are not usually available in a mobile setting. Advancements in nanotechnology, microfluid-ics, and solid-state sensors, however, offer the promise of miniature, low-cost, chip-sized sensors that can provide “lab-on-chip” capabilities.

Examples include

Nanowire sensors, which are fabricated at low cost in high-yield semiconductor foundries and packaged into disposable electronic strips.3 These embedded sensors can detect the presence of specific molecules such as cardiac troponins in a mobile setting to facili-tate the diagnosis, monitoring, and risk stratification of suspected acute myocardial infarction in cardio-vascular patients. All-electronic digital microfluidic devices operated by electrowetting-on-dielectric actuators, which can stimulate highly precise, programmable microreac-tions to allow performing a wide range of assays in a mobile setting. DNA microarrays, which can capture the epigenetic information necessary to understand the protein-DNA interactions that underlie many biological processes and disease states.Low-power solid-state chemical sensors, which use inorganic materials for potentiometric and resistive sensing of trace gases in human breath in real time and at low parts per billion (ppb) concentrations. Research has associated the presence of trace gases with various diseases, such as nitric oxide for asthma, acetone for diabetes, and hydrogen for gastroenteric ailments.

Imaging. Passive and active imaging methods such as ultrasound, x-rays, MRI, and CT scans are mainstays of modern healthcare. However, factors such as optical

pathway impose size constraints and the need to gener-ate powerful or even dangerous signals (such as x-rays) or fields (such as MRI) make it impossible to embed them in a smartphone. Indeed, cameras are the only imaging devices widely available in a mobile environment. Using computer vision methods, researchers have incorporated smartphone cameras in mHealth applications to perform tasks such as detecting the heart rate from microblushes and estimating refractive errors in the eye. For imaging to scale in mHealth, it must move beyond the current human-in-the-loop approach for interpreting images to perform computational triage. Two solutions are emerging that provide complementary capabilities.

The first one is lens-free computational microscopy and tomography4 running on a smartphone that can algorithmically overcome optical constraints to provide high-resolution 3D imaging of biological samples with a wide field of view and a large depth of field. These methods have been shown to assay blood samples for malaria.

The second solution is radio frequency (RF) imag-ing, which is an attractive option for smartphones as they already have several built-in radio transmitters and receivers—for example, cellular, Wi-Fi, and Bluetooth—that researchers can potentially repurpose. Wideband signals from an RF transmitter can penetrate and illu-minate the interior of the human body, and the mobile device can analyze the interferogram image resulting from the reflected signal waveforms to infer a variety of internal variables, such as heart motion, blood flow, respiration, and fluid accumulation. Because RF imag-ing is inherently contactless—that is, it does not require coupling transducers to the body via gels or fixed elec-trodes—it allows unobtrusive, real-time physiological sensing.

Bioelectric sensing. The measurement and analysis of surface biopotentials is a powerful, and often the only, sensing modality for diagnosing and monitoring many disorders. Examples include electrocardiography (ECG) for the heart, electroencephalography (EEG) for the brain, and electromyography (EMG) for muscles.

Although devices for measuring surface bioelectric sig-nals in mobile environments have been available for many years (Holter monitors, for example), they are too cumber-some for long-term monitoring, which requires affixing multiple electrodes to the body. With ongoing develop-ments in low-power electronics, smarter and compressive sampling, and energy harvesting, batteryless wireless patches could offer a less obtrusive bioelectric sensing approach that eliminates the tangle of wires and elec-trodes.5 For sporadic monitoring, emerging contactless bioelectric sensors that use through-the-clothing capaci-tive coupling could allow building the entire sensor into a smartphone.

For imaging to scale in mHealth, it must move beyond the current human-in-the-loop approach for interpreting images to perform computational triage.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 31

Mobile computing for mHealthAn effective mHealth computing platform must be able

to efficiently make semantically rich and medically trusted inferences about physical, physiological, psychological, cognitive, and behavioral states from sensor information, and correlate these inferences with environmental, social, and other factors.

Current mobile operating systems offer sophisticated network communications capabilities and provide appli-cations with rich abstractions of communication patterns such as Web services. Rich networking stacks and frame-works free individual applications from handling low-level data types such as bits, bytes, and packets or managing low-level protocols for media access, routing, and trans-port. However, these systems provide rather primitive software support for sensory applications. mHealth ap-plications are developed as ad hoc stovepipes handling all the system layers, from raw transducer signals to semanti-cally rich inferences to actionable information. The lack of a standard approach is largely due to the failure to provide reusable and semantically rich abstractions of sensory information for mobile operating systems.

We envision the emergence of a “sensing stack” analo-gous to the networking stack. The sensing stack might similarly be organized as a layered set of modules that systematically transforms sensor measurements into rich inferences that applications subscribe to. The sensing stack would have layers reflecting the needs of typical sensing applications:

a bottom layer that samples front-end signal process-ing to extract feature vectors, an inference layer that maps classifiers and recog-nizers such as support vector machines (SVM) into semantic labels, and a top layer that extracts complex events from the time series of inference labels.

Applications would access the stack via a suitable API. For example, an application might ask to be notified when the user is engaged in a particular activity or is in a particu-lar behavioral state—for example, “running,” “stressed,” or “talking to Bob.” However, unlike networking stacks, which deal with relatively few network interface types, the sens-ing stack would need to deal with many diverse sensors and corresponding processing.

Energy is another challenge. The realistic lifetime of current mobile phones when engaged in continuous sens-ing is woefully short. As smartphones begin to use sensors for continuous inferencing, they will need sensing subsys-tems for efficient sampling and duty cycling. In addition, the supporting architecture will need to be both efficient and high performance, providing dedicated sensor proces-sors that applications can configure and program.

Addressing the current shortcomings will require revis-iting assumptions underlying the hardware and software organization of mobile devices, which are optimized for sporadic interactive computing and communication. For example, designers will need to optimize the analog-sampling-computation-communication chain using ad-vances in adaptive and sub-Nyquist sampling,6 tiered processing architectures, in situ measurement of system performance to cope with environmental and platform variability affecting power performance, and emerging ultra-low-power wireless technologies.

mHealth analyticsAs the variety and availability of sensing and mobile

computing technologies for mHealth increase, the data collected from the mobile environment will grow expo-nentially; multiple sensors will continuously collect data at high frequency for the user’s lifetime. A major challenge is extracting actionable information and knowledge from

this data deluge to provide accurate health information for users and professionals and help researchers and policy-makers reach optimal conclusions.

For the data to be useful in making health decisions, they must represent their provenance, quality, and validity. mHealth applications can use well-characterized data to make inferences using models of the relationship between the acquired data and the phenomena of interest. For ex-ample, if the applications can estimate the uncertainties associated with the placement of the device relative to the user’s body, they can use the data obtained from a mobile device’s accelerometer to estimate an individual’s actions and activities.

Data quality. Data collection in mHealth sensing intro-duces various data-quality challenges. Sensors such as ECG electrodes might be placed incorrectly on the body, or, even if initially placed at the correct location, could subsequently slip or become detached. Sensor measure-ments might be noisy not only because of placement and attachment errors, but also because of the variability inherent in a patient’s daily activities and the mobile envi-ronment. Wearable wireless communication devices are convenient, but can be another source of signal distortion. Finally, a device might intentionally degrade data quality to conserve battery life. These issues call for new research in data quality.

Current mobile operating systems offer sophisticated network communications capabilities and provide applications with rich abstractions of communication patterns such as Web services.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






32 COMPUTER

COVER FE ATURE

We need metrics to characterize the distortions and uncertainties associated with collecting the data and the resulting inaccuracy because decisions are based on that data. The specification of such metrics is an open ques-tion. For example, can the metric indicate the inaccuracy expected in computing heart rate from a poor-quality ECG signal? Having the same or similar metrics for different sensors will make interpreting and processing the data much easier. But is a universal data-quality metric fea-sible for a single sensor, let alone for different sensors? In an ECG, for example, can the same metric measure inaccuracy in computing both heart rate and heart rate variability? Metrics should preserve information quality even when the data quality varies.

We also need methods to restore data quality. These methods will need to not only detect data quality dete-rioration, but also identify the deterioration source—for example, poor attachment or component malfunction—and use this diagnosis to engage the sensor user in taking appropriate action to restore the data quality.

Model-based inference. For mHealth data to be usable in making health-related decisions—for example, to initi-ate a just-in-time intervention, alert a caregiver, or even to solicit more information via self-reporting, mHealth tools will need higher-level inferences regarding health status, behavior, and context. Such inferences require new com-putational models that relate the observable variables to the quantities of interest.

Researchers are making progress in inferring physical state (such as posture and activity using accelerome-ters7), psychological state (such as stress using sensory measurements8), social context (such as conversation based on respiratory patterns9), and environment (such as place and commuting status using GPS). Significant work, however, is required to make these models and inferences reliable enough to use in the real world with a diverse sample of participants, so they can provide the basis for real-time inferences, decisions, and ac-tions. A key concern is the potential for high false-alarm rates that can render the entire system annoying and ineffective.

Under limited circumstances, researchers can derive the computational models from principles such as bio-mechanics.10 In most cases, however, they develop these models by combining principled approaches with machine learning and statistical pattern-recognition techniques. With this approach, however, the difficulty in obtaining

representative labeled data limits the development of reli-able and validated models.

Although data collected in the lab might have valid labels, it might not represent the natural environment in which mHealth systems ultimately need to work.11 Labels collected in a mobile environment represent the natural environ-ment, but they can be noisy, uncertain, biased, missing, or spurious. For example, researchers could use self-reported times of smoking to develop a model for automated de-tection of smoking, but these labels might not accurately represent the actual start and end times of the smoking episodes. Subjects might forget to report an episode or might even falsely report it to earn compensation. The models for inferring health states must also account for between-subject differences and changes in context without requiring subject-specific training with the model parameters.

Inference models should also adapt to changes in re-source and data quality and be able to separate out the effects of various confounds. For example, having both ECG and respiration measurements available would im-prove the models’ ability to infer psychosocial stress, but the models should work even if only one measurement is available. Further, activity, smoking, speaking, and so on, affect physiology. The model should therefore be able to demultiplex the effects of these and other confounding events to reliably infer the stress level.

A model based on engineering, physical sciences, and health and behavioral sciences principles will have a better chance of addressing these issues and generalizing across subjects, contexts, and environments.

Design of mHealth interventionsBecause mobile devices are usually continuously with

an individual and have the sensing and computational ca-pacity needed to collect and analyze health-related data in real time to infer health and behavior, they offer a powerful platform for delivering just-in-time adaptive intervention.

A mobile device that is aware of an individual’s health status and environment can adjust an intervention’s content and timing accordingly. For example, if it can rec-ognize vulnerable moments for a newly abstinent smoker, the device can deliver interventions to help resist the urge to smoke and prevent a relapse.

Mobile devices could also adapt the intervention’s con-tent to the vulnerability type (stress, alcohol, tobacco, and so on) as well as personalize the intervention to both the individual and the context (for example, “at work” versus “at home”). They could then apply their sensing capabilities to evaluate adherence and response to an intervention and use this information to adapt the interventions.

In general, an intervention can be viewed as analogous to closing a loop in an automatic control system. In classic control system theory, the system derives the input con-trol signal by comparing the desired output to the actual

Mobile devices offer a powerful platform for delivering just-in-time adaptive intervention.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 33

system behavior. Modern automatic control systems com-pute the optimal control signal from the desired outcome and a learned computational model of the target system.

Thus, achieving an optimal outcome for an intervention would require characterizing the human user’s behaviors in terms of a predictive computational model, which is an open transdisciplinary research problem. Capturing the com-plexities, nonlinearities, and uncertainties associated with human response to interventions requires developing new predictive computational models for the controlled systems.

In addition to developing an appropriate theoretical framework for control, researchers need simulation tools that leverage technological advances to create models that account for the many factors that determine health. They can then use both the theoretical model and simulation tools to analyze the safety and efficacy of mHealth inter-ventions prior to conducting human user studies.

mHEALTH SAFETY, EFFICACY, PRIVACY, AND SECURITY

Unlike other human-cyber-physical systems, where the human is just an operator, a sensor, or an actuator (such as in avionics or cars), in mHealth systems, humans are the “plant” whose health and well-being is to be affected and controlled. The life-and-death implications and the associated economic and legal burdens place a high degree of responsibility on mHealth system designers. Moreover, the human body is complex, highly variable, and not well understood. Thus, traditional computing verification and validation approaches are not particularly effective.

Designing trustworthy mHealth systems requires ex-tensive collaboration between engineering and health professionals to create and evaluate effective mechanisms. The overall goal of ensuring that mHealth systems are trustworthy comprises several dimensions.

mHealth safetyWhen assessing health products, the first question

evaluators ask is whether the device, medication, or treat-ment is safe. Likewise, researchers must ask whether an overall mHealth system is safe from both a health and an engineering perspective.

From a health perspective, safety means the mHealth system produces information that is valid and of adequate quality for critical decision making. For example, given the rapid onset of a heart failure event and the potentially cata-strophic impact of a missed detection, an mHealth system that predicts heart failure in patients with congestive heart disease must maintain high-quality information continu-ously over an extended time period. For more advanced mHealth systems that might also trigger autonomous physiological and behavioral interventions, safety from a health perspective means that the interventions are medi-cally safe and appropriate.

From an engineering perspective, safety means that devices, such as sensors, used in mHealth systems will not cause their users unanticipated harm or discomfort because of design or manufacturing errors. For example, a wearable sensor with poor circuit and thermal design could lead to excessive heating or a battery fire. Engineering safety is not just a hardware matter—errors in the embed-ded software are also a source of concern about safety. For example, faulty firmware in a defibrillator can cause unintended shocking, and an ill-designed user interface might confuse the user and elicit incorrect responses. Even when a device is functioning as engineered, the human body’s variability combined with ambient conditions can result in safety issues in unexpected contexts, such as in extreme environmental conditions. For a device to be safe, designers must address all of these factors.

mHealth efficacyEfficacy—the evaluation of whether a device does

what it claims, as well as for whom and in what context—is crucial for devices used in health applications. Health researchers ask whether the device is valid, that is, it measures what it claims, and reliable, that is, it generates reproducible measurements. Although these are common healthcare issues, defining and assessing such metrics is difficult in a mobile environment. For example, what’s the best way to test the reliability of a device that is designed to assess temporal variability or address a concept’s validity for which there is no ground truth (stress, for example)? These issues call for new research and specific metrics.

The evaluation of health interventions usually occurs in multiple phases. Early in development, carefully de-signed studies with individual patients (n-of-1 design) or small groups (pretest, post-test, or time-series designs) are useful. Researchers can use these studies to assess safety, feasibility, and usability and determine the interventions’ potential effect.

Later in development, when an intervention is mature, the use of randomized clinical trials (RCTs) is common. To minimize biases, RCTs involve randomly assigning a large number (perhaps thousands) of potential participants to a treatment group and a control group, in which some participants receive a placebo. The RCT lets researchers estimate the treatment effects’ statistical significance and size. Although statistically sound, RCTs are generally ex-pensive, inefficient, and lengthy. Other designs, such as regression-discontinuity and stepped-wedge, have similar

Designing trustworthy mHealth systems requires extensive collaboration between engineering and health professionals to create and evaluate effective mechanisms.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






COVER FE ATURE

34 COMPUTER

methodological rigor and might be suitable alternatives for mHealth.

Because technology evolves rapidly in this area, freezing an intervention during evaluation, which usually lasts for years, is not feasible. Therefore, further work is needed to develop research designs suitable for evaluating mHealth interventions.12

mHealth privacy and security Patient information privacy has long been an area of

concern in healthcare. Health information is managed under the 1996 Health Insurance Portability and Ac-countability Act (HIPAA), which aims to ensure that the information remains private (by regulating the use and disclosure of data to various parties) and secure (by man-dating administrative, physical, and technical safeguards against intrusion by unauthorized parties). Created in an era when health information had just begun to be digi-tized, HIPAA could not have anticipated the many ways in which health information is now collected, exchanged, and processed.

mHealth systems pose enormous challenges for HIPAA. One challenge comes from the openness of mHealth sys-tems. Sensors, computers, and networks that collect, transmit, process, and act upon healthcare information are owned and operated by multiple parties with complex trust relationships and technical competence. Ensuring privacy and security in such a setting, while well under-stood from a theoretical perspective, is difficult in practice where concerns of usability, cost, legacy, and conflicting interests intrude.

mHealth systems pose an even bigger privacy challenge because the data they collect from wearable sensors and personal mobile devices presents fundamentally new risks and vulnerabilities.13 Embedded in this data, which users share willingly and often not anonymously with others, are many privacy-revealing behaviors such as addictive behaviors and movement patterns. Smart algorithms can fuse these behaviors with digital footprints—that is, information from other sensors and publicly available information—to construct a near-real-time virtual biogra-phy of previously private behaviors and lifestyle patterns.

Although certain architectural principles, such as giving individuals more control over the data that sensors collect about them, have begun to emerge, designing mHealth systems that are sensitive to the needs of both producers and consumers of information remains an open challenge. Privacy-preserving mHealth systems would need to go beyond the traditional focus on data and identity privacy to providing behavior privacy.

In addition, because mHealth systems are continuously connected to the network, they are more vulnerable to tampering and jamming of sensing and communication functions by third parties, resulting in denial of service

and incorrect operation that could cause medical harm and threaten users’ well-being.14

mHEALTH REGULATION In the US, mHealth devices come under the regulatory

authority of the Food and Drug Administration (FDA). In mHealth, as in other healthcare areas, the FDA regulates products that are intended for medical use. Products can include software, hardware, and devices that combine them.

The intent of the person or vendor who develops or sells a product gives it a medical-use quality. Some items, such as artificial heart valves, clearly have an intended medical use. But, when there is less than 100 percent clarity about intended use, the FDA considers a vendor’s words, actions, and recommendations to customers to determine intended use. For example, if a vendor promotes an mHealth prod-uct such as a sensor or a mobile app as a sleep aid, the FDA can label it as a wellness device rather than as a product for medical use. If the same technology is marketed as a treatment for a sleep disorder, it will likely have a medical-use quality. In this case, the device will be subject to regu-lation and must be tested and evaluated using the FDA’s scientific standards and processes. While using a product beyond its approved uses is not illegal, promoting it for something beyond what it is approved for is. The FDA is allowed some regulatory discretion; it can choose not to regulate something or can classify it for a specific type or level of regulation.

In 2011, the FDA released draft guidance for mHealth,15

and it now has congressional authority to formalize the guidelines.

Multiple US federal agencies that touch on various aspects of mobile computing also regulate and govern mHealth. This includes the Federal Communications Commission (FCC), which governs wireless operators and spectrum, and the Federal Trade Commission (FTC), which regulates interstate commerce and works for consumer protections, including mobile data security. Additionally, mHealth operates within the frameworks of other federal agencies that address mobility issues within the greater healthcare and standards systems. This includes the Office of the National Coordinator for Health IT, which provides guidance and support for the nation’s health information technology infrastructure, and the Department of Health and Human Services’ Office for Civil Rights, which is re-sponsible for implementing and enforcing HIPAA. These agencies work together to regulate and guide the mHealth domain.16

m Health systems herald an exciting new era in health with a shifting focus of healthcare to well-ness and prevention. These devices also portend

a transformed health research environment, where most



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 35

data is collected remotely and entire clinical trials might be run without the researcher and the participant ever meeting face to face. The data from these systems, in combination with the multiple fixed sensors in the envi-ronment, will also create a rich database for exploring new ways of understanding health. Thus, as mHealth systems become more prevalent and versatile, their use will not only enable myriad disruptive transformations in healthcare delivery and medical research, but will also present many scientific, engineering, and regulatory challenges. mHealth’s success will, therefore, depend on transdisciplinary research collaboration among comput-ing, engineering, and medical researchers.

References1. United Nations General Assembly, “Prevention and Control

of Non-Communicable Diseases,” Report of the Secretary-General, A/66/83, 19 May 2011.

2. R.S.H. Istepanian, E. Jovanov, and Y.T. Zhang, “Guest Edi-torial Introduction to the Special Section on m-Health: Beyond Seamless Mobility and Global Wireless Health-Care Connectivity,” IEEE Trans. Information Technology in Biomedicine, vol. 8, no. 4, 2004, pp. 405-414.

3. F. Patolsky and C.M. Lieber, “Nanowire Nanosensors,” Ma-terials Today, Apr. 2005, pp. 20-28.

4. H. Zhu et al., “Optical Imaging Techniques for Point-of-Care Diagnostics,” Lab on a Chip, no. 1, 2013, pp. 51-67.

5. V. Pop et al., “Human++: Wireless Autonomous Sensor Technology for Body Area Networks,” Proc. 16th Asia and South Pacific Design Automation Conf. (ASPDAC 11), IEEE, 2011, pp. 561-566.

6. H. Mamaghanian et al., “Compressed Sensing for Real-Time Energy-Efficient ECG Compression on Wireless Body Sensor Nodes,” IEEE Trans. Biomedical Eng., vol. 58, no. 9, 2011, pp. 2456-2466.

7. M. Rabbi et al., “Passive and In-Situ Assessment of Mental and Physical Well-Being Using Mobile Sensors,” Proc. 13th Int’l Conf. Ubiquitous Computing (UbiComp 11), ACM, 2011, pp. 385-394.

8. K. Plarre et al., “Continuous Inference of Psychological Stress from Sensory Measurements Collected in the Natu-ral Environment,” Proc. Conf. Information Processing in Sensor Networks (IPSN 11), ACM, 2011, pp. 97-108.

9. M. Rahman et al., “mConverse: Inferring Conversation Episodes from Respiratory Measurements Collected in the Field,” Proc. ACM Conf. Wireless Health (WH 11), ACM, 2011; doi:10.1145/2077546.2077557.

10. A.D. Kuo, “An Optimal Control Model for Analyzing Human Postural Balance,” IEEE Trans. Biomedical Eng., vol. 42, no. 1, 1995, pp. 87-101.

11. J. Gall et al., “Hough Forests for Object Detection, Tracking, and Action Recognition,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 33, no. 11, 2011, pp. 2188-2202.

12. S. Kumar et al., “mHealth Evidence Workshop: Exploring Innovative Methods to Evaluate the Efficacy and Safety of Mobile Health,” Am. J. Preventive Medicine, 2013, in press.

13. A. Raij et al., “Privacy Risks Emerging from the Adoption of Innocuous Wearable Sensors in the Mobile Environment,” Proc. Conf. Human Factors in Computing Systems (CHI 11), ACM, 2011, pp. 11-20.

14. W. Burleson et al., “Design Challenges for Secure Implant-able Medical Devices,” Proc. Design Automation Conf. (DAC 12), ACM, 2012, pp. 12-17.

15. US Food and Drug Administration, “Draft Guidance for Industry and Food and Drug Administrative Staff—Mobile Medical Applications,” 21 July 2011; www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm263280.htm.

16. Office of the National Coordinator for Health Information Technology, “Mobile Devices Roundtable: Safeguarding Health Information,” 16 Mar. 2012; http://healthit.hhs.gov/portal/server.pt/community/healthit_hhs_gov__mobile_devices_roundtable_agenda/3846.

AcknowledgmentsThe views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the US National Institutes of Health or any other author-affiliated organizations.

Santosh Kumar is an associate professor in the Department of Computer Science at the University of Memphis. His mHealth research interests include scalable physiological sensing, reliable inference of human health and behaviors, and realization of just-in-time interventions. He received a PhD in computer science and engineering from the Ohio State University. Kumar is a senior member of IEEE. Con-tact him at [email protected].

Wendy Nilsen is a health scientist administrator at the US National Institutes of Health Office of Behavioral and Social Sciences Research. Her scientific focus is on human behavior and behavior change, including utilizing mobile technology to better understand and improve health, adher-ence, the mechanisms of behavior change, and behavioral interventions in complex patients in primary care. Nilsen’s current mHealth efforts include addressing methodology and barriers to the utilization of mobile technology in re-search, serving on numerous federal mHealth initiatives, and leading the mHealth training institutes. Contact her at [email protected].

Misha Pavel is a professor in the Department of Biomedical Engineering, with a joint appointment in the Department of Medical Informatics and Clinical Epidemiology, at the Oregon Health and Science University. He currently over-sees the Smart Health and Well-being program at the US National Science Foundation. Pavel received a PhD in ex-perimental psychology from New York University and an MS in electrical engineering from Stanford University. He is a senior member of IEEE. Contact him at [email protected].

Mani Srivastava is a professor in the Department of Elec-trical Engineering and is affiliated with the Department of Computer Science at the University of California, Los Angeles. His research interests include embedded wireless systems, power-aware systems, wireless networks, and pervasive sensing. Srivastava received a PhD in electrical engineering from UC Berkeley. He is an IEEE Fellow. Contact him at [email protected].

Selected CS articles and columns are available for free at http://ComputingNow.computer.org.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_________

___________

_____________

____________________

____

_______________________

__________________________________

____________________


http://www.qmags.com/clickthrough.asp?url=http://healthit.hhs.gov/portal/server.pt/community/healthit_hhs_gov__mobile_devices_roundtable_agenda/3846&id=17866&adid=P35E3

http://www.qmags.com/clickthrough.asp?url=www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm263280.htm&id=17866&adid=P35E2









COVER FE ATURE


HiLCPS applications offer benefits in many realms—for example, the population of functionally locked-in individuals would benefit tremendously from such sys-tems. Because these individuals cannot interact with the physical world through their own movement and speech, they often must rely heavily on support from caregivers to perform fundamental everyday tasks, such as eating and communicating. As the “Fundamen-tal Autonomy for Functionally Locked-In Individuals” sidebar describes, a HiLCPS could aid in restoring some autonomy by offering alternative interfaces to the cyber-physical environment for interaction, com-munication, and control.

MULTIDISCIPLINARY CHALLENGESDesigning and implementing a HiLCPS poses tremen-

dous challenges and is extremely time-consuming. Experts from many disciplines need to join forces to successfully solve these challenges.

Transparent interfacesTraditional dedicated interfaces to the virtual world,

such as the keyboard, mouse, and joystick, are less suit-able for augmenting human interaction in the physical world. This environment requires transparent interfaces that use existing electrophysiological signals such as electroencephalography (EEG), electrocardiography (ECG), and electromyography (EMG), which measure electrical signals emitted by the brain, heart, and skel-etal muscles, respectively. Additional auxiliary sensors

H uman-in-the-loop cyber-physical systems (HiLCPSs) comprise a challenging and promis-ing class of applications with immense potential for impacting the daily lives of many people. As

Figure 1 shows, a typical HiLCPS consists of a loop involv-ing a human, an embedded system (the cyber component), and the physical environment. Basically, the embedded system augments a human’s interaction with the physi-cal world.

A HiLCPS infers the user’s intent by measuring human cognitive activity through body and brain sensors. The embedded system in turn translates the intent into robot control signals to interact with the physical environment on the human’s behalf via robotic actuators. Finally, the human closes the loop by observing the physical world interactions as input for making new decisions.

Examples of HiLCPSs include brain-computer inter-face (BCIs), controlled assistive robots,1 and intelligent prostheses.

A prototyping platform and a design framework for rapid exploration of a novel human-in-the-loop application serves as an accelerator for new research into a broad class of systems that augment human in-teraction with the physical world.

Gunar Schirner, Deniz Erdogmus, and Kaushik Chowdhury, Northeastern University

Taskin Padir, Worcester Polytechnic Institute

The Future of Human-in-the-Loop Cyber-Physical Systems



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 37

to monitor the respiratory rate, pulse oximetry, and skin resistance can help provide a more comprehensive view of the whole body.

The challenge with analog interfaces is in accurately detecting electrophysiologi-cal signals: electric potentials can be as low as in the mV range. Moreover, connect-ing a human to a host of wires to centrally gather these signals is not only impractical but also too restrictive. The optimum solu-tion is to have a distributed sensor network with power-efficient and reliable commu-nication, for example, through wireless body area networks (WBANs).

Human intent inferenceHiLCPSs put high demands on intent

inference algorithm design because input signals are inherently noisy. Intelligent sensor fusion can help compensate for inconsistent measurements of individual sensors and form a complete, coherent picture from the multimodal sensor input. One approach to deal with the interpreta-tion of noisy signals is to take the physical world context into account, eliminating contextually impossible decisions such as actions that are not physically possible given the current state of robot control, or letters that do not make sense given the language used for typing interfaces. Be-cause a HiLCPS continuously interacts with the physical environment, real-time intent inference is crucial for keeping up with the constantly changing environment.

Fundamental Autonomy for Functionally Locked-In IndividualsA human-in-the-loop cyber-physical system (HiLSPS) can offer assistive

technology that helps to restore fundamental autonomy—self-feeding and communication—for people who are functionally locked-in due to various neurological or physical conditions. Depending on their clinical diagnosis and condition, these individuals might have full cognitive capabilities yet lack the ability to execute any motor actions that can generate movement or speech. Consequently, they rely heavily on caregivers to accomplish everyday tasks.

We are developing a HiLCPS that augments the neurophysiological capabili-ties of a functionally locked-in individual to facilitate self-feeding, communi-cation, mobility, and digital access. As depicted in Figure A, we intend to build a brain-computer interface (BCI)-controlled wheelchair as a mobility platform, con-struct a robotic arm for self-feeding, and establish a communication interface. In addition to restoring the ability to meet basic needs, a HiLCPS can help to close the digital divide, making it possible for users to access the informational and social resources that computers offer and contributing to a sense of self-fulfillment that is essential for a productive life.

Figure A. Restoring fundamental autonomy for functionally locked-in individuals through a HiLCPS.

Sensory perception/processing

Body/brain sensors Embedded system(HW/SW)

Inferenceengine/control

Environment

Wirelessbody areanetwork

Human Cyber Physical

Sensors Actuators

Figure 1. Human-in-the-loop cyber-physical system (HiLCPS). The loop consists of a human, an embedded system, and the physical environment.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






38 COMPUTER

COVER FE ATURE

Robotics with shared governanceIn spite of recent advances in robotics research, the

design and control of robotic systems capable of au-tonomous operation remains a challenge. In the HiLCPS context, additional issues arise—for example, robots oper-ate in close proximity with the human, which poses strict safety requirements.

Decision algorithms also must divide governance be-tween human and machine. While the human can make top-level decisions, their local realization is better done autonomously. The overall aim is to require only concep-tual decision making from the human. However, safety overrides might be required to avoid implausible actions, depending on the overall physical state.

Building a HiLCPS requires tackling various multidisci-plinary design challenges, including

efficient embedded system design;cognitive intent detection algorithms using brain or other neurophysiological signals;actuator and robotics to realize an intended outcome or effect in the physical world; and distributed sensor architectures with suitable, power-efficient communication mechanisms.

We use a holistic design process to approach these multidisciplinary challenges. Our envisioned methodol-ogy and unifying framework for HiLCPS design offers an automated path for implementing body/brain computer interface (BBCI) algorithms for intent inference as well as for robot control on an embedded real-time platform. Automating the path to implementation lets algorithm developers explore real-time integration and simplifies exploring the shared human/machine governance.

A HOLISTIC DESIGN FRAMEWORKA HiLCPS must be realized in an efficient embedded

implementation that fulfills both functional and nonfunc-tional requirements. Unfortunately, algorithm designers typically are not embedded systems experts, so they need an integrative framework that bridges disciplines. Ideally, such a framework allows algorithm designers to achieve embedded implementations at the push of a button.

The key to HiLCPS adoption and integration is an efficient, robust, and reliable embedded implementation. As Figure 2 shows, an embedded control platform is at the heart of the HiLCPS that we are building. The sensing inputs (primarily EEGs) are directly connected through a specialized analog front end (AFE) and digital-to-analog converter (DAC). Auxiliary sensors interface with the embed-ded control platform through a body area network. Sensor fusion and intent inference algorithms execute mainly on an embedded processor assisted by a custom hardware compo-nent implemented in a field-programmable gate array. The FPGA is essentially dedicated to signal preprocessing to clean up the noisy input signal. A network interface trans-mits top-level decisions to the robotic actuators, which in turn interact with the environment.

In addition to being reliable and efficient, the algorithms developed for intent inference and robotics navigation/control also must be robust from a nonfunctional per-spective. Of particular concern are maintaining power efficiency—to allow battery-powered operation—and meeting real-time performance constraints as mandated by interaction with the physical world. In addition, fusing sensor data from multimodal distributed sensors and shared human/robot governance demands distributed operations.

Traditionally, algorithm design and its embedded implementation were approached sequentially, first by al-

ESL tool suite

Controller

CPU FPGA

Memory bus

IFC IFC

CPU.bin HW_1.v

AFE / DAC

Act. I

F

{EEG, ECG, … EMG}

Wheelchaircontrol,

home automation

NetIF

Body areanetwork

Brain/Body

Sensor

Brain/Body

Sensor

Brain/Body

Sensor

Brain/body

sensorNet

RoboticActuator

RoboticActuator

RoboticActuator

Roboticactuator

AlgorithmsRobotSensing

Figure 2. Framework for automatically generating embedded code (HW/SW) from abstract brain-computer interface algorithms. An electronic system-level (ESL) tool suite analyzes the algorithm at its input for computation complexity. It then generates a distributed implementation across hardware and software.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 39

gorithm and then embedded system experts, respectively. But this sequential process creates a long delay between algorithm conception and its embedded realization. In ad-dition to prolonging the time to market, this delay forces algorithm designers to make simplifying assumptions about the physical environment until the algorithms finally are translated to a real platform. Consequently, much of the cross-discipline optimization potential is lost.

Using a holistic methodology for developing design automation concepts can overcome most problems and consequences associated with sequential design.2 In our methodology, designers develop algorithms in the high-level languages with which they are familiar. They then enter the input algorithm together with a description of the underlying platform—constant, in our case—into the electronic system level (ESL) tool suite, which analyzes computation demands on the granularity of function blocks. The ESL tool suite generates code for both the CPU and the FPGA. As part of its overall synthesis process, the tool inserts interface code for hardware/software com-munication automatically.3 In effect, the ESL tool suite operates as a system compiler, as it compiles a high-level application to run atop an embedded platform across hard-ware/software boundaries.

The ESL flow paves the path for cross-discipline op-timizations. For example, it allows exploring different distributions of sensor fusion and intent inference. An event classification algorithm could directly execute on an intelligent sensor, which would increase the pro-cessing demand on the sensor but dramatically reduce communication—transmitting just the events of interest instead of a constant stream of data samples. This design freedom helps embedded architects devise low-power, high-performance systems.

BBCI researchers can use the automation framework to develop embedded algorithms without requiring specialized embedded knowledge. The ESL flow hides implementation-level details, enabling designers to focus on the important issue of algorithm and model develop-ment. Through automation, BBCI researchers will be able to directly test their algorithms in an embedded setting, enabling the development of a new class of real-time al-gorithms that can exploit the combination of sensing, analysis, and decision making.

CONTEXT-AWARE SENSING OF HUMAN INTENT

The use of multimodal physiological signals from the operator’s body and brain is an established idea in human-computer interaction and more broadly in human interface design for control systems. With the advent of portable and affordable systems and increased computing power in recent decades, growing interest has focused on the use of physiological signals easily measurable from the skin

on the arms or legs and also from the scalp. EEG, which measures electric potential on the scalp, has become the mainstay of noninvasive BCI design.4 There is considerable interest in developing not only BCI-controlled systems but also systems that combine EEG and other physiological signals such as EMG and gaze position.

The convergence of improved and less costly technolo-gies now makes it possible to develop prototypes using these multimodal input mechanisms for HiLCPS applica-tions. We foresee that, in the next decade, commercial applications using such interfaces will emerge. Some start-up companies in the gaming and entertainment markets are already experimenting with these ideas and are offer-ing reasonably successful products. Of course, the real challenge is to design commercial systems with a higher threshold for success in terms of accuracy, robustness, response time, and reliability.

That said, it is much harder to reliably infer a user’s intent with physiological signal-based interface designs than with a classical joystick or keyboard. Electrophysi-ological signals are inherently orders of magnitude noisier than their engineered electromechanical interface coun-terparts. Consider, for example, recently popularized speech- and gesture-based interfaces, which are struggling with many real-world issues such as relevant source sepa-ration in ambient noise or relevant object segmentation with moving background clutter as they find their role in the marketplace.

BBCI designs that rely on signals like EEG and EMG are prone to similar problems in terms of signal-to-noise and interference ratios. Clearly, the operator’s brain and body are being used for other internal physiological functions that have nothing to do with the intent that needs to be conveyed to the BBCI system. Therefore, although careful signal processing design and feature engineer-ing are crucial, they might not be sufficient in some cases. The low signal-to-noise and interference ratios simply make incorporating context a requirement in intent inference.

To improve the intent inference accuracy, BBCI design-ers must develop algorithms that adaptively take into account the current application as well as the operator’s preferences and historical behavior. For example, a BCI-based keyboard interface uses EEG traces to select letters, but there is room for improvement in the prediction suc-cess rate. Incorporating language models, which capture

The ESL flow hides implementation-level details, enabling designers to focus on the important issue of algorithm and model development.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






40 COMPUTER

COVER FE ATURE

the likelihood of character sequences, in the inference logic with proper Bayesian fusion helps to achieve improved re-sults in terms of both accuracy and speed. Figure 3 shows the speed/accuracy tradeoff for different n-gram model orders on the RSVP Keyboard, a recently developed BCI-based keyboard interface.5

Generalizing from this example, the principle of utiliz-ing contextual information and application-specific priors as routinely prescribed in machine learning theory be-comes essential and could make the difference between success and failure—that is, the human’s acceptance or rejection of the system/interface.

In our robotics applications, we are building and modi-fying contextual information and probabilistic models of

desired behavior and outcome sequences, with the intent of creating HiLCPS designs that will eventually operate successfully in the real world. For this, we use tools from adaptive signal processing, machine learning, and robot-ics when they are available, and we develop new tools and methods when necessary. Although generic recursive Bayesian modeling and inference procedures have been well established, existing parametric and nonparametric models that can be used within these frameworks might be insufficient. We anticipate that most of the effort in incorporating contextual information and application-specific priors in inference and intent detection will be spent on modeling.

So far, we have developed a preliminary brain-controlled robot prototype, shown in Figure 4, that allows an operator to remotely navigate a robotic platform, such as a wheelchair, using steady-state visual evoked potentials (SSVEPs) induced by flickering light patterns in the operator’s visual field. A monitor shows four flickering checkerboards that emit periodic square waves with different frequencies. Each checkerboard and frequency corresponds to one command to control the robot.

In our prototype, the four commands represent the de-sired target locations D1, D2, and D3 as well as the stop command. To select a command, the operator focuses his or her attention on the desired checkerboard on the monitor. After the operator focuses on one checkerboard, the visual cortex predominantly synchronizes with the checkerboard’s flickering patterns—fundamental and harmonic frequencies. Like any brain activity, the visual cortex’s activity will result in voltage fluctuations that can be measured on the scalp. Accordingly, we place an elec-

1/3 1/2 10

0.2

0.4

0.6

0.8

1.0

Speed

Accu

racy

Not first letter of word

0-gram1-gram4-gram8-gram

Flickering checkerboards Predefined destinations

EEG amplifier

High-passfilter and 60-Hz

notch filter

Classification Context-sensitiveprobablistic filter

InternetTCP/IP

Global planner

Sensor model(human)

Transitionmodel (context)

Local planner

Destination

D2

Stop D1

D3

Figure 3. Letter selection among 28 possible symbols dem-onstrating accuracy versus speed for the preliminary RSVP Keyboard design when fusing EEG evidence from multiple trials for each symbol. Speed is inversely proportional to the number of trials.

Figure 4. System architecture for the BCI-based control of an intelligent wheelchair as an example of a HiLCPS. (left) The semi-autonomous wheelchair receives brain signals from the user for a high-level activity; (right) it then executes the tasks of path planning, obstacle avoidance, and simultaneous localization and mapping.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 41

trode on the scalp near the occipital lobe where the visual cortex is located to pick up these EEG signals.

Distinguishing which checkerboard the operator has focused on requires analyzing the power spectrum of signals gathered close to the visual cortex. The power spectrum shows the power contri-bution over different frequencies; Figure 5 shows a typical power spectrum for a flickering frequency of 15 Hz. Clearly visible is the peak at 15 Hz, which is the fundamental frequency, as well as at 30 Hz and 45 Hz, which are the second har-monic and third harmonic, respectively. By recognizing the peaks in the power distribution, the system can infer the checkerboard the operator is focusing on, and thus identify which command the operator would like to select. The in-ference system then sends the detected command via TCP/IP to the robot for execution.

In a more advanced approach, this frequency encoding can be replaced by showing different pseudorandom sequences, for example, m- or Gold-sequences. Then, the system can use template matching or other temporal-model ing-based decision-making mechanisms.

Figure 6 shows the average responses of the visual cortex for four separate 31-bit m-sequences flickering the visual stimuli at 30 bits per second.6 Notice that the visual cortex response clearly varies for different m-sequence visual inputs. Thus, this approach might make it possible to utilize ideas from digital communication theory, where commands can be encoded with unique signature sequences and in-ference methods, robust to interference from neighboring flickering objects, and more reliably distinguished (as in CDMA communications) using code-based filter-ing. This latter approach might further have the benefit that natural brain activity can strongly influence the power spectrum, so pseudorandom-code-based stimuli can be potentially detected and classified more reliably than frequency-based visual stimuli.

In our first prototype design, we found that an operator could achieve greater than 99 percent accuracy in selecting between four commands for the robot using one-second EEG signals per decision along with the m-sequence coded

flickering paradigm to induce SSVEPs in the brain with simple nearest-template classification. Further research is needed to make more commands available in the limited visual field, which requires better signal processing be-cause interference from nearby flickering patterns causes a reduction in accuracy.

Our future research will focus on improving signal pro-cessing and statistical inference with the help of better physiological signal modeling and improved contextual

0 10 20 30 40 50

10−4

10−5

10−6

10−7

Frequency (Hz)

Mag

nitud

e(dB

)

Alphaactivity

15 Hzfundamental

15 Hz2nd fundamental

15 Hz3rd fundamental

0 0.2 0.4 0.6 0.8 1−15−10−5

05

1015

Oz te

mpla

te EE

G (μV

)

−15

−10−5

05

1015

Oz te

mpla

te EE

G (μV

)

−15−10−5

05

1015

Oz te

mpla

te EE

G (μV

)

−15

−10−5

05

1015

Oz te

mpla

te EE

G (μV

)

Time (ms)0 0.2 0.4 0.6 0.8 1

Time (ms)

0 0.2 0.4 0.6 0.8 1Time (ms)

0 0.2 0.4 0.6 0.8 1Time (ms)

M-sequence 1 M-sequence 2

M-sequence 3 M-sequence 4

Figure 5. SSVEP power spectrum for a flickering LED visual stimulus blinking on/off following a 15-Hz square wave with 50 percent duty cycle.

Figure 6. SSVEP average waveforms in response to four separate m-sequences controlling the flickering patterns of checkerboard patterns displayed on an LCD monitor. All signals are measured at Oz in the international 10-20 con-figuration, with the stimulus located central to the visual field at a distance of approximately 60 cm.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






COVER FE ATURE

42 COMPUTER

modeling. We still need to include other physiological sig-nals and signal processing on the embedded platform for optimal real-time performance and integration.

ROBOT ASSISTIVE TECHNOLOGYRobotics, the integration of sensing, computation, and

actuation, is integral to a HiLCPS as robots provide the in-teraction with the physical world.7 Even though robotics research and enabling technologies for practical appli-cations are making significant progress, it is essential to develop novel methodologies for the design, modeling, and control of robotic systems that can work safely with people in shared spaces. Modular and reconfigurable designs, plug-and-play integration of cyber and physical components, composability, and optimizing the role of the human com-prise a short-list of current HiLCPS research challenges.7

New applicationsWhile robot assistive technologies cover a range of ap-

plications from helping persons with autism to eldercare to stroke rehabilitation, an essential area of research is the development of intelligent wheelchairs and safe robotic arms to assist physically locked-in individuals.1 State-of-the-art wheelchair-arm systems can perform obstacle avoidance, simultaneous localization and mapping (SLAM), path planning, and motion control with a shared autonomy framework.

However, important research questions for implement-ing shared control of an intelligent system still remain: Who controls the system—human or machine—and when? Under what circumstances does the human or the machine override a decision? How can HiLCPSs decide adaptively on the level of autonomy? Early efforts in the development of smart wheelchairs tackled these issues by providing the user with an external switch or button to trigger a change in operation mode. Another approach is to imple-ment the mode change automatically, where the shared control switches from human control to machine control and vice versa.8

Within the experimental setup and control architecture for a HiLCPS testbed developed by our research teams at Northeastern University and Worcester Polytechnic In-stitute, the semiautonomous wheelchair receives brain signals from the user for a high-level task such as Navigate-to-Kitchen, and then executes the tasks of path planning, obstacle avoidance, and SLAM. Using the robot operating

system for wheelchair navigation results in a modular com-munications and software design.

Shared controlIn the HiLCPS context, there are certain tasks in which

humans are and probably always will be superior to robots, such as perception, intuitive control, and high-level deci-sion making. On the other hand, robots can and probably should perform tasks such as precise low-level motion planning, solving an optimization problem, and operat-ing in dirty, dull, and dangerous situations. Therefore, the investigation of new control interfaces and shared control methods that can effectively delegate tasks and blend the control between robots and human operators will make it possible to field robot systems that act in direct support of humans.

We can classify most currently deployed robots in two categories: fully autonomous, performing specific tasks; and teleoperated, with little to no intelligence. Although not all human-robot interactions fall into these two catego-ries, they represent most currently available systems. The development of control techniques that will dynamically shift the level of control between the human operator and the intelligent robot will be the key to increased deploy-ment of HiLCPSs.

Within this shared control framework, addressing the tight physical interaction between the robot and human remains a key research problem.9 To operate robots in close vicinity of humans, global safety protocols should be developed, and fail-safe modes should be implemented to realize a practical system. In addition, force and tactile sensing interfaces can be used for physical human-robot interaction to enable safer operation of a robot near a human operator.

Modularity and reconfigurabilityModular and reconfigurable robot design is another

important aspect of engineering the future HiLCPS. Modu-larity requires reusable building blocks with well-defined mechanical, communication, and power interfaces. It allows low-cost development, reusable hardware and software components, and ease of maintenance as well as improvements in design time and effort.

Reconfigurability brings together modules such as sen-sors, actuators, and linkage in various configurations to compose robotic systems for environment interaction. Modular and reconfigurable cyber and physical com-ponents will enable the accelerated and cost-effective composition of HiLCPSs.

WIRELESS BODY AREA NETWORKS AND PHYSIOLOGICAL SIGNALS

In addition to EEGs, auxiliary sensors that measure physiological changes in blood pressure, muscle activity,

Robots can and probably should perform tasks such as precise low-level motion planning, solving an optimization problem, and operating in dirty, dull, and dangerous situations.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 43

and skin conductivity, among others, can significantly enhance a HiLCPS’s capabilities. In addition to intent in-ference, these sensors can detect any sudden abnormal changes or stress that the human cannot otherwise com-municate. In addition to alerting caregivers to medical emergencies, such a sensor network also provides a con-text for the inference engine/control module. Thus, an assistive WBAN makes it possible to understand both the intent and the condition of the human signaling that intent.

A WBAN comprises small interconnected sensors that can either be placed noninvasively on the subject’s body or surgically implanted within it. These sensors can monitor a wide range of physiological and emotional states and communicate this sampled data to a centralized mon-itoring entity.10 Because our goal is to develop an open platform for the holistic and automatic design of embedded HiLCPSs, our work will address several unique architec-tural and functional characteristics of WBANs related to the limitations of energy (especially for implanted sensors), heterogeneity, and interference.

Our general approach leverages the human body as the communication channel, resulting in significant reduction in the energy used compared to RF transmission using electromagnetic waves. In this new body-coupled com-munication (BCC) paradigm, the signals are placed through electrical impulses directly in or on the surface of human tissue, at the point of data collection by the sensors. As a key motivation, the energy consumed in BCC is shown to be approximately 0.37 nJ/bit—three orders of magnitude less than the low-power classic RF-based network created through IEEE 802.15.4-based nodes.

The wide variety of available monitoring applications requires transmitting periodic scalar data or continuous pulses, ranging from cardiovascular state monitoring to one-shot emergency notifications, such as indicating the onset of an epileptic seizure, that must take precedence over all other forms of periodic monitoring. The high bandwidth availability of BCC, approximately 10 Mbps, sufficiently accommodates the needs of such varied sensor measurements. Moreover, this form of communication offers considerable mitigation of fading as it is not impaired by continuous body motion and disruption to a clear line of sight, as is common in the RF environment. This allows for simpler modulation and signal generation/reception schemes that the sensor’s limited onboard capability can accommodate.

Finally, BCC can overcome the typical problems of ex-ternal interference in the ISM band’s various channels, which typically carry transmissions from wireless local area networks, including Bluetooth and radiation from microwave ovens. However, this injected signal should remain in the 100-KHz to 60-MHz range: at the lower end of the frequency scale, there is a risk of interfering with the internal and implanted electrical signals from devices within the human body, such as a cardiac pacemaker. At the higher end, above 100 MHz, the average height of a human body approaches the same length as the signal’s wavelength, making it function as a lossy antenna and causing it to radiate the energy externally.

Our work mitigates this potential problem by using the electrical circuit-equivalent representation of the body channel,11 in which different types of body tissue—skin,

Z (bone)

Z (skin)

Z (muscle)

Z

Electrical circuit representationof the body channel

Signal path at frequency f1

Equivalent impedance

Signal path at frequency f2

Sensor

Electrodeimpedance

Figure 7. Overview of the WBAN through body-coupled communication. Different types of body tissue—skin, fat, muscles, and bone—each offer varying but measurable levels of signal impedance.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






COVER FE ATURE

44 COMPUTER

fat, muscles, and bone—each offer varying but measurable levels of signal impedance, as Figure 7 shows. As an exam-ple, the fat layer’s relative permittivity varies from 1.0 E + 2 at 100 KHz to 2.0 E + 1 at 5 MHz. This frequency-specific change in signal conduction levels must be considered in link-layer design. Moreover, the electrodes’ contact imped-ance means that the signal is differentially applied to the input point.

Our link-layer design uses this channel model to iden-tify the loss of signal strength and construct simple error correction schemes that ensure reliable packet delivery. If multiple sensors report a signal being forwarded to a distant pickup point on the body, the total charge density must be less than 350 μC/cm2, which will determine which sensors can concurrently access the body channel to send their measurements.12

Depending on node placement, the transmitting sensor will also need to optimize both the injected signal power and the frequency, as these signals propagate to a differ-ent extent within the human tissue and along the surface distance from the generation point. This will lead to power/frequency tuples uniquely assigned to each neighbor node, such that only a single node is addressed with that com-bination, further reducing the packet header lengths and interference possibility.

As an initial demonstration, we will use skin conduc-tivity and muscle activity monitoring sensors placed at five locations—both palms, both arms, and the torso. These sensors will send inputs periodically via BCC to a predetermined collection point on the body, from which the physiological data will be transferred to the embedded control system. The channel characteristics will define the required complexity of signal modulation and error correc-tion capability, which both the signal-generating sensor and the embedded controller must support. These inputs will provide clues to the system when the human operator reg-isters his or her intent. For example, heightened stress levels alter skin conductivity and cause involuntary muscle action, a factor that influences the subsequent robotic actuation.

The BCC paradigm will usher in a new communication method for the BCI system to gather enhanced knowledge of the human condition, which will empower it to make better situational decisions on the needs of the integrated intent-decision closed-loop system.

H iLCPSs offer an exciting class of applications both for restoring or augmenting human interaction with the physical world and for researchers faced with

the interdisciplinary challenge of combining semiautono-mous robotics, WBANs, embedded system design, and intent inference algorithm development.

Our holistic design methodology enables cross-disciplinary optimizations and facilitates the cross-polli-

nation of ideas across four previously disjoint disciplines, thus leading to otherwise unachievable advances. In addi-tion, our outlined project establishes an open prototyping platform and a design framework for rapid exploration of a novel human-in-the-loop application, serving as an accelerator for new research into a broad class of cyber-physical systems.

AcknowledgmentsThis article is based on work supported by the National Science Foundation (NSF) under award nos. CNS-1136027, 1135854, IIS-0914808, and IIS-1149570, as well as the National Institutes of Health (NIH) 5R01DC009834. Any opinions, find-ings and conclusions, or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of NSF or NIH.

References1. X. Perrin et al., “Brain-Coupled Interaction for Semi-

Autonomous Navigation of an Assistive Robot,” Robotics and Autonomous Systems, Intelligent Robotics and Neuro-science, vol. 58, no. 12, 2010, pp. 1246-1255.

2. D.D. Gajski et al., Embedded System Design: Modeling, Syn-thesis and Verification, Springer, 2009.

3. G. Schirner, R. Dömer, and A. Gerstlauer, “High-Level Development, Modeling and Automatic Generation of Hardware-Dependent Software,” Hardware-Dependent Software: Principles and Practice, W. Ecker, W. Müller, and R.Dömer, eds., Springer, 2009.

4. J. Wolpaw and E.W. Wolpaw, eds., Brain-Computer Inter-faces: Principles and Practice, 1st ed., Oxford University Press, 2012.

5. U. Orhan et al., “RSVP Keyboard: An EEG-Based Typing In-terface,” Proc. IEEE Int’l Conf. Acoustics, Speech and Signal Processing (ICASSP 12), IEEE, 2012, pp. 645-648.

6. H. Nezamfar et al., “Decoding of Multichannel EEG Activ-ity from the Visual Cortex in Response to Pseudorandom Binary Sequences of Visual Stimuli,” Int’l J. Imaging Sys-tems and Technology, June 2011, pp. 139-147; http://dx.doi.org/10.1002/ima.20288=0pt.

7. E. Lee, “Cyberphysical Systems: Design Challenges,” Proc. 11th IEEE Int’l Symp. Object-Oriented Real-Time Distributed Computing (ISORC 08), IEEE, 2008, pp. 363-369.

8. I. Iturrate et al., “A Noninvasive Brain-Actuated Wheel-chair Based on a P300 Neurophysiological Protocol and Automated Navigation,” IEEE Trans. Robotics, June 2009, pp. 614-627.

9. S. Ikemoto et al., “Physical Human-Robot Interaction: Mutual Learning and Adaptation,” Robotics & Automation Magazine, Dec. 2012, pp. 24-35.

10. H. Cao et al., “Enabling Technologies for Wireless Body Area Networks: A Survey and Outlook,” IEEE Communica-tions Magazine, Dec. 2009, pp. 84-93.

11. Y. Song et al., “The Simulation Method of the Galvanic Coupling Intrabody Communication with Different Signal Transmission Paths,” IEEE Trans. Instrumentation and Measurement, Apr. 2011, pp. 1257-1266.

12. D.P. Lindsey et al., “A New Technique for Transmission of Signals from Implantable Transducers,” IEEE Trans. Biomedical Eng., vol. 45, no. 5, 1998, pp. 614-619.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


http://www.qmags.com/clickthrough.asp?url=http://dx.doi.org/10.1002/ima.20288=0pt&id=17866&adid=P44E1





JANUARY 2013 45


Gunar Schirner is an assistant professor in the Department of Electrical and Computer Engineering at Northeastern University. His research interests include embedded system modeling, system-level design, and the synthesis of em-bedded software. Schirner received a PhD in electrical and computer engineering from the University of California, Irvine. He is a member of IEEE. Contact him at [email protected].

Deniz Erdogmus is an associate professor in the De-partment of Electrical and Computer Engineering at Northeastern University. His research focuses on statistical signal processing and machine learning with applications to contextual signal, image, and data analysis with applications in cognitive signal processing, including brain-computer interfaces and technologies that collaboratively improve human performance. Erdogmus received a PhD in electrical and computer engineering from the University of Florida. He is a senior member of IEEE. Contact him at [email protected].

Kaushik Chowdhury is an assistant professor in the Department of Electrical and Computer Engineering at Northeastern University. His research interests include wireless cognitive radio networks, body area networks, and energy-harvesting sensor networks. Chowdhury received a PhD in electrical and computer engineering from the Geor-gia Institute of Technology. He is a member of IEEE. Contact him at [email protected].

Taskin Padir is an assistant professor in the Robotics Engineering program at Worcester Polytechnic Institute. His research interests include robot control, cooperating robots, and intelligent vehicles. Padir received a PhD in elec-trical and computer engineering from Purdue University. He is a member of IEEE. Contact him at [email protected].

stay connected.Keep up with the latest IEEE Computer Society

publications and activities wherever you are.

| IEEE Computer Society| Computing Now

| facebook.com/IEEEComputerSociety| facebook.com/ComputingNow

| @ComputerSociety| @ComputingNow

TM

| youtube.com/ieeecomputersociety



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_______________

________

__________________

__________

http://www.qmags.com/clickthrough.asp?url=www.youtube.com/ieeecomputersociety&id=17866&adid=P45A1

http://www.qmags.com/clickthrough.asp?url=http://www.facebook.com/ComputingNow&id=17866&adid=P45A3

http://www.qmags.com/clickthrough.asp?url=http://www.facebook.com/IEEEComputerSociety&id=17866&adid=P45A2










COVER FE ATURE


(online courses, games), and contemplative (filmmaking) domains. Experiential mediation is evolving quickly and increasing the amount of discrete hybrid physical-digital knowledge. However, few existing integration structures can merge these instances into a coherent system of reflective knowledge and values that is foundational to a truly hybrid physical-digital culture. In turn, the lack of such a mature culture is hindering the development of integrative cultural solutions to complex, societal problems.

We propose the emergence of a truly integrative hybrid physical-digital culture, which will require the broad in-tegration of knowledge across engineering, arts, design, humanities, and the social sciences. It will also demand a value system and related design approaches that em-phasize the integration of efficiency, contemplation, and quality of experience. Society can then advance a genera-tion of experiential media systems in which utilitarian applications engage an activity’s meaning—not just the activity itself—and in which contemplative forms are metaphors of the emergent complexity of the everyday, communication and social applications promote reflec-tive transactions, and learning systems motivate through experiential quality.

HYBRID KNOWLEDGEEarlier cultural epochs typically included a small privi-

leged class of individuals versed in the arts and sciences and imbued with decision-making authority. At the same time, such cultures also contained rich folk traditions within the masses, with a slowly evolving overlap between the two. The Age of Reason, for example, saw the prolif-

T he global problems that impact today’s densely interconnected society present stark resource, energy, and sustainability challenges. Rather than being singular component problems, as in

the need for better data analysis, these complex issues are primarily cultural, as in the need to develop a sustain-ability culture. As such, understanding and tackling them requires exploring culture as an integration of knowl-edge and values arising from the diverse experiences of its constituents.

From this perspective, utilitarian, social and communi-cation, structured learning, and contemplative experiences contribute to developing an integrative body of knowledge and values. For example, the contemplative experience of reading a metaphorical passage can help the reader re-think everyday moments, while the structured compilation of carefully observed experiences can give rise to broadly applicable metaphors.

Many experiences now involve digital media. Expe-riential media—the integration of computational media in everyday experiences—is apparent in the utilitarian (traffic networks), social (participatory media), learning

Multidisciplinary value structures and a design approach focusing on combining efficiency, reflection, and quality of exper-ience will foster the true hybrid physical-digital culture that is foundational to solv-ing complex societal problems.

Thanassis Rikakis and Aisling Kelliher, Carnegie Mellon University

Nicole Lehrer, Arizona State University

Experiential Media and Digital Culture



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


VI

DE

OV

ID

EO







JANUARY 2013 47

eration of popular texts and the emergence of amateur female musicians.

Analysts now understand these patterns as catalysts in the emergence of a flattening popular and transactional culture, with smaller outposts remaining within the folk or traditional and the trained practitioner realms. The op-portunity lies in integrating these varying forms in a way that elevates their richness, depth, and cultural longevity in potentially transformative ways.

The philosophical field of phenomenology advances the case that much of human knowledge is empirical and arises from embodied, physical experience. Phenom-enology contrasts this empirical approach to exclusive rationalist or Cartesian theories that approach knowledge as a purely mental representation of the world. Building on the principles of phenomenology, hermeneutics intro-duces the notion of interpreting experiences in relation to both personal understanding and the experiences of other cultural members. Paul Ricoeur, a leading herme-neutics’ proponent, attests that nothing is simply “given” to consciousness; rather, mediation and interpretation arise through language, symbol, culture, and history.1

Within the specific context of technology, phenome-nology, hermeneutics, and embodied cognition theories have significantly aided the development of successful principles for intuitive human-computer interaction and the emergence of embodied computing and experiential media—the latter being systems that integrate computing and digital media with the physical and social experience. The main goal of these hybrid media systems is to facilitate and enhance an individual’s experience in a way that will lead to more integrated knowledge.

EXPERIENTIAL MEDIA SYSTEMSFigure 1 is a continuum of experiential media sys-

tems representing areas of activity that contribute to the generation of cultural knowledge and values. Although some examples, such as advanced gaming controllers and smartphones, are encountered mostly in postindustrial na-tions, other systems, such as musical instrumentation and cameras, are more widespread. Along the continuum are various degrees of purposeful engagement and experien-tial activity—how deliberately the system involves the user in the digital experience. Systems on the far left represent little or no direct engagement, while systems at the far right allow the user to create and shape the experience.

Sensor networksTraffic control systems combine sensor networks, compu-

tational analysis tools, and control feedback signals as part of a smart environment that rarely directly engages users.

The computational complexity and intelligence of these systems is growing to the point that adaptive models can now allow traffic signals to partially self-organize.2 These systems learn from users while the users learn to func-tion within the systems’ constraints. Although the result benefits society by improving traffic flow, users rarely consider the computational intelligence in relation to their own travel activities. Without meaningfully engaging in key parts of the everyday hybrid physical-digital experi-ence, users will find it difficult to productively reflect on that experience.

Digital thermostats are another example of a sensor network, but in these systems, user engagement goes up a notch. Thermostats highlight the notion that even the slightest direct engagement with an experience’s digital elements can produce baseline meaningful interaction. Just the responsibility of briefly adjusting the device causes users to consider its location, how its output relates to the room temperature they experience, and how a seemingly unrelated activity, such as leaving a door open, can affect that relation. The brief involvement of thermostat setting causes reflection on all these factors.

Swiping an access card with a magnetic strip is another example of a simple sensor system interaction that can generate baseline knowledge. Most people realize that the strip contains digital information and understand that the sensor reader accesses that information based on the user card movement. During this simple interaction, the user can also reflect on issues such as digital identity, even though the sole engagement activity with the swipe system is utilitarian: to make a purchase or gain building access.

It is easy to create an intuitive experience with a simple system that addresses a narrow function like access. However, as direct engagement with digital components increases, developing solid experiential mediation and related knowledge by carefully considering demonstrated principles of experiential media design will become more important. For example, the mouse and graphic display interface took advantage of key innate human motor and spatial reasoning abilities to allow intuitive control and organization of the computing workspace and to facilitate computing access for the nonexpert.

Little or nodirect digitalengagement

Direct expertdigitalengagement

Figure 1. Distribution of experiential media systems. At the far left are sensor networks that require little direct engagement. At the far right is filmmaking, which requires direct expert participation. (Mouse icon design courtesy of Camila Bertoco, smartphone icon design courtesy of Stefan Bumbeck, Kinect design courtesy of Luca Erbifori from The Noun Project.)



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






48 COMPUTER

COVER FE ATURE

SmartphonesModern smartphones and tablets have extended com-

puting beyond the desktop and into daily living. These devices have advanced experiential mediation to the point at which users now interact with complex digital media at the level of meaning, as opposed to just carrying out an ac-tivity. For example, users navigating new territory can view a map or directions on the smartphone touchscreen and use pinching gestures to rescale the map. The computer extracts and reacts to the gesture’s meaning, opening to expand and closing to compress; the action is more than a simple gesture trace.

Although reading a map on a smartphone might be construed as utilitarian because the user needs direc-tions or location information, this interaction can also support user social and entertainment activities. Texting and social media communication (tweeting, status up-dates) are farther to the right on the continuum depicted in Figure 1. Together, the device and applications not only

enhance users’ understanding of their environment but also cause them to reflect on the complex mediation network of sensors, transmitters, analytical tools, and displays involved in their experience. Thus, engagement in the digital part of this experience increases, and the result is the generation of true hybrid physical-digital knowledge.

Social activitiesSocial entertainment and communication activities

strongly orient around hybrid and customized content con-sumption. Users can start and stop a Netflix movie, select a music genre on a Pandora station, and organize pictures in iPhoto. Selecting and customizing created content gener-ates a fairly elaborate relationship with a variety of digital delivery components, and arguably, customization is the first step toward content creation.

Digital recordings are easy to edit and remix using simple, intuitive interfaces such as Gimp and Audacity. Content capture has also become much easier. Smart-phones have one-click photo and videocameras, and many inexpensive but dependable interfaces enable digital recording. Indeed, the compilation of experiential media—smartphones, touchscreens, gestural interfaces, and intuitive experience recording and editing—has sim-plified and popularized amateur content creation. The many available avenues for publishing amateur content, such as Flickr and YouTube, incentivize casual creative

activity and content sharing and make it easier to grow the supply of user-generated content.

These opportunities to create content are changing the generated and collective knowledge within the digital com-munity. The amateur content creator is gaining not only digital knowledge (how to touch up a picture), but also complex hybrid knowledge (how pixelation relates to light and sampling definition). This knowledge might not be born of rigorous formal training, but it is powerful enough to enrich social media and contribute to the growth of social media platforms.

Beyond being a way to share interests, stories, events, and photos, Facebook is primarily a mechanism for constructing and experiencing a digital community. In this process, participants become significantly involved with computational facets, including issues of complex computer networks, recommendation mechanisms, and connectivity.

Online gaming similarly engages participants in the digital aspects of their gaming experiences. In some of the more advanced participatory games, such as World of Warcraft and Guild Wars, players can design digital elements within the game, as well as actively create fan art outside the game. Before computing’s popularization, games of all kinds encapsulated rich physical experiences with full engagement of every aspect of a player’s mental and physical capacity.

In modern gaming and in social media in general, users are increasingly engaged in their experience’s digi-tal components, which is leading to an increase in hybrid knowledge. In some particular applied instances, such as mediated therapy, the amount of digital knowledge might even be overwhelming physical knowledge.

Learning and educationExperts in mediated learning have long recognized the

importance of games in furthering knowledge. Modern computer games have adaptive and participatory struc-tures that allow for active, customizable learning along with offering sheer entertainment value. This combina-tion is proving very effective for 21st-century learning.3

Online learning systems, such as MITOpenCourseWare and Coursera, are enabling knowledge dissemination at unprecedented levels, even questioning some of the basic assumptions of knowledge-delivery financial models. The creators of these learning vehicles are deeply engaged with the systems’ digital components to the extent that their experience of and concepts about learning and education have changed.

Professional content creators are also directly engaged with digital creation, and most already are or are be-coming hybrid knowledge experts. Much contemporary music is created through interaction with digital tools—from acoustic sound that is mixed or sequenced through

Social entertainment and communication activities strongly orient around hybrid and customized content consumption.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 49

digital means to music performed on hybrid (physical-digital) instruments, and music composed using com-puter algorithms and performed by computers.

Filmmaking is an equally hybrid experience, again ranging from primarily physical performance captured and post-produced through digital means, to films in which some or all of the content is realized through digital means, including 3D animation, special effects, and machinima. Finally, expert videogame creators also exhibit hybrid skills and approaches as they combine interaction design and storytelling (Shigeru Miyamoto), computation ( John Carmack), real-world problems (Jane McGonigal), and complex systems (Will Wright). All these professional expert creators of original, rich experiences have deep hybrid knowledge that can prove critical to advancing the maturity of a hybrid physical-digital culture.

EVIDENCE OF FRAGMENTED CULTUREAs the “Effects of Fragmented Culture” sidebar de-

tails, most of these examples exist in a fully or partly discrete space, as opposed to coexisting in an integra-tive metaspace—a fragmented state that is hampering the evolution of a truly hybrid physical-digital culture. Examples of this fragmentation are evident in experi-ential media systems from traffic control to the arts and social media.

Traffic controlThe engineers developing sensor networks and al-

gorithms for analyzing and controlling traffic are now working with transportation experts, city planners, and, occasionally, public policy experts, includ-ing social scientists and economists. Conspicuously absent are the perspectives of artists, designers, and humanists, despite the long history of such view-points generating rich experiences motivating change in beliefs, behavior, and knowledge. Rather, devel-opers assume that speed and efficiency or financial incentives are sufficient to motivate behavioral change. As Nobel Laureate Lee Hartwell recently re-marked at the Emerge Symposium (http://vimeo.com/tag:emerge2012asu), the first and most important step in achieving sustainability is directly engaging people, not analyzing behavioral data on the disengaged.

Learning systemsThe idea of engagement as a key path to learning

continues to drive significant advances in serious games, educational games, and mediated learning. However, many mediated learning efforts are not inte-grating the full range of expertise required to produce gaming designs, in which the quality of the experience is the main incentive for engagement. Instead, they at-

tempt to incentivize engagement through the need to win (gain points), to discover what happens next, or even just to enjoy the novel digital gadgets.

Developers are missing the idea that people play games to have a rich and rewarding experience. Many of the most

Effects of Fragmented CultureW ebster’s defines one understanding of culture as “the

characteristic features of everyday existence (as diversions or a way of life) shared by people in a place or time” (www.merriam-webster.com). It also offers an alternative definition, “the integrated pattern of human knowledge, belief, and behavior that depends upon the capacity for learning and transmitting knowledge to succeeding generations.”

By the first definition, a hybrid physical-digital culture already exists, and indeed many characteristic features in everyday experience are hybrid and produce various levels and types of hybrid knowledge. However, the second defini-tion speaks to an integrative body of deep knowledge, in which society builds beliefs and reflective behaviors. In that sense, current hybrid experiences are overly fragmented and thus cannot yet produce the deep transferable knowledge and values associated with culture. Underlying that fragmen-tation is the communicative distance between the people and communities who develop and study experiential media.

Earlier work1,2 argued that the creation of meaningful mediated experiences requires the integration of knowledge from the arts, design, engineering, science, and humanities. Although connectivity among these areas is increasing, inte-grative discourse and activity is still not widespread. The most significant fragmentation occurs between the quantita-tive constituents of experiential media development (engineering and science) and the qualitative components (arts, design, and humanities). Merging the different approaches and priorities of these contributing communities is difficult.

Fragmentation among system developers leads to biases in design that affect subsequent usage patterns of experien-tial media and diminish the value of the resulting knowledge. In the 1930s, John Dewey decried the “hurried and impatient human environment … where what is called experience becomes so dispersed and miscellaneous as hardly to deserve the name.”2 Similarly, Phoebe Sengers argues that much expe-riential technology development is based on optimization (Taylorism) models, with the main goal being to do more and do it faster. The result is a culture with optimization as the leading value; people consistently try to do more until they exhaust themselves and seek refuge in mindless entertain-ment.3 Digital technology clearly has a role in the misunderstanding of the relationship between doing and undergoing.

References1. H. Sundaram and T. Rikakis, “Experiential Media Systems,” Ency-

clopedia of Multimedia, 2nd ed., B. Furtht, ed., Springer Verlag, 2008.

2. J. Dewey, Art as Experience, Penguin Group, 1934.3. P. Sengers, “The Engineering of Experience,” Funology: From

Usability to Enjoyment, M. Blythe et al., eds., Kluwer, 2003.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________

http://www.qmags.com/clickthrough.asp?url=http://vimeo.com/tag:emerge2012asu&id=17866&adid=P49E2

http://www.qmags.com/clickthrough.asp?url=www.merriam-webster.com&id=17866&adid=P49E1





50 COMPUTER

COVER FE ATURE

intriguing and fun games, such as hide and seek, embody the physical, involve others, and promote continuous discovery, such as finding the best places to hide and an-ticipating where the searcher will go next. Similarly, people play music primarily for the quality of the experience and personal creative discovery, not to win points or find out what happens next or because they want to experience a new gadget. Most people play music they have played before, and they are deeply attached to their familiar and comfortable instruments.

Rehabilitation is another learning domain that suf-fers from a lack of rich experiences. Therapists rely on off-the-shelf computer games and culturally uninformed incentives to rehabilitate adults who yearn for high-quality

experiences. Adults whose quality of life has suffered be-cause of a health condition are not interested in beating a random computer character and can find the recom-mendation to play these games patronizing. Fortunately, alternative designs of interactive rehabilitation systems are beginning to use more meaningful feedback structures.4

Arts and entertainmentIncreasing embodiment in experiential media is also

constrained by fragmented development expertise. An example is Dance Central, the first popular dance game cre-ated for the Kinect interface. The game, which requires that all players face the screen, relies on traditional computer vision and computer graphics perspectives as it mimics players’ movements. However, most dancing is not simply mimicking but rather an interaction in which dancers imitate and improvise. It is about moment-to-moment interaction but it also involves interaction that draws on space, time, body relation patterns, and a fine balance of repetition and variation. Advanced dance and choreog-raphy knowledge could have provided insight to create a more realistic, engaging, and culturally aware game.

Music performance embodies centuries of knowledge about how to achieve mastery of performance involving complex interfaces and how to tackle motor control issues related to that objective. However, little of this knowledge has found its way into labs that develop embodied com-puting interfaces and control and movement-assistance systems. Theme park creators know that humans use all their senses to explore novel, complex environments, and they leverage that idea by using audition to monitor com-plex time patterns, vision for spatial relations, and touch for detail investigation and intimacy. However, complex

data navigation remains primarily a two-dimensional visualization-based activity.

Social mediaSocial media platforms aim to increase the size of target

communities and the number of transactional exchanges within them (consequently generating advertising revenue). This goal seems to largely preclude paths for generating the deep contemplation and high-level creative activity that produce and sustain a mature culture. Many important cultural eras, notably Athenian democracy, the Renais-sance, the Age of Reason, and Modernism, formed through rich social interactions that advanced the scientific, social, learning, and creative aspects of cultural development in an integrated manner. Although much earlier dialog was constrained to the middle and upper classes, modern so-ciety can move beyond this privileged cultural separation by using digital media to engage a broader mass audience. The “Evolving the Artistic Perspective” sidebar describes the role of the design and arts communities in exploiting this opportunity.

TOWARD A DIGITAL CULTUREAlthough we have highlighted some unfortunate trends,

there is also reason to be optimistic given the number of ongoing efforts to promote the integrative development of experiential media systems. For example, many design and art schools now require undergraduates to be proficient in programming by their sophomore year. Computer sci-ence programs are also promoting minors that connect computing to the humanities and arts. Digital media un-dergraduate and graduate programs combining arts and engineering faculty are rapidly growing at Arizona State University (ASU), the University of Southern California, and the University of California at Santa Barbara, among others.

Media-centric curriculaSchools that recognize the importance of hybrid physical-

digital media are combining arts and engineering faculty and generating innovative curricula that are media- rather than discipline-centric. ASU has developed an under-graduate degree in digital culture (https://digitalculture.asu.edu) that combines 50 courses from 12 academic units spanning engineering, design, arts, and social and cogni-tive sciences. These courses form a network connected via a set of 20 proficiencies, which are metalevel learning outcomes. Students gain a proficiency by taking courses from a range of disciplines through different paths, such as obtaining a modeling and inference proficiency through computer science, industrial design, digital music, or an-thropology courses. Thus, a digital music major might take computer science or vice versa.

The ASU curricula and similar programs strike a bal-ance between media-centric outcomes that are discipline

Developers are missing the idea that people play games to have a rich and rewarding experience.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 51

agnostic and disciplinary expertise that can bring many years of cultural practice knowl-edge to the mix. As such, they pose a model for 21st-century liberal education, in which students aim for balanced knowledge across myriad dimensions of the human experience and can choose their own path to achieve that knowledge.

We have informally observed this model in action and have seen it produce student agency and innovation, such as students designing capstone projects aimed at the potential future job they wish to create—rather than simply accept—within an existing company. More formal studies are needed to determine how such communities can advance the integra-tive development of mediated experiences and culture.

Integrative industrial communitiesIndustry is also promoting integrative com-

munities for the development of experiential media. Apple has long viewed the overall ex-perience provided by its products to be the driving outcome rather than any one indi-vidual component of that experience. Intel has created an interaction and experience research group that combines engineering, computing, cognitive and social science, and arts and design expertise. Architecture and design companies such as Diller Scofidio + Renfro (architecture, visual arts, and per-forming arts), and Hoberman Associates (art, architecture, and engineering) specialize in developing integrative approaches for transforming the built environ-ment. Their diverse workforce produces a broad range of outcomes—from video art to complete buildings and transformable building facades.

Industrial design firms, such as Herman Miller, are also driving innovation in product, service, and commercial development, where social change and sustainability are core company values. Frogdesign describes its company as creating “connected experiences that span multiple technologies, platforms, and media” (www.frogdesign.com).

These exemplary industrial leaders epitomize the clear value of cultivating diverse teams of experts and integra-tors collaborating on complex problems that they can only solve by working respectfully in close tandem.

Integrative projects A growing collection of individual projects is also ex-

ploring integrative approaches to experiential media and digital culture. We are deeply familiar with the projects we

describe here, which span disruptive arts, aesthetics in the classroom, performance practice, mediated rehabilitation, and experience capture.

Disruptive arts. Disrupting our everyday experience is the goal of interventionists, who provide surprising mech-anisms and tools for raising awareness and inviting action. The arts community has long used unexpected reversal to generate contemplation and reflection; an example is Aristotle’s Poetics. Current works explore the idea of reverse surveillance and self-surveillance. Neal White’s Dark Places (2009) is a field guide to secret intelligence space in the UK. David Rokeby’s Guardian Angel (2001) challenges notions of benevolent control as part of the surveillance apparatus. Camille Utterback’s Abundance(2007) uses public plaza movement as the driver for archi-tectural projection. All these works tackle the notion of the watcher and the watched from a variety of critical perspectives.

Injecting pleasure, benevolence, playfulness, and confrontation into social interactions with everyday sur-veillance technologies produces project outcomes that

Evolving the Artistic PerspectiveE xperiential media is prevalent in postindustrial nations, but ensuring its

continued maturity requires design and arts expertise. This prevalence and need for artistic expertise offer the opportunity to structure a modern culture that emulates older, more balanced models in which arts and humanist perspectives influenced everyday experience and in turn elevated its quality.

However, directly involving artists in experiential media systems design and development will require a significant change of perspective within the profes-sional arts communities. In this new perspective,

deterministic creation models make space for probabilistic approaches; paradigms of singular creators who develop finished products make space for participatory models with evolving forms; andpractices that aim for concrete block experiences, as in presenting a com-plete musical composition in one location at one time, make space for experiences in which smaller instances of related sound and music aggre-gate to create one sonic experience (occuring in different ways and at different times).

Most important, arts practice must consider what David Wessel calls “the emerging large amateur base” (http://cnmat.berkeley.edu). Wessel proposes that mediated creative practice tools and processes must have a low entry bar-rier, thus attracting a large amateur base, but an unlimited ceiling, thus encouraging continuous aspiration.

This proposition is similar to the idea of playing the piano. A five-year-old child can play a simple tune on the piano after a week of practice, but the ceil-ing for playing the piano or composing piano music is unlimited. Likewise, a five-year-old child can use social media with very little practice, but the poten-tial for endless creative enrichment and transformation is more constrained.

Providing avenues for social media tools and processes to support the gen-eration and refinement of advanced cultural knowledge would infuse users with cultural aspirations and lead to the development of expert creative practi-tioners who could contribute to the growth and value of a large community.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___

http://www.qmags.com/clickthrough.asp?url=www.frogdesign.com&id=17866&adid=P51E2

http://www.qmags.com/clickthrough.asp?url=http://cnmat.berkeley.edu&id=17866&adid=P51E1





COVER FE ATURE

52 COMPUTER

move beyond the technology used, the data generated, or the fixed art product created. Rather, the outcome is a complex, contemplative, interactive process.

These examples highlight a key concept of 21st-century creative practice: the contemplative outcome—the direct reflective participatory process—is the overt focus rather than the actual product or installation.

Classroom aesthetics. David Tinapple and Loren Olson at ASU’s School of Arts, Media and Engineering are rethinking the mediated learning paradigm as the practice of creative aesthetics. They are helping middle school students in the Herberger Academy create products of high aesthetic value using advanced digital processes, including algorithmic programming, 3D modeling and printing, image aggrega-tion, and motion rendering. The overt focus is on creating aesthetically rich artifacts, but to reach that goal, students need to master complex science, technology, engineering, and mathematics (STEM) concepts. The engagement is very high from inception; students indirectly learn STEM con-cepts, but the learning environment remains focused on the quality of the experience.

Tinapple and Olson’s process shows the value of a cultural model in which STEM is a means for having a contemplative rich experience rather than an end in itself. This model avoids the awkward practice of masquerading a learning exercise as a game and using points or badges as an incentive to complete it.

Performance practice. At Carnegie Mellon University, Roger Dannenberg has also turned a traditional gaming paradigm on its head. Instead of converting creative prac-tice outcomes into an entertainment game, as in Guitar Hero, he has developed a signal recognition algorithm that follows actual guitar players, and tutors them on their errors via the Rock Prodigy app. Popular reviews and the commercial success of Dannenberg’s app indicate that users are willing to pay more to experience increasing levels of creative challenge. This finding also supports the idea of using an unlimited creative ceiling as an incentive rather than winning points.

Mediated rehabilitation. Stroke is the leading cause of serious, long-term disability in the US. The World Health Organization describes disability as an experience that is influenced by

the health condition’s underlying pathophysiology (the stroke),

impairments at the bodily function and structural level (muscle activity patterns), disability at the activity level (task performance), handicaps at the participation level (life and self-care involvement), andvarious personal and environmental factors.

Given that inclusive definition, effective stroke rehabili-tation must address physiological constraints, motivation for participation, trust, information communication, the therapy’s adaptability, and the long-term sustainability of therapy and outcomes. Consequently, experience design must integrate optimization with reflection and quality of experience within a broader awareness of cultural context.

Over the past six years, we have worked collaboratively with experts from engineering, computing, design, arts, medicine, and neuroscience in developing a mixed-reality system that can aid the rehabilitation of a stroke survivor’s upper extremities.

Figure 2 shows a diagram of how image size coupled with harmonic progression indicates progress to a target. The system tracks key components of movement and pro-vides multimodal digital feedback to help users execute and plan their movements. A recent study shows that the system can enhance physical rehabilitation by advancing movement quality to aid in functional recovery.5

Experience capture. These examples point to a grow-ing body of work and support structures for developing hybrid experiential systems that fall on the right side of the continuum in Figure 1. The development of such sys-tems returns value not only from the final product, but also—and more essentially—from the development pro-cess itself and the issues being considered. This idea of deriving value from considering the process of developing hybrid experiential systems is at the heart of our recent project to capture and summarize event experiences. Our ultimate goal is to provide insights into how a hybrid cul-ture emerges.

For the three-day Emerge symposium held at ASU in March 2012, we developed a multimodal documentation framework for capturing and representing the event expe-rience (http://emerge.asu.edu) and presented findings from this work as part of a four-month interactive exhibition. The captured and assembled data will be available online in late January 2013.

We also aim to disseminate exemplar, current, and in-progress work integrating science, engineering, art, and design through our virtual eXchange for Science, Engineering, Arts and Design (XSEAD) portal (http://xsead.org), which is sponsored by the National Science Foundation.

Our goal is to reach diverse audiences, such as expert practitioners, emerging researchers, parents and stu-

Primary participation in physical experiences offers many ways to ignite contemplation and increase the appreciation of life’s deeper meaning.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____

______

http://www.qmags.com/clickthrough.asp?url=http://xsead.org&id=17866&adid=P52E2

http://www.qmags.com/clickthrough.asp?url=http://emerge.asu.edu&id=17866&adid=P52E1





JANUARY 2013 53

dents, and the general public with a rich variety of performance animation code, video, and prototypical artifacts. These works can stem from distinct sources including workshops, cultural probes, site-specific work, open source frameworks, and participatory design.

We anticipate that the XSEAD portal will highlight hybrid collaboration and the active process of engagement as paramount in crafting experiential systems of societal and cultural value.

Funding supportCreating and sustaining a truly hybrid digital cul-

ture requires funding and support from agencies and foundations willing to invest in transdisciplinary proj-ects that produce diverse outcomes. The NSF’s former CreativeIT program supported transformative projects spanning creativity, computer science, and informa-tion technology. The Andrew W. Mellon Foundation has been instrumental in supporting several initiatives, including establishing the Center for Art, Science, and Technology at the Massachusetts Institute of Technol-ogy and providing generous support to the ArtsEngine at the University of Michigan.

The MacArthur Foundation and the Bill and Melinda Gates Foundation recently awarded Electronic Arts more than $10 million to create the Games, Learning and As-sessment Lab (GlassLab) to develop next-generation educational games and assess game-related learning outcomes. Corporations and private donors are also ex-hibiting interest in supporting art-science integrative work. The Frank-Ratchye Studio for Creative Inquiry at CMU recently received a $1 million naming endow-ment to establish the Fund for Art and the Frontier, while one of the new Intel-funded centers at Georgia Tech is directed toward humanities-driven social computing research.

T he aggregate experience that experiential media produces needs to integrate optimization with con-templative value and quality of experience in terms

of both sensory engagement and depth of meaning. For example, readers return to a Shakespearean play again and again because it communicates its meaning effec-tively (even optimally), uses beautiful language (engaging in form), and promotes contemplation of individual and societal topics. Each rereading of one of these plays offers an opportunity to extract new meaning. In contrast, sys-tems and experiences that do not integrate deep meaning structures try to attract attention and sustain engagement through surface ornamentation.

Many social media practices reinforce this point, fo-cusing on continuous redesign or the addition of features to maintain attention. In contrast, primary participation in physical experiences offers many ways to ignite con-templation and increase the appreciation of life’s deeper meaning. As evidence, consider how a short walk can provide diversions that induce deep contemplation and meaningful discovery.

As experiential media systems evolve and encapsulate a large part of daily life, promoting development approaches that integrate diverse expertise is crucial. This will result in architectures that address optimization within the bigger context of quality of experience and meaningful knowledge and ultimately will support a more harmonious culture and society.

References1. R. Kearney, Paul Ricoeur: The Hermeneutics of Action, Sage,

1996.2. M. Buchanan, “Why Complex Systems Do Better without

Us,” New Scientist, 6 Aug. 2008; www.newscientist.com/article/mg19926681.500-why-complex-systems-do-better-without-us.html.

Distance

Space

Velocity

Speed

Subject

z’: 0.290.0 0.19 0.850.630.50 1.0

IV V7/IV viIma7

V7 I ii7

ReachRest

1.00.8 0.80.6 0.60.4 0.40.2 0.20.0 0.0z’/sec:

Particledisplacement

Chordprogression

Activity

Harmonicprogression

ReturnGrasp

Figure 2. Mixed-reality rehabilitation system. Image size couples with harmonic progression to indicate the user’s progress to a target. The system uses arts principles to develop optimal displays of high aesthetic value and to generate creative para-digms such as metaphoric mappings and participatory narrative development to engage and motivate the user.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________________________________________

http://www.qmags.com/clickthrough.asp?url=www.newscientist.com/article/mg19926681.500-why-complex-systems-do-better-without-us.html&id=17866&adid=P53E1





Featuring industry technology solutions you can use.

Visit http://computingnow.computer.org

NEW + EXPANDEDIEEE COMPUTER SOCIETY’S COMPUTING NOW WEBSITE

Subscribe today for the latest in computational science and engineering research, news and analysis, CSE in education, and emerging technologies in the hard sciences.

www.computer.org/cise

COVER FE ATURE

54 COMPUTER

3. H. Jenkins, Confronting the Challenges of Participatory Culture: Media Education for the 21st Century, MacArthur Foundation, 2006.

4. N. Lehrer et al., “Exploring the Bases for a Mixed Reality Stroke Rehabilitation System,” J. Neuroeng. and Rehabilita-tion, Oct. 2011, pp. 1-15.

5. M. Duff et al., “Adaptive Mixed-Reality Rehabilitation Improves Quality of Reaching Movements More Than Traditional Reaching Therapy Following Stroke,” Neuro-rehabilitation and Neural Repair, 2013; sagepub.com/journals/Journal201625.

Thanassis Rikakis is vice provost for design, arts, and tech-nology at Carnegie Mellon University. His research interests include experiential media, mixed-reality rehabilitation, and computer music. Rikakis received a DMA in music from Columbia University. He is a member of IEEE. Contact him at [email protected].

Aisling Kelliher is an associate professor in the School of Design at Carnegie Mellon University. Her research in-terests include interaction design, multimedia, and social networks. Kelliher received a PhD in media arts and sci-ences from the Massachusetts Institute of Technology. She is a member of ACM. Contact her at [email protected].

Nicole Lehrer is a doctoral student in the School of Arts, Media and Engineering at Arizona State University. Her research interests include mixed-reality rehabilitation and interactive graphics. Lehrer received a BSE in biomedical engineering and a BFA in painting from Tulane University. She is a member of ACM. Contact her at [email protected].




qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___________________

_____

__________

______________

__

______________

http://www.qmags.com/clickthrough.asp?url=http://computingnow.computer.org&id=17866&adid=P54A2

http://www.qmags.com/clickthrough.asp?url=www.computer.org/cise&id=17866&adid=P54A1


http://www.qmags.com/clickthrough.asp?url=www.sagepub.com/journals/Journal201625&id=17866&adid=P54E1








computer.org/software

More value, more content, more resourcesThe new multi-faceted IEEE Software offers exclusive video and web extras that you can access only through this enhanced digital version. Dive deeper into the latest technical developments with a magazine that is:

Searchable—Quickly fi nd the latest information in your fi elds of interest. Access the digital archives, and save what’s most relevant to you.

Linked—Click on table of contents links and instantly go to the articles you want to read fi rst. Article links go to additional references to deepen your new discoveries.

Engaging—Experience concepts as they come to life through related audio, video, animation, and more. Email authors directly. Even apply for jobs through convenient ad links.

Mobile—Read the issue anytime, anywhere, at your convenience—on your laptop, iPad, or other mobile device.

Now Available in Enhanced Digital Format



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________

http://www.qmags.com/clickthrough.asp?url=www.computer.org/software&id=17866&adid=P55A1








COVER FE ATURE


nology is already being used at many medical facilities, including the Forth Valley Royal Hospital in Scotland. At this facility, which opened in July 2011, robots deliver meals, sort hospital mail, move bed linen, and dispose of clinical waste through a dedicated network of corridors beneath the hospital. Further plans call for robotic beds to transport patients to the operating room and remotely controlled robots to conduct ward rounds to monitor pa-tients’ postoperative recovery.

But autonomous surgery, in which a machine performs the operation entirely without human assistance, could be many years away. Duke University researchers have made some progress in developing a robotic arm that takes input from ultrasound sensors to guide a plunger device in taking biopsy samples. The system is still at an early stage, and has so far been demonstrated using turkey breasts in place of human breast and prostate tissue.

In the future, there will no doubt be debates about whether surgical procedures should be undertaken with-out the direct supervision and presence of people. It is not clear how acceptable autonomous surgery would be to patients.

ROBOTIC SURGERYRobotic surgery systems fall into three main categories:

controlled systems that depend entirely on a surgeon’s actions, translating them into precise movements; auto-matic systems that perform surgery directly after being programmed by a surgeon or other human operators; and semi-automatic systems that constrain a surgeon’s

S onia Dickerson, 24, of Brooklyn, New York, entered the Flushing medical clinic and went into one of the private examination booths. She inserted her insurance card in the machine in

front of her and stared fixedly into the retinal identifier. “Okay,” the machine stated, “please specify your symp-toms.” After Dickerson told it what was wrong with her, the machine instructed her to remove her clothing and lie down in the booth’s scanner. A few minutes later, the procedure was complete.

“The only medical option available to you with your insurance policy is automatic surgery,” the machine an-nounced. Would you like to proceed?” “Yes,” she replied, and the machine activated a conveyor belt in the scanner that took her to a surgical machine. Later that day, Dicker-son left the medical center to eat at her favorite vegetarian restaurant, the New Bodai.

Although science fiction, this scenario offers a glimpse of a potential future in automated healthcare. Such tech-

Robotic surgery is a relatively new technol-ogy that promises unparalleled innovation with minimization of pain and time spent in recovery, but surgeons must follow re-alistic ethical guidelines if this treatment option is to thrive.

Noel Sharkey and Amanda Sharkey, University of Sheffield, UK

Robotic Surgery: On the Cutting Edge of Ethics



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 57

movements. The “Robotic Surgery Systems” sidebar provides examples and describes these systems in more detail.

Like many new technologies, robotic surgery promises innovation and improvement over conven-tional methods, but only time will tell if it proves to be truly revolutionary. As the engineer and futurist Roy Amara famously stated, “We tend to overestimate the effect of a technology in the short run and under-estimate the effect in the long run.” While investing in robotic surgery does not guarantee its immediate success, we believe that the technology ultimately may be highly beneficial to humanity and is thus worth pursuing.

Consider, for example, an idea proposed by the nonprofit research institute SRI International. With funding from the US Defense Advanced Research Proj-ects Agency, SRI developed a prototype “trauma pod,” a portable covered operating table that enables surgeons working at a remote location to perform acute stabi-lization procedures on wounded soldiers who might otherwise die waiting for treatment. Robots would fetch a wounded soldier from the battlefield and place him in a trauma pod to undergo remotely controlled scrubbing (cleanliness is one of the great challenges of battlefield surgery), diagnostic tests (including x-rays), and selected surgical intervention.1

Imagine the great benefit of this technology to the civilian world. Many people die in ambulances en route to a hospital after being injured in a serious accident. Trauma pods could be taken to the scene so that sur-geons could remotely render victims safe for travel. Similarly, remotely located surgeons could operate on patients in disaster areas that are too dangerous to visit or difficult to reach.2

However, the future of robotic surgery will be highly dependent on patient outcomes and on the public’s perceptions of how good the technology is. People must receive fair and reasonable treatment that re-spects their human rights and dignity. The promise of a transformational technology must not blind us to the underlying ethical issues.3

ETHICAL CHALLENGESThe first recorded robotically assisted surgical pro-

cedure was performed in 1985, yet discussions of its ethical application have been limited.4-7

To gain traction on this issue, we must first con-sider the ethics of innovative surgery in general. The challenge is to achieve a reasonable balance between stifling overregulation and underregulation that leads to harmful and dangerous practices.

Francis Moore8 offered three criteria for making sur-gical innovation acceptable:

Robotic Surgery Systems

T hree basic types of robotic surgery systems are in use today.Controlled systems are entirely dependent on the surgeon’s

direct actions. Intuitive Surgical’s da Vinci Surgical System (www.intuitivesurgical.com/products/davinci_surgical_system), which was cleared for use by the US Food and Drug Administration (FDA) in 2000, is an example of a “hands-on” robot designed for soft-tissue surgery in the abdominal cavity.

A surgeon observes and controls the da Vinci system’s three to four interactive arms, which are equipped with various instru-ments and a high-resolution endoscopic camera, while seated at a console. The system is similar to the Zeus Robotic Surgery System developed earlier by Computer Motion (which merged with rival Intuitive Surgical in 2003) and cleared by the FDA in 1994, but its jointed-wrist arms have more degrees of freedom. The normal mode is for the surgeon to operate the console near the patient in the same room, but the da Vinci system has also been operated by surgeons who perform the procedure while located at a consider-able distance from the patient.

Controlled robotic surgery systems are designed to improve upon traditional minimally invasive surgery, also known as laparo-scopic or keyhole surgery, in which the surgeon performs the operation through small incisions using a camera inside the body. Such systems have a wide range of applications including gastroin-testinal surgery, gynecology, cardiothoracic surgery, orthopaedics, and neurosurgery.

In contrast, automatic systems follow precise, preprogrammed instructions given by a surgeon. Several of the first such systems were adapted from industrial robots. For example, ROBODOC,1

which is used for total knee and hip arthroplasty, was developed in the early 1990s by Integrated Surgical Systems (now Curexo Tech-nology) based on a selective compliant assembly robot arm (SCARA) device. The surgeon uses 3D data from a computed tomography scan of the patient to plan the implant location and orientation, then ROBODOC executes the plan.

Another example is the CyberKnife Robotic Radiosurgery System (www.cyberknife.com), an adapted KUKA industrial robot that uses multiple beams of high-dose radiation from various angles for high-precision, noninvasive irradiation of cancerous cells. Imaging software provides a continuous update of the patient’s position, and allows the robotic manipulator to correct for changes in patient position during treatment. The system does not require the head-mounted frames needed for other radiosurgical systems such as Gamma Knife.

Surgeons’ unease at being merely observers of programmed systems has led to the development of semi-automatic systems such as the Acrobat Company’s Acrobat Sculptor for orthopaedic surgery.2 Rather than the surgeon cutting bone directly, the robot actively constrains the surgeon to cut accurately within a safe region by means of a force-controlled handle. Similarly, NeuroArm (www.neuroarm.org) also constrains a neurosurgeon’s move-ments. The device resides in a magnetic resonance imaging machine to enable 3D imaging of the patient’s brain in near real time. The surgeon sets up a surgical “corridor” that provides force feedback for any attempt to move outside it.

References1. J. Pransky, “ROBODOC—Surgical Robot Success Story,” Industrial Robot,

vol. 24, no. 3, 1997, pp. 231-233. 2. P.-L. Yen and B.L. Davies, “Active Constraint Control for Image-Guided

Robotic Surgery,” Proc. Institution of Mechanical Engineers, Part H: J. of Eng. in Medicine, vol. 224, no. 5, 2010, pp. 623-631.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___

http://www.qmags.com/clickthrough.asp?url=www.neuroarm.org&id=17866&adid=P57E3

http://www.qmags.com/clickthrough.asp?url=www.cyberknife.com&id=17866&adid=P57E2

http://www.qmags.com/clickthrough.asp?url=www.intuitivesurgical.com/products/davinci_surgical_system&id=17866&adid=P57E1





58 COMPUTER

COVER FE ATURE

sufficient laboratory experience,sufficient intellectual and technical expertise, and “institutional stability.”

In line with these guidelines, hospitals conducting robotic surgery must ensure that surgeons have the requisite level of knowledge and experience, the right tools and re-sources, and a well-trained support staff familiar with the technology.

Jane Johnson and Wendy Rogers9 raise four main ethi-cal challenges associated with surgical innovation. As a discipline, robotic surgery must meet these four challenges if it is to move forward and demonstrate its revolutionary potential.

Compromised informed consent. Difficulties in achieving valid informed consent for innovative procedures could put patient autonomy at risk. Barriers to valid consent in-clude patients’ unawareness of a procedure’s novelty, the unavailability of evidence about the risks and comparative effectiveness of innovative surgery, and a general tendency to equate newness with increased benefit. Trust in the authority of a surgeon or institution can exacerbate this tendency.

Conflicts of interest. A surgeon might be influenced unconsciously by the career benefits and elevated social status associated with being an innovator. A surgeon might also be biased toward a novel device due to a large

investment in training time. The institution likewise could have dedicated considerable resources to a new technol-ogy and enjoy the reputation of being on the leading edge. All of these motives can give rise to conflicts of interest.

Harm to patients. Surgical intervention increases the risk of mortality and morbidity compared to most diagnostic procedures. Surgery in itself is not benign, and there are dan-gers associated with the use of anesthesia and longer hospital stays, such as infection. Finan-cial or psychological harm as well as loss of trust in the medi-cal profession are also possible consequences.

Unfair allocation of healthcare resources. Surgeons and institu-tions alike tend to overestimate the positive effects of new, typically costlier, treatments. Surgical innovations that are

more expensive than existing treatments can divert lim-ited healthcare resources from less expensive and possibly safer and more effective options.

COMPROMISED INFORMED CONSENTIntuitive Surgical could not have asked for a better ad-

vertisement for its da Vinci Surgical System. In July 2009, President Obama visited the Cleveland Clinic in Ohio and was photographed trying his hand at operating the system, named for the famous 15th-century Italian artist, scien-tist, engineer, and inventor of, among other wonders, the first humanoid automaton (www.intuitivesurgical.com/company/history). A da Vinci robot was also featured on the cover of U.S. News & World Report’s Best Hospitals issue in 2012, and one was even featured on the hit TV show Grey’s Anatomy. Figure 1 shows a promotional image of the system from the company’s website.

It is hardly surprising, then, that the system’s popularity has exploded. According to Intuitive Surgical, the number of robot-assisted prostatectomies performed in the US in-creased from about 5,000 in 2002 to 73,000 in 2009, an increase of 1,400 percent.10 The da Vinci system is currently the number one option among US patients seeking treat-ment for prostate and gynecologic cancer as well as for minimally invasive hysterectomies, and around 2,500 of the devices are installed in more than 1,900 hospitals worldwide (www.intuitivesurgical.com/company/profile.html).

Figure 1. Da Vinci Surgical System. A surgeon observes and controls the system’s robotic arms, which are equipped with various instruments and a high-resolution endoscopic camera, while seated at a console. (Image ©2013 Intuitive Surgical, Inc.)



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___________

http://www.qmags.com/clickthrough.asp?url=www.intuitivesurgical.com/company/profile.html&id=17866&adid=P58E2

http://www.qmags.com/clickthrough.asp?url=www.intuitivesurgical.com/company/history&id=17866&adid=P58E1





JANUARY 2013 59

Favorable media reporting is partly responsible for the system’s appeal. But critics have questioned the company’s claims and the glowing da Vinci ads on hospital websites. They point to a considerable gap between the optimistic, even euphoric, descriptions of the system made available to patients and the lack of clear supporting clinical evidence. Some have charged Intuitive Surgical and many hospital clients with exaggerating the system’s efficacy to attract business.

When Intuitive Surgical sought premarket approval of the da Vinci robot from the US Food and Drug Adminis-tration (FDA) in 1999, it did so on the grounds that using the system was equivalent to performing conventional laparoscopic surgery. The company never claimed that its system was superior. It stated that the objective of a three-month clinical trial of the system at a Mexico City hospital was to “demonstrate device equivalence in safety and effectiveness to the control of standard laparoscopic equipment in performance of general lapa-roscopic tasks.”11

In fact, several studies have shown that, at best, robotic surgery achieves results similar to other procedures:

A review of Medicare records for 8,837 men who un-derwent a prostatectomy between 2003 and 2007 concluded that those who underwent minimally inva-sive surgery (including robotic surgery) had a shorter hospital stay and experienced fewer respiratory and surgical complications and strictures, but were more likely to later suffer from genitourinary problems, incontinence, and erectile dysfunction.12

A study of 1,904 patients whose prostate glands were removed by experienced urological surgeons at one institution between 2003 and 2008 found that using robotic surgery was no more effective than using con-ventional open surgery.13

Of eight randomized-control studies of general surgi-cal procedures published between 2001 and 2009, seven showed no patient benefit from the use of ro-botics while the eighth showed only a decrease in postoperative hospitalization.14

In 2009, the Institute for Clinical and Economic Review examined the relative effectiveness of robotic surgery over 111 studies and rated robot-assisted prostatectomy among patients with low-risk prostate cancer as “unproven with potential” pending further long-term controlled studies.15

Despite these findings, a systematic analysis of 400 randomly selected US hospital websites found that 41 de-scribed robotic surgery and that 86 percent of these touted the technology’s clinical superiority.15 None of the websites mentioned potential dangers. The authors concluded that “materials provided by hospitals regarding the surgical

robot overestimate benefits, largely ignore risks, and are strongly influenced by the manufacturer.”

Interestingly, in 2001 the FDA twice warned Intuitive Surgical that the materials on its websites made mislead-ing claims for the da Vinci robot, stating that the system “has use in a variety of applications that have not received agency clearance.”16

Failing to state the risks associated with robotic surgery or the comparative effectiveness of other procedures in marketing material is clearly unethical and compromises informed consent.9 It reduces patient autonomy by failing to provide the full information required to select the best choice from among all available options. It also biases deci-sions in favor of one technology over others that might be safer and less expensive.

CONFLICTS OF INTERESTSurgeons and hospitals endorsing the da Vinci system

might also be subject to conflicts of interest. At the time of this writing, the Intuitive Surgical web-

site contains two promotional videos about the system (www.intuitivesurgical.com/specialties). One video pres-ents a string of testimonials from mostly female patients who have undergone da Vinci surgery, while in the other, a series of doctors explain to potential patients why “I am a da Vinci surgeon.”

Like anyone else, former patients and surgeons have the right to give their personal opinion. However, using such opinions to promote a controversial medical technology is ethically questionable, especially in the case of surgeons, in whom the public places high trust. Sick people and their families are vulnerable to claims about miraculous pain-free treatments, and the laudatory comments in the videos are punchy and edited for maximum effect.

Selling robotic surgery is very different from selling a bar of soap. When people are seriously ill, they need independent advice not tainted by commercial inter-ests. However, Intuitive Surgical’s website includes no negative opinions about the da Vinci robot, nor does it indicate any risks or possible complications associated with its system beyond the obligatory disclaimer in small print at the bottom of the page. Open abdominal surgery is mentioned as an alternative but not tradi-tional laparoscopic surgery, which studies have shown to be at least as effective as robotic surgery in many cases.

Failing to state the risks associated with robotic surgery or the comparative effectiveness of other procedures in marketing material is clearly unethical and compromises informed consent.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


http://www.qmags.com/clickthrough.asp?url=www.intuitivesurgical.com/specialties&id=17866&adid=P59E1





60 COMPUTER

COVER FE ATURE

“There’s a medical arms race,” declared Paul Levy, chief executive at Boston’s Beth Israel Deaconess Medical Center, to The Wall Street Journal in 2010.17 He initially resisted purchasing a da Vinci device because there was no reliable evidence that it was better for patients but gave in when it became clear that the hospital was losing business to competitors who had surgical robots. “Technologies are being adopted and becoming widespread based on the marketing prowess of equipment makers and suppliers,” he contended, “not necessarily on the public good.”

Dr. Jeffrey Cadeddu, a urologist at the University of Texas Southwestern Medical Center in Dallas, made simi-lar comments to The New York Times around the same time.10 Before switching to the da Vinci system, he had been conducting traditional laparoscopic prostatectomies for many years, but prospective patients were walking away. According to Cadeddu, “They say: ‘Do you use the robot? O.K., well, thank you.’ And they leave.”

As Francis Moore8 pointed out in a 2000 article on the ethics of surgical innovation, “The surgeon should be aware of the fact that patients threatened by severe illness display a surprising and sometimes alarming readiness to accept uncertainty and reach out for something new. The surgical scientist must avoid exploiting this willingness of patients to try something new in a desperate situation.” Surgeons should exercise care and not let conflicts of inter-est determine their course of action in choosing the best methods for patients.9

With intense marketing efforts fueling consumer demand for robotic surgery, small hospitals have been particularly susceptible to the commercial pitch. In 2010, 131 out of 853 US hospitals with at least one da Vinci system were small facilities with 200 or fewer beds.17 How-ever, the system requires a large capital outlay, costing from $1.1 million to more than $2 million depending on the model and requiring an annual service contract of up to $180,000.18 Because small hospitals typically do not have the volume of operations necessary to justify such an investment, surgeons can feel pressured to make up the difference by increasing the number of surgical procedures they perform.

A marketing staff typically is tempted to make hyper-bolic claims on behalf of a product its company makes or provides to clients. Their job does not require any special-

ized knowledge of the product, only what will help them sell it. While misrepresentation is perhaps too strong a word to describe the promotional material for the da Vinci and other robotic surgery systems, hospitals and surgeons must be cognizant of the potential for overly enthusiastic marketing of new, complex technological products about which little is known.

Consider, for example, charges of false advertising made against Curexo Technology’s ROBODOC system for total hip and knee arthroplasty.

An analysis of 10,000 ROBODOC operations per-formed in Germany between 1994 and 2004 revealed that patients were presented with unsubstantiated claims about the technology and received insufficient informa-tion about alternatives.4 The evidence did not point to superior outcomes for patients—in fact, a comparison study found that ROBODOC operations took longer than conventional manual procedures and caused more blood loss.19

The 2007 information leaflet of a private orthopaedic hospital in Essen, Germany, implied that ROBODOC was the best option: “You do not need to be a prophet to know that … operations on the bone skeleton will be increasingly performed by computer-controlled robotic processing, be-cause they guarantee a much higher precision and safety for the patient.”4 Although the risk of muscle damage from the system was known at the time, the leaflet made no mention of it; consequently, a former patient successfully sued the hospital for 30,000 euros.

In addition, a lawyer for hundreds of ROBODOC patients treated at another German clinic successfully argued that surgeons often did not respond appropriately to patients’ postoperative complaints of a permanent limp and fre-quent and considerable pain. He said that the surgeons played down the complaints and even suggested that these were risks associated with total hip arthroplasty, not use of the ROBODOC system per se.4

Caregivers must take great care to manage patients’ ex-pectations and provide them with substantiated evidence about potential outcomes for robotic surgery as well as available alternative procedures. Ignoring negative results and misrepresenting or even slightly exaggerating positive results from studies is morally unacceptable. Conflicts of interest must not impugn the integrity of the doctor-patient relationship.

HARM TO PATIENTSIn January 2007, Juan Fernandez, a 49-year-old Chicago

man, died two weeks after a surgical procedure in which a da Vinci robot was used to remove his spleen. He left behind a wife and young child. His family sued the doctor, who had never used the system before on a living person, for medical malpractice and was awarded $7.5 million. The attorney argued that using the robot, whose arms

Caregivers must take great care to manage patients’ expectations and provide them with substantiated evidence about potential outcomes for robotic surgery as well as available alternative procedures.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 61

had inadvertently punctured the man’s lower intestine, was unnecessary. “As our expert said, it was like using an 18-wheeler to get a quart of milk at the market. And they just weren’t careful.”18,20

Sherry Long, a 42-year-old woman from Rochester, New Hampshire, was told at Wentworth-Douglass Hospital in Dover that robotics would be used to perform her hysterec-tomy in March 2009 because it was less invasive than other methods. However, during the procedure, both of her ure-ters were severed, causing her kidneys to shut down and leaving her with permanent damage. In May 2010, Long filed a suit against her two surgeons for gross negligence, alleging that neither had sufficient training and experience to competently use the da Vinci robot. The suit acknowl-edged that the device “allows for better magnification of the surgical site,” but “the slightest improper movement by the physician controlling the device can result in cata-strophic injuries to the patient.”17,21

Two other female patients who underwent gynecological surgery at Wentworth-Douglass using the da Vinci system accidentally had their bladders lacerated, and one had to be sent to another hospital to have the injury repaired.17

These are not isolated cases. In June 2010, retired US Air Force Colonel David Antoon filed a suit against the Cleveland Clinic after complications arising from a ro-botically assisted prostatectomy that left him incontinent and permanently impotent. The suit alleged that Antoon had signed a consent form authorizing one Dr. Kaouk to perform the surgery and that the surgeon had expressly agreed to this condition. However, a trainee present during the surgery reportedly did not recall Dr. Kaouk ever par-ticipating, and Antoon suspected that the trainee actually performed the procedure, and Dr. Kaouk, if he was even there, merely observed.22

The FDA panel had a robust discussion about training procedures before granting da Vinci premarket approval. Nancy Dubler, a bioethicist on the panel, said, “I fear we need more data on risk, which would be related in my mind to the dimension of the training that will be required. People who are not sufficiently trained could, in fact, pose a risk using this procedure.”11

But another panelist, Dr. Cedric Walker, countered that he was “very uncomfortable with the notion of an FDA regulatory mandate for a certain number of hours or a certain type of training.” He did not want to include training as a condition for approval: “The reason is that if there are patient injuries from untrained surgeons using this device, then the sponsor is going to be the deep-pocket co-defendant sitting in court with the sur-geon. I think that probably is going to have the effect of ensuring adequate training before the device is sold, probably a better program of adequate training than we, sitting here as non-training experts today, could possibly come up with.”11

Yet in all of the cases we examined, including those reported here, Intuitive Surgical has never been held legally liable for operators’ insufficient training. Dr. Barry Gardiner, principal investigator for the company’s Mexico City premarket trials, told the FDA panel, “It is pretty clear to me that you are safe and effective sitting down and operating with it initially, but there is no question you are going to get better as you go—I don’t know, 10 or 15 cases maybe.”11

The FDA ultimately approved the da Vinci system for use on the condition that Intuitive Surgical would provide a “detailed description” of the training it en-dorsed. We have been unable to locate such a document or what the FDA thought of it. However, we did find that the company provides two days of training on pigs and cadavers to staff members of hospitals that purchase the system. The hospital can then schedule cases on real patients using the da Vinci that are proctored by a surgeon who has performed at least 20 robotically as-sisted procedures.

If other experienced clinicians are to be believed, Dr. Gardiner appears to have seriously underestimated the difficulty of becoming competent in use of the system. Dr. Jim Hu, a surgeon at Brigham and Women’s Hospital in Boston who has performed more than 1,000 roboti-cally assisted procedures, told The Wall Street Journal that it takes a urologist anywhere from 250 to 700 cases to master the da Vinci robot.17 Dr. Ashutosh Tewari from Weill Cornell Medical College in New York, a urologist who has more than 3,000 robotically assisted operations under his belt, told The New York Times that it takes 200 to 300 operations to become highly proficient in performing these procedures.10

Wentworth-Douglass Hospital, where Sherry Long un-derwent surgery, is a small regional facility where the da Vinci robot had only been used in about 300 operations over a four-year period—an average of less than 80 per year.17 This is far too low a number to ensure sufficient expertise, according to surgeons highly experienced with robotic surgery.

Insufficient training in robotic surgery clearly increases the risk of harm to patients.9 Like any technology, the learning curve will vary among individuals: some sur-geons will pick it up quickly while others will never feel comfortable with it. A federal regulatory body should thus be established to certify that a surgeon is qualified, not based on the number of procedures performed but

Insufficient training in robotic surgery clearly increases the risk of harm to patients.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






COVER FE ATURE

62 COMPUTER

on an established level of skill. Anything less is ethically indefensible.

UNFAIR ALLOCATION OF HEALTHCARE RESOURCES

All people have the right to life, and many believe that this right extends to the quality of the healthcare they re-ceive. Of course, not all societies put equal value on access to universal healthcare, which varies greatly among coun-tries. In the US, for example, hospitals are generally private, with patients and insurance companies providing the bulk of payments. Other countries use a single-payer or two-tier system with the government taking a more active role.

Whatever system is implemented, cost of treatment is a central issue. Most innovative surgical systems are ex-tremely expensive. For example, beyond the capital and maintenance expenditures for the da Vinci system, the cost of disposable parts can add up to $2,000 per procedure.17

At the FDA panel on the da Vinci robot, Nancy Dubler said that “there is a school of ethical analysis these days that argues the following: that it is, in fact, unethical to approve new technologies that will add to the cost of medicine, given the number of people without health insurance and access to healthcare, unless there is a measurable benefit that proceeds from that technology.”11

Even in countries with mandated universal healthcare, costs limit patient access to medical technologies. Article 35 of the European Union’s Charter of Fundamental Rights grants all citizens “the right of access to preventive health-care and the right to benefit from medical treatment under the conditions established by national laws and practices.” However, in some cases, taxpayer-funded hospitals might not be able to afford to deliver the best treatment.

Numerous da Vinci systems are in use in the UK, but National Health Service (NHS) hospitals do not necessarily provide the technology to patients even if they request it due to its high cost. For example, those seeking CyberKnife radiosurgery have been denied authorization despite evi-dence that it delivers superior surgical outcomes. In April 2011, the UK’s Telegraph reported that although almost £9 million had been spent buying the technology for three hospitals, NHS officials had rejected most referrals, includ-ing those for many cancer patients who saw CyberKnife surgery as their last hope of survival, forcing them to try to raise the funds on their own.23

Such controversies illustrate the difficulty in balancing an individual’s right to the best medical treatment avail-able with a fair distribution of limited resources. Robotic surgery could be seen as an ethical imperative if it had demonstrably superior comparative outcomes, dramati-cally decreased hospital bed occupation, and resulted in considerably less aftercare and fewer complications. None-theless, it would still not be ethical if treating some people deprived others of access or inhibited the development of less expensive, equally beneficial procedures.

LONG-DISTANCE TELESURGERYIn 2001, Dr. Jacques Marescaux, a surgeon in New York

City, performed the first transoceanic surgery by remov-ing the gallbladder of a woman nearly 4,000 miles away in Strasbourg, France. “Operation Lindbergh,” which cost France Telecom $1.5 million in equipment and $150 million in R&D over 2.5 years to achieve the necessary bandwidth and transmission delay,24 heralded the arrival of long-distance telesurgery. The technology made it possible to remotely operate on patients in underdeveloped countries where there are few, if any, highly skilled surgeons.

However, long-distance telesurgery has yet to become an everyday reality, largely due to the prohibitive costs of ensuring fast, secure, and reliable connectivity. However, the technology also presents several ethical challenges.

Protecting the privacy of patient data is one issue. In 1999, the World Medical Association asserted that “regard-less of the telemedicine system under which the physician is operating, the principles of medical ethics which are globally binding upon the medical profession must never be compromised.”25 As Bernard Dickens and Rebecca Cook6 pointed out, we must ensure that there is no elec-tronic eavesdropping or inadvertent copying of electronic communications such as diagnoses. We must also make sure that nonphysician intermediaries who collect and transmit data about patients observe confidentiality, and that patients treated at home by distant practitioners have consented to give family caregivers access to their medi-cal information.

A clear chain of accountability must be established to assign responsibility for errors in surgical operations that cross national borders. Dickens and Cook6 highlighted the potential difficulties with a simple example: if a medical practitioner in jurisdiction A harms a patient, who is a resident of C, in B while conducting robotic surgery using equipment designed in D and manufactured in E, the in-jured patient could commence a claim for negligence under any of the five jurisdictions’ legal systems. This situation can lead to a form of legal gamesmanship, with the plain-tiff and the defendant asserting those jurisdictional rights most beneficial to their interests.

In addition, there must be a way to deal with procedures that are banned in some countries, such as abortion or

Robotic surgery would not be ethical if treating some people deprived others of access or inhibited the development of less expensive, equally beneficial procedures.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 63

fertility enhancement. A surgeon could be breaking the law in her own country by carrying out a long-distance telesurgical procedure in a country without restrictions, or she could be breaking the law in the country where the patient resides.

Another ethical dilemma is the technology’s poten-tial impact on the doctor-patient relationship. Aimee van Wynsberghe and Chris Gastmans7 suggested that this is likely to be diminished or disrupted by distance. They expressed concern about the loss or reduction in human contact and the objectification or dehumanization of patients, although they acknowledged that this is less important in drastic situations such as those found on a battlefield or at a disaster site.

We also need protection mechanisms to prevent hack-ers from interfering with operations and harming patients, either intentionally or unintentionally. Data encryption would make long-distance telesurgery more secure, but there is a tradeoff in the delays associated with robot control.

Given the massive potential benefits of long-distance telesurgery to people in underdeveloped countries, it would be shameful if we could not surmount these ethi-cal obstacles by securing international legal agreements on establishing jurisdictional liability and developing pa-tient confidentiality guidelines. Nevertheless, telesurgery is unlikely to become a mainstream technology until the cost of dedicated high-speed communication decreases considerably.

R obotic surgery promises unparalleled innovation that minimizes pain and decreases the time spent in recovery. Although this relatively new technology

has been steadily increasing in popularity among both surgeons and patients, that could end quite suddenly if public perceptions change. Thus far, the advertising cam-paigns have gone very well, but grumblings against the technology are growing louder, and medical malpractice suits are mounting. If patients turn against surgical robots, the market could shrink, driving up costs and retarding development.

One of the strongest reasons the FDA panel gave for approving the da Vinci system was its future potential,11

and we agree. But stakeholders must address various ethi-cal issues surrounding robotic surgery if this treatment option is to thrive. While these issues are similar to those arising from surgical innovation in general, a major differ-ence is the increased risk of disrupting the doctor-patient relationship, particularly through the use of long-distance telesurgery and other advances in automation that dis-tance surgeons from those in their care.

An ethical framework for robotic surgery is needed that guarantees patient autonomy, dignity, well-being, and privacy, as well as the fair allocation of resources. Market-

ing claims of clinical superiority by device designers and manufacturers must be well-founded and based on evi-dence, with the costs as well as benefits clearly delineated. Hospitals and surgeons must be honest about possible negative outcomes, and offer safer or less expensive equiv-alent procedures if available. They must also be wary of conflicts of interest to avoid compromising informed pa-tient consent. Finally, there must be more effective federal regulation of training to limit patient harm.

References1. P. Garcia et al., “Trauma Pod: A Semi-automated Telerobotic

Surgical System,” Int’l J. Medical Robotics and Computer As-sisted Surgery, vol. 5, no. 2, 2009, pp. 136-146.

2. N. Sharkey, “Don’t Dismiss Robot Surgeons,” The Guardian,26 Aug. 2008.

3. N.E. Sharkey and A.J.C. Sharkey, “Robot Surgery and Ethical Challenges,” Medical Robotics: Minimally Inva-sive Surgery, P. Gomes, ed., Woodhead Publishing, 2012, pp. 276-291.

4. R.J. Weber, “Surgery Robotics,” Deliverable D5: Techno-Ethical Case-Studies in Robotics, Bionics, and Related AI Agent Technologies, R. Capurro, G. Tamburrini, and J. Weber, eds., Ethicbots Consortium, 2008, pp. 65-83; http://ethicbots.na.infn.it/restricted/doc/D5.pdf.

5. E. Datteri and G. Tamburrini, “Ethical Reflections on Health Care Robotics,” Ethics and Robotics, R. Capurro and M. Nagenborg, eds., IOS Press, 2009, pp. 35-48.

6. B.M. Dickens and R.J. Cook, “Legal and Ethical Issues in Telemedicine and Robotics,” Int’l J. Gynaecology and Ob-stetrics, vol. 94, no. 1, 2006, pp. 73-78.

7. A. van Wynsberghe and C. Gastmans, “Telesurgery: An Ethical Appraisal,” J. Medical Ethics, vol. 34, no. 10, 2008, e22; doi:10.1136/jme.2007.023952.

8. F.D. Moore, “Ethical Problems Special to Surgery: Sur-gical Teaching, Surgical Innovation, and the Surgeon in Managed Care,” Archives of Surgery, vol. 135, no. 1, 2000, pp. 14-16.

9. J. Johnson and W. Rogers, “Innovative Surgery: The Ethical Challenges,” J. Medical Ethics, vol. 38, no. 1, 2012, pp. 9-12.

10. G. Kolata, “Results Unproven, Robotic Surgery Wins Con-verts,” The New York Times, 13 Feb. 2010, p. A1.

11. Dept. of Health and Human Services Food and Drug Ad-ministration Center for Devices and Radiological Health, Summary minutes of the open session of the General and Plastic Surgery Devices Panel, Medical Devices Advisory Committee, 16 June 1999; www.fda.gov/ohrms/dockets/ac/99/transcpt/3523t1.rtf.

12. J.C. Hu et al., “Comparative Effectiveness of Minimally Invasive vs Open Radical Prostatectomy,” JAMA, vol. 302, no. 14, 2009, pp. 1557-1564.

13. D.A. Barocas et al., “Robotic Assisted Laparoscopic Pros-tatectomy versus Radical Retropubic Prostatectomy for Clinically Localized Prostate Cancer: Comparison of Short-Term Biochemical Recurrence-Free Survival,” J. Urology,vol. 183, no. 3, 2010, pp. 990-996.

14. L.X. Jin et al., “Robotic Surgery Claims on United States Hospital Websites,” J. Healthcare Quality, vol. 33, no. 6, 2011, pp. 48-52.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_______________

http://www.qmags.com/clickthrough.asp?url=www.fda.gov/ohrms/dockets/ac/99/transcpt/3523t1.rtf&id=17866&adid=P63E2

http://www.qmags.com/clickthrough.asp?url=http://ethicbots.na.infn.it/restricted/doc/D5.pdf&id=17866&adid=P63E1





COVER FE ATURE

64 COMPUTER

15. D.A. Ollendorf et al., “Active Surveillance & Radical Prostatectomy for the Management of Low-Risk, Clinically-Localized Prostate Cancer,” Inst. for Clinical and Economic Review, draft appraisal document, 16 July 2009; www.icer-review.org/index.php/ASRP/View-category.html.

16. L. Spears, acting director of Office of Compliance, Dept. of Health and Human Services Food and Drug Admin-istration Center for Devices and Radiological Health, warning letter to L.M. Smith, president and CEO, Intuitive Surgical, 12 Apr. 2001; www.fda.gov/downloads/ICECI/EnforcementActions/WarningLetters/2001/UCM069548.pdf.

17. J. Carreyrou, “Surgical Robot Examined in Injuries,” The Wall Street J., 4 May 2010, p. A1.

18. T. Tsouderos, “Remote Control Surgery Grows, Despite Inconclusive Evidence,” Chicago Tribune News, 25 Feb. 2012; http://articles.chicagotribune.com/2012-02-25/news/ct-met-robot-20120225_1_robot-assisted-intuitive-surgical-prostate-removal-surgeries.

19. W.L. Bargar, A. Bauer, and M. Börner, “Primary and Revi-sion Total Hip Replacement Using the Robodoc System,” Clinical Orthopaedics and Related Research, vol. 354, Sept. 1998, pp. 82-91.

20. ABC7 Chicago News, “Family of Man Who Died after Sur-

gery Awarded $7.5M,” 17 Feb. 2012; http://abclocal.go.com/wls/story?section=news/local&id=8548618.

21. A.D. Krauss, “Da Vinci Patient Suit Expected against WDH: Woman Claims Permanent Damage When Both Ure-ters Severed during Surgery,” Fosters.com, 22 May 2010; www.fosters.com/apps/pbcs.dll/article?AID=/20100522/GJNEWS_01/705229896/-1/fosnews.

22. D. Suchetka, “Cancer Patient Sues Cleveland Clinic, Claim-ing Doctors-in-Training Performed Prostate Surgery That Left Him Incontinent and Impotent,” Cleveland.com, 7 Jun. 2010; http://blog.cleveland.com/metro/2010/06/cancer_patient_sues_cleveland.html.

23. L. Donnelly, “Cancer Patients Denied ‘Last-Hope’ Robot Surgery,” The Telegraph, 3 Apr. 2011; www.telegraph.co.uk/health/healthnews/8423891/Cancer-patients-denied-last-hope-robot-surgery.html.

24. Espace Multimedia, “Operation Lindbergh”—A World First in TeleSurgery: The Surgical Act Crosses the Atlantic,” press conf., 19 Sept. 2001; www.ircad.fr/event/lindbergh/lindbergh_presse_en.pdf.

25. World Medical Assoc., “WMA Statement on Accountability, Responsibilities and Ethical Guidelines in the Practice of Telemedicine,” Oct. 1999; www.wma.net/en/30publications/10policies/20archives/a7.

Noel Sharkey is a professor of AI and robotics as well as a professor of public engagement at the University of Sheffield, UK. His research interests include the ethical application of ro-botics in areas such as the military, policing, crime, child care, elder care, transportation, and sex and medicine. He received a PhD in ex-perimental psychology/cognitive science from the University of Exeter, UK, and a DSc from the University of Ulster, UK. He is a Fellow of the Institution of Engineering and Technol-ogy, the British Computer Society, the Royal Institute of Navigation, and the Royal Society of Arts and is both a chartered electrical engi-neer and a chartered information technology professional. Contact him at [email protected].

Amanda Sharkey is an associate professor (senior lecturer) in the Department of Com-puter Science, University of Sheffield, UK. Her research interests include the ethics of robot care, human-robot interaction, and swarm robotics. She received a PhD in psycholinguis-tics from the University of Essex, UK. She is a member of the Institution of Engineering and Technology and the Experimental Psychology Society. Contact her at [email protected].


IEEE offers a variety of open access (OA) publications:

Discover top-quality articles, chosen by the IEEE peer-review standard of excellence.

Unrestricted access to today’s groundbreaking research

via the IEEE Xplore® digital library

Learn more about IEEE Open Access

www.ieee.org/open-access



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___

_____________

____________________________

______

_________________

__________________________________

__________________________________

__

_______________

_______________

__________________________________

_______________

_________________

_____________________

__________________________

____

____________________

__________

_______


http://www.qmags.com/clickthrough.asp?url=www.wma.net/en/30publications/10policies/20archives/a7&id=17866&adid=P64E11

http://www.qmags.com/clickthrough.asp?url=www.ircad.fr/event/lindbergh/lindbergh_presse_en.pdf&id=17866&adid=P64E10

http://www.qmags.com/clickthrough.asp?url=www.telegraph.co.uk/health/healthnews/8423891/Cancer-patients-denied-last-hope-robot-surgery.html&id=17866&adid=P64E9

http://www.qmags.com/clickthrough.asp?url=http://blog.cleveland.com/metro/2010/06/cancer_patient_sues_cleveland.html&id=17866&adid=P64E8

http://www.qmags.com/clickthrough.asp?url=www.Cleveland.com&id=17866&adid=P64E7

http://www.qmags.com/clickthrough.asp?url=www.fosters.com/apps/pbcs.dll/article?AID=/20100522/GJNEWS_01/705229896/-1/fosnews&id=17866&adid=P64E6

http://www.qmags.com/clickthrough.asp?url=www.Fosters.com&id=17866&adid=P64E5

http://www.qmags.com/clickthrough.asp?url=http://abclocal.go.com/wls/story?section=news/local%26id=8548618&id=17866&adid=P64E4

http://www.qmags.com/clickthrough.asp?url=http://articles.chicagotribune.com/2012-02-25/news/ct-met-robot-20120225_1_robot-assisted-intuitive-surgical-prostate-removal-surgeries&id=17866&adid=P64E3

http://www.qmags.com/clickthrough.asp?url=www.fda.gov/downloads/ICECI/EnforcementActions/WarningLetters/2001/UCM069548.pdf&id=17866&adid=P64E2

http://www.qmags.com/clickthrough.asp?url=www.ieee.org/open-access&id=17866&adid=P64A1

http://www.qmags.com/clickthrough.asp?url=www.icer-review.org/index.php/ASRP/View-category.html&id=17866&adid=P64E1







PERSPECTIVES


The ability to recover quickly from large-impact, hard-to-predict, and rare incidents without incurring sizable permanent damage is vital to major stakeholders in ICT infrastructures of national importance.

E very industrialized nation has information systems that provide services of great importance to its popu-lation. Examples include payment, mobile phone, identity management, and electronic voting systems.

Nationwide information systems tend to be stable over long periods, punctuated by large-impact, hard-to-predict, and rare (LHR) incidents.

Researchers traditionally measure the risk associated with a future incident by estimating its likelihood and impact. While the likelihood of a recurrent incident is esti-mated by considering incidents in the past, it is very hard to accurately estimate the likelihood of a rare incident because the past might not contain any such event. We say that the risk of an event is hidden when its likelihood is unknown and the impact is difficult to estimate.

In a holistic approach to the assessment of hidden risks in ICT infrastructures of national importance, an infra-structure is modeled as a complex adaptive system1,2

consisting of many interdependent parts, and LHR events are modeled as surprising and extreme global behavior caused by the numerous interactions between these parts. Three system properties—being unique, connected, and closed—identify infrastructures with large hidden risks.

Here, Norwegian infrastructures and LHR events during 2011 are presented as examples. These events, as well as those described in an earlier publication on risk assessment,3 suggest that LHR events dominate major stakeholders’ total risk. Mitigation of LHR events is thus

Kjell Jørgen Hole, University of Bergen, Norway

emphasized, especially incidents caused by functional dependencies between infrastructures.

The risk management approach presented here is based on Nassim Taleb’s seminal work,4 Dan Geer’s Source Boston 2008 and 2012 keynotes,5,6 my own research,3,7,8 and pub-licly available reports from Norwegian system owners and government agencies.

COMPLEX SYSTEMS AND RARE EVENTSA stakeholder is a person or institution with an inter-

est in a system. As Figure 1 shows, a complex adaptive information system consists of stakeholders, a networked computer system, security and privacy policies, and threats such as equipment failure, extreme weather, hacking, sabotage, and terrorism. The system’s complexity is due to the numerous interactions between the users and the networked system, the large amounts of communications between the networked subsystems, and the influence of changing policies and threats. Because a complex system operates in an ever-changing environment, its global behav-ior changes over time. The system’s high complexity and lack of well-defined boundaries make it nearly impossible to predict extreme global behavior from the local opera-tions of the many parts.1,2

A black swan in a complex system is a metaphor for a large-impact, rare event that comes as a complete surprise to all stakeholders.3,4 This type of event is the “unknown unknown,” a rare bombshell that no stakeholder has con-

Management of Hidden Risks



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






66 COMPUTER

PERSPECTIVES

sidered. The catastrophic consequences of 11 September 2001 (9/11) are negative black swans. A gray swan is a metaphor for a large-impact, rare event that is somewhat predictable, yet most stakeholders overlook it. It is the “known unknown,” a rare event that some know is possible, but no one knows when or whether it will occur. Because a gray swan is not a complete surprise to all stakeholders, it tends to have less impact than a black swan. However,

its impact is still huge. Figure 2 illustrates the differences in likelihood and impact between a nonrecurrent black or gray swan and “normal,” recurrent incidents: swans are outliers.

While black swans cannot be assessed, we can partly assess gray swans. Because the global behaviors of complex adaptive systems change significantly over time, using fixed probability distributions to rep-resent the probabilities of rare incidents is problematic. Even if we could, it would be hard to determine the tail of a distribution where the negative gray swans reside since we have little or no relevant data about non-recurrent events. All in all, it is not possible to determine good point estimates for the probabilities of gray swans. (Readers who are skeptical about this claim should study Taleb’s technical papers at www.fooledbyrandomness.com.) It is, however, possible to determine wide interval estimates for these probabilities.9

It is natural for a system owner or a governmental regulatory agency to view an incident’s impact as a function of the number of affected users. National infra-structures generally have impact functions with superlinear growth. For example, the sketched impact function in Figure 3 grows

linearly for a few affected users, but then grows faster for an increasing number of users due to factors such as class action lawsuits or new, more stringent requirements from governmental agencies.

Impact functions with superlinear growth allow for outli-ers. A hard-to-predict incident is a gray swan if the impact function’s value is well into the superlinear region in Figure 3. It is difficult to determine highly accurate impact functions for rare future incidents. However, if it can be argued that an event’s impact is a superlinear function of the event’s size, then the system is likely vulnerable to gray swans.

ASSESSMENT AND MITIGATION OF GRAY SWANS

Norwegian ICT infrastructures are complex adaptive sys-tems because they contain millions of interacting devices, have millions of users, are regulated by either national or international laws, and are exposed to many threats. Major stakeholders are solution architects, software developers, security experts, owners, operators, users, and governmen-tal regulatory agencies. Rare and extreme global behaviors represent negative gray swans. The following three system properties help to identify infrastructures vulnerable to gray swans:

Figure 1. A complex information system’s global behavior is caused by dynamic interactions among stakeholders and the networked computer system, as well as by interactions among the network’s many subsystems. Changes to policies, threats, or subsystems can cause sudden large changes in global behavior.

Figure 2. Likelihood and impact of a single nonrecurrent swan incident versus recurrent incidents. The swan is an outlier.

Figure 3. Superlinear impact function. The sketched impact function grows linearly for a few affected users, but then grows faster for an increasing number of users due to varying factors.

Stakeholders

Interactions

Networked system

Polic

ies

Thre

ats

Impact

Likeli

hood

Swan

Affected users

Impa

ct

Linear regionSuperlinear

region



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___________

http://www.qmags.com/clickthrough.asp?url=www.fooledbyrandomness.com&id=17866&adid=P66E1





JANUARY 2013 67

A system is unique when some alternative sequence of operations cannot replicate its key functions in another system. A unique infrastructure is likely vul-nerable to gray swans because the many users cannot switch to an alternative infrastructure during pro-longed downtime.A system is connected when its normal operation depends on the normal operation of another system. If an infrastructure is connected to another infra-structure, then the large overall complexity of the infrastructures is likely to cause gray swans.A system is closed when stakeholders refuse to share technical and legal information. If only a small group of experts have deep knowledge of an infrastructure, they have a tendency to develop similar mental models for how the infrastructure works during discussions. This propensity toward “groupthink” is especially strong when most group members belong to the same organizational culture. A uniform group cut off from external expertise overlooks gray swans.

A national infrastructure is usually characterized by some of these properties. Two known incidents illustrate why these properties indicate vulnerability to gray swans and offer suggestions for how to limit the impact of future swans.

Mobile phone systemsMobile phone networks are essentially unique systems

because users of one network cannot generally connect to another. If an entire network goes down, millions of mobile phones become useless.

After a long period with stable service, the largest mobile phone network in Norway went down for about 11 hours on 10 June 2011. The restart of a central node with upgraded software initiated a “signal storm” that exceeded the net-work’s signaling capacity. The outage affected nearly 3 million customers—approximately 60 percent of all Nor-wegians. According to the system’s chief executive officer, “this was not supposed to happen” because earlier restarts of the same node had not caused any problems. The man-agement’s surprise and the rare, highly negative outcome of a common operation qualify the signal storm as a gray swan incident.

On 17 June 2011, parts of the same mobile phone net-work went down again due to a new signal storm. According to the Norwegian Post and Telecommunications Authority (NPTA), both signal storms were caused by insufficiently understood interdependencies between central nodes in the network combined with too little capacity to handle increasing signal traffic from many new smartphones.

The difficulty in pinpointing the true root cause of the extreme behavior became evident when, more than a month later, careful technical analyses of the events finally revealed that the signal storms were primarily due to a

programming error, and not insufficient signal capacity to serve new smartphones, as the NPTA suggested. However, the network owner also discussed the need to change the system design and to increase the signal traffic capacity.

The network owner’s difficulty in determining the root cause illustrates that the mobile phone network is indeed a complex adaptive system with surprising global behavior. The NPTA stated publicly that the owner needed to improve the network’s risk management.

While better risk management can assess and mitigate more incidents, perhaps providing longer periods of stable

network operation, large-impact incidents will still occur because the network has too many dynamic interactions for humans to reliably foresee rare and extreme behavior.

All mobile phone networks in Norway are connected to the national power grid. The normal operation of each net-work depends on a nearly continuous supply of electricity. These networks went down in a large area of Norway when a storm with hurricane-force winds damaged many power lines in late December 2011, leaving more than 700 base stations belonging to the different networks without elec-trical power. While the base stations had backup batteries, most lasted a maximum of only four hours. Because land-line phone and fixed Internet access were also disrupted in the same area, most people were without communication capabilities following the storm. Although both the power and phone companies worked hard to repair the extensive damage, it took more than a week to restore services to all customers.

The networks’ strong dependence on the power grid came as a surprise to leading Norwegian politicians. Their initial response was to severely criticize the mobile phone companies. According to the NPTA’s director general, the people of Norway had come to depend more on the mobile phone systems than the agency had realized before the storm.

To mitigate the impact of future incidents, the NPTA required the network owners to improve their backup power solutions to ensure that the 1,000 base stations cov-ering the most critical areas of Norway have backup power for at least 72 hours. All other base stations were required to have at least six hours of backup power. In addition, the owners were told to prepare more resources and develop

A unique infrastructure is likely vulnerable to gray swans because the many users cannot switch to an alternative infrastructure during prolonged downtime.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






68 COMPUTER

PERSPECTIVES

better contingency plans to enable local crews to quickly repair damaged base stations.

Electronic voting systemsNorway’s governmental electronic voting (e-voting)

system is built on fiberoptic, ADSL, and mobile phone net-works.8 All voters run a Java applet in their Web browsers to submit ballots over the Internet to a central infrastruc-ture. Because there is no alternative to this remote e-voting system, there is a real possibility that voters might not be able to cast a vote due to failure of the central infrastructure or one of the underlying networks.

The e-voting system lets voters detect malware attacking the applet on their computers. Before the e-voting period starts, each voter is given a card with one verification code per political party. When casting a ballot, the voter receives a verification code for the chosen party via SMS. If the received

code does not match the code on the card, it is likely that malware has tampered with the ballot.8 However, it is still possible for malware to obtain the identity of voters and the content of their ballots without being detected. Such privacy violations and information leakages are a major concern.

The impact functions associated with prolonged down-time and information leakage are, most likely, superlinear functions in the number of affected voters. While very few affected voters might be tolerable since problems also occur with paper-based voting, even a relatively modest number of affected voters will generate massive press cov-erage, leading to widespread distrust of the e-voting system among the population. Hence, the e-voting system by itself is vulnerable to surprise incidents that can quickly scale to become gray swans. There are also other hidden risks due to the system’s source code being available on the Internet as well as remote e-voting’s inherent vulnerability to large-scale vote buying and voter coercion.8

While the Norwegian e-voting system has only seen limited use so far, the responsible government ministry has described two countermeasures to mitigate surprise incidents leading to prolonged downtime and information leakage during a national election. First, the e-voting system will be operational for multiple weeks, allowing voters time to cast their ballots later if the system goes down temporar-ily. Second, because a widespread information leakage is intolerable, a “traditional” election day will be established to allow all voters to cast paper ballots after the e-voting period has ended.

A cast paper ballot will nullify any earlier electronic ballot. In other words, e-voting does not replace paper-based voting; it only provides an additional means of voting. It is unknown whether the two countermeasures are suf-ficient to maintain enough trust in the e-voting system after a serious system failure or malware attack.

This discussion illustrates that a unique infrastructure should not realize a service of critical importance when the impact of a gray swan is intolerable. Hidden risks can be reduced by realizing a critically important service using two different systems.3 It is vital that a failure in one system does not cause a failure in the other. It is not sufficient to deploy two identical systems because they obviously have common vulnerabilities; the systems must have diverse architectures, designs, and implementations. Of course, simultaneous attacks could still bring down both systems, but the diversity would make it very costly and difficult to successfully carry out such attacks.

PUBLIC-KEY INFRASTRUCTUREA closed infrastructure can cause hidden risks. The public-

key infrastructure providing most Norwegian Internet banks with authentication and digital signature services is a closed system. There is a need for information shar-ing to mitigate gray swans associated with this system’s digital signatures.7 The bank customers’ cryptographic keys are stored on a central server. To authenticate, a cus-tomer enters a username, a one-time personal identification number, and a fixed password into a Java applet running on a Web browser. The applet then accesses cryptographic functionality on the central server to digitally sign a docu-ment with the customer’s private key.

The results of a study conducted in 2007 while working with five PhD students revealed that the unusual central key storage and a problem with the Java applet’s implementation made the authentication procedure especially vulnerable to combined phishing/man-in-the-middle attacks. Following the responsible disclosure model, we informed the relevant enti-ties about our finding, but little happened until the Financial Supervisory Authority of Norway was given a proof-of-concept demonstration. A later public disclosure of several problems with the infrastructure created a fierce debate, leading a Nor-wegian parliamentarian to bring up our security concerns during one of Parliament’s question sessions.

Following the 2009 publication of our risk analysis of the public-key infrastructure and its digital signature service,7 the NPTA requested an independent security analysis. While the details of this analysis have not been published, a publicly available summary written in Nor-wegian confirms that the signature service is vulnerable to client-side malware attacks (www.bankid.no/documents/dokumenter%20for%20alle/BSK-NO-5100770-ETR_Sammendrag_v01_0.pdf), but the likelihood of a success-ful attack is deemed small. In 2011, real-world malware

E-voting does not replace paper-based voting; it only provides an additional means of voting.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


______________________________________

________________

http://www.qmags.com/clickthrough.asp?url=www.bankid.no/documents/dokumenter%20for%20alle/BSK-NO-5100770-ETR_Sammendrag_v01_0.pdf&id=17866&adid=P68E1





JANUARY 2013 69

bypassed the user authentication, and funds were fraudu-lently withdrawn from a few bank accounts. While there might be only a small likelihood that a similar attack could sign an electronic document on an unsuspecting user’s behalf, a successful attack is possible.

We cannot determine the probability of a successful attack on the signature service, but we can carry out a thought experiment to show that the impact can be severe.

According to the bank in our fictitious experiment, a customer digitally signed an electronic contract. Because the customer claims not to have signed the contract, the bank must convince a judge that the customer did in fact do so. As part owner of the public-key infrastructure, the bank has access to experts with detailed system knowledge and information about the signature action. Because the customer’s lawyer does not have access to the same level of expertise or information, the customer could lose the case even though malware signed the disputed contract.

Let us consider the impact function from the bank’s point of view. It is reasonable to assume that the function is superlinear in the number of affected customers since even a rather limited attack impacting few customers could lead to widespread mistrust of both the public-key infra-structure and the involved banks. A situation in which an attack scales and customers must go to court to defend themselves against malware-generated signatures quali-fies as a gray swan.

It is possible to defend against malware attacks using a callback procedure similar to the one deployed by the Norwegian e-voting system. In an improved version of the infrastructure, a customer receives a confirmatory message on a separate channel—for example, a mobile phone—when digitally signing a contract. The signature is only legally binding after the customer has acknowledged the message. Further, an independent and trusted third party collects and stores information about the signature process—so-called nonrepudiation information.7,10 During a conflict between a customer and a bank, the trusted third party is obliged to provide the customer’s lawyer with the same level of technical expertise and nonrepudiation infor-mation available to the bank.

To limit the impact of unplanned downtime associated with the infrastructure’s authentication service, some Inter-net banks can activate alternative authentication services. When the infrastructure went down unexpectedly on 6 July 2011, customers of Internet banks without authentication fail-safes could not access their accounts for nearly six hours, while customers of banks with alternative authen-tication services only experienced a minor inconvenience because the authentication procedure had changed.

ELECTRONIC PAYMENT INFRASTRUCTURESElectronic payment infrastructures in Norway are

connected systems. On 20 April 2011, automatic teller

machines (ATMs) and in-store payment terminals stopped working when a primary payment server malfunctioned and the secondary backup server could not handle the increased level of payment transactions associated with Easter shopping.

According to the Financial Supervisory Authority of Norway, 140,000 customers tried in vain to pay with their debit cards or get cash from an ATM. Each payment transaction was supposed to be processed by a system con-taining the malfunctioning server, followed by additional processing in a second system. Corrections made by the operators of the first system and delays created by the sec-ondary server caused the two systems to accept multiple

payment attempts for a single item, leading customers to incorrectly believe that their accounts were empty. While the first system was restored after seven hours, it took a week before all account balances were corrected.

This incident is a gray swan because the team correct-ing the problems in the first system did not appreciate the large negative effect these changes had on the processing of payment transactions in the second system. While a more diverse risk management team3 with knowledge of both systems could perhaps have avoided the incident, the connected payment infrastructures’ high overall complex-ity makes it impossible to completely avoid gray swans.

A more realistic goal is to reduce the impact of gray swans and improve the ability to handle swans when they occur. As an example, the Finance Sector Union of Norway wants to scrap paper money and use only electronic pay-ment systems. The connected payment infrastructures are known to fail on rare occasions, thus there is a need for an alternative payment system that is not connected to exist-ing infrastructures before cash is abolished.

Since the Norwegian mobile phone networks are made more robust to power interruptions, it is possible to build a new payment system using the mobile networks and a central infrastructure that is separate from existing pay-ment infrastructures. A careful avoidance of functional dependencies between the new and old payment systems could make it highly unlikely that gray swans occur in these systems at the same time. Consequently, consumers

To limit the impact of un-planned downtime associated with the infrastructure’s authentication service, some Internet banks can activate alternative authentication services.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






PERSPECTIVES

70 COMPUTER

W hile here I have only considered hidden risks in national infrastructures, the potential for surprising incidents with huge impact is even greater in worldwide systems. The major

outages suffered by Google App Engine in July 2009 and Amazon Elastic Compute Cloud in April 2011 show that even very large and resourceful companies cannot entirely avoid gray-swan incidents. The ability to recover quickly from large-impact incidents without sizable permanent damage is thus vital to major stakeholders.

References1. J.P. Crutchfield, “The Hidden Fragility of Complex

Systems—Consequences of Change, Changing Conse-quences,” Cultures of Change: Social Atoms and Electronic Lives, G. Ascione, C. Massip, and J. Perello, eds., ACTARD Publishers, 2009, pp. 98-111.

2. J.H. Miller and S.E. Page, Complex Adaptive Systems: An Introduction to Computational Models of Social Life,Princeton Univ. Press, 2007.

3. K.J. Hole and L.-H. Netland, “Toward Risk Assessment of Large-Impact and Rare Events,” IEEE Security & Privacy,vol. 8, no. 3, 2010, pp. 21-27.

4. N.N. Taleb, The Black Swan: The Impact of the Highly Improbable, 2nd ed., Random House, 2010.

5. D.E. Geer, “Dan Geer Keynote,” Source 2008 Conf., 13 Mar. 2008; http://geer.tinho.net/geer.sourceboston.txt.

6. D.E. Geer, “Dan Geer Keynote,” Source 2012 Conf., 18 Apr. 2012; http://geer.tinho.net/geer.sourceboston.18iv12.txt.

7. K.J. Hole et al., “Risk Assessment of a National Security Infrastructure,” IEEE Security & Privacy, vol. 7, no. 1, 2009, pp. 34-41.

8. L.H. Nestås and K.J. Hole, “Building and Maintaining Trust in Internet Voting,” Computer, May 2012, pp. 74-80.

9. B.J. Garrick, Catastrophic Risk, Academic Press, 2008.

10. J. Zhou, Non-Repudiation in Electronic Commerce, Artech House, 2001.

11. T. Aven, Quantitative Risk Assessment: The Scientific Platform, Cambridge Univ. Press, 2011.

Kjell Jørgen Hole is a professor in the Depart-ment of Informatics, University of Bergen, Norway. His research interests include risk management of complex adaptive systems. Hole received a PhD in computer science from the University of Bergen. He is a member of IEEE. Contact him at [email protected].

would have at least one payment system they could use at any time.

To further ensure the availability of the payment sys-tems, it is important that the owners prepare resources and arrange regular training exercises to allow a quick return to normal operation after a gray-swan incident.

BETTER MANAGEMENT OF HIDDEN RISKSIt is tempting for decision-makers to ignore possible gray

swans in ICT infrastructures because identifying swans is difficult, the incidents are unlikely, and implementing countermeasures often is expensive. However, insufficient focus on swans during an infrastructure’s life cycle can lead to huge negative consequences affecting a significant per-centage of a nation’s population. It is particularly important to mitigate gray swans due to connected infrastructures.

To evaluate the functional dependency between two infrastructures, we can determine scenarios for each system with a potential to seriously impact the normal operation of the other system. Such scenarios are likely to exist when the systems are colocated, contain identical software or hardware, utilize the same networks, exchange much signaling traffic, or process each other’s data. It is important to consider previous incidents and to remember that a small-impact incident in one system can trigger a large-impact incident in the other system.

It is hypothesized that gray swans dominate major stake-holders’ total risk.3 The choice of risk assessment method3,11

is important since “traditional” methods underestimate the risk of possible rare and extreme incidents, especially incidents caused by functional dependencies between infrastructures. If owners cannot determine functional dependencies between systems, they can still secure the systems over the short term, if not over the long term.6

A stakeholder’s trust in an infrastructure will continue to grow due to stable operation, until a gray swan occurs with, perhaps, intolerable impact, resulting in the loss of trust, if not generating a powerful distrust, that might be difficult to overcome.




qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________

__________________

_________________________


http://www.qmags.com/clickthrough.asp?url=http://geer.tinho.net/geer.sourceboston.18iv12.txt&id=17866&adid=P70E2

http://www.qmags.com/clickthrough.asp?url=http://geer.tinho.net/geer.sourceboston.txt&id=17866&adid=P70E1






http://www.qmags.com/clickthrough.asp?url=www.computer.org/intelligent&id=17866&adid=P70A1

RESE ARCH FE ATURE


Reminders to take medications. Various message ser-vices and over-the-counter devices are available to help patients remember to take medications at cer-tain times. Some devices also regulate the dispensing of medicines to ensure that patients take the correct doses.Improved scheduling. If a patient is routinely delaying or missing doses, the healthcare provider can revise the regimen by prescribing medications with a longer half-life to allow for fewer doses per day or by schedul-ing stronger doses fewer times per day depending on the patient’s tolerance.

I am exploring ways to use wireless technologies to im-prove medication adherence by patients who are not in a specialized care center such as a nursing home or assisted living facility, where medications are professionally man-aged, but are caring for themselves at home. Specifically, I have been working to develop a smart medication manage-ment system (SMMS) that is flexible, portable, and integrates well with smart environments.

SMART MEDICATION MANAGEMENT SYSTEMS

To help patients—particularly those with cognitive deficiencies, including the elderly—remember which medications to take, in what doses, and when, medication management systems issue reminders using some combi-nation of alarms and messages. Some systems also monitor

M edication adherence is critical to managing chronic conditions such as cardiovascular dis-ease and diabetes. However, patients use only 50-60 percent of medications as prescribed,1

leading to 125,000 unnecessary deaths and $90 billion in additional medical procedure and hospitalization costs every year in the US alone.2

Patients fail to adhere to drug prescriptions for many reasons, including forgetfulness (30 percent), having other priorities (16 percent), choosing not to do so (11 percent), lacking the proper information (9 percent), and emotional instability (7 percent).3

In general, 80 percent medication adherence is con-sidered satisfactory; however, a higher level is needed to treat some conditions, such as 95 percent for HIV medi-cations. Increasing medication adherence to such levels, both to improve outcomes and to help contain rising costs, constitutes a grand challenge for the healthcare industry.

Interventions to help patients manage prescription drug use generally fall into three categories:1,3-7

Motivation and support. This can come from various caregivers, including family, friends, and healthcare professionals—physicians, nurses, pharmacists, thera-pists, social workers, and so on. It can also come from insurance companies in the form of waived premi-ums or price reductions for medication management devices.

Smart medication management systems can remotely monitor and regulate patients’ prescription drug use, support communication with caregivers to improve scheduling and increase motivation, and generate context-aware reminders.

Upkar Varshney, Georgia State University

A Smart Approach to Medication Management



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






72 COMPUTER

RESE ARCH FE ATURE

medication consumption to ensure that patients adhere to their prescribed regimen.

The simplest systems are automated pill dispensers that beep when it is time to take a medication and also record how many times the containers have been opened and closed. More expensive systems include remote moni-toring to notify patients and caregivers—by phone, text, or email—that a dose has been changed or missed. The highest-end systems support personalized intervention. For example, if a patient misses a dose, the system will notify a licensed practitioner, who if necessary can consult with the patient’s physician and vary the timing and quantity of leftover doses or instruct the patient to skip the dose if variations are not possible and it is medically safe.

Prototype SMMSs use sensors to obtain context-aware information, such as vital signs and other biomedical parameters, to monitor general health as well as prescrip-tion drug use and to support caregiver intervention. For example, the Magic Medicine Cabinet4 incorporates face recognition technology, RFID medicine labels, vital sign monitoring, voice synthesis, and Web-based interaction with healthcare professionals. The Smart Medicine Cabinet6

also uses RFID labels as well as Bluetooth-enabled active tags to communicate with a cell phone. These systems are not portable, however, which limits use by patients with varying social and travel schedules.

As Figure 1 shows, an SMMS could be integrated with other wireless technologies—including social interaction and entertainment, memory support, fall prevention and detection, and daily activity support—in a smart house.8,9

Sensors used for medication management and health moni-

toring could be implanted, wearable, portable, or installed in the patient’s surroundings depending on the level of intrusiveness required or the patient’s mobility.2,10,11 Wear-able sensors in particular must be reliable, robust, durable, unobtrusive, and able to identify users and communicate with other sensors and devices, as well as require minimal maintenance and fault recovery.

SMMS DESIGN AND OPERATIONI have architected a portable SMMS that provides context-

sensitive reminders, dispenses prescription drugs, moni-tors the use of medications and vital signs, and supports intervention by healthcare professionals to improve sched-uling.12 The system can also communicate with family members, friends, and other designated persons to moti-vate providing support for medication adherence.

Basic functionsTo achieve these goals, the SMMS incorporates vari-

ous sensors as well as RFID and Bluetooth technology. As Figure 2 shows, the system links to both wired networks and wireless local area and wide area networks. In addition to issuing alerts and notifications, the system can transmit and receive information—for example, electronic medi-cal records and patient schedules—to increase safety and convenience as well as to encourage medication adherence.

Figure 3 shows some basic SMMS functions. These in-clude obtaining information on medications and doses from healthcare professionals, authorizing people to be notified, determining the duration and type of monitoring, and managing network connectivity.

Figure 1. A smart medication management system (SMMS) could be integrated with other sensing and monitoring technologies in a smart house.

Breathingsensor

Posture-detectionsensors

Swallowing-detection sensors

Temperature sensors

Blood pressuresensor

Oxygen saturationsensor

ECG sensors

Memory support system

Fall prevention and detection systems

Smart medication management system

Energy-aware

Context-aware

Daily activity support system

Social interaction/entertainment system

Aler

ts, no

tifica

tions

, and

rem

inder

s

Real-

time m

onito

ring



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 73

ADDICTION AND OVERDOSE PREVENTION According to the National Institutes of Health, 20 percent

of US residents age 12 and older have used prescription drugs for nonmedical reasons at least once in their life-time.13 Doctors increasingly prescribe drugs to treat various conditions, and many medications—especially opioids,

The system dispenses medications, in prescribed doses, to the patient at certain times as programmed by healthcare professionals. The timing of doses can be ad-justed to accommodate certain situations, such as when the patient is traveling. The SMMS verifies medication use and computes patterns of adherence. It implements interventions that range from context-aware reminders to scheduling revisions based on consultations with health-care providers. It can be programmed to recommend changes in the intervention if adherence falls below a given threshold.

Context-sensitive remindersThe system generates reminders using rule-based pro-

tocols, in which logical connections among a wide range of data inputs weighted for the patient determine the rules. As Figure 4 shows, these include prescribed medication schedules and actual use; medication adherence goals and history; the patient’s health status (changes in vital signs); the patient’s current location and activity (for example, sitting, walking, running, or sleeping); and any special in-structions or data, such as drug interaction, expiration, and recall information.

The protocols are designed to assist, not replace, deci-sion making by caregivers. If the system cannot acquire the context, it will notify the appropriate person, who can request more information and make a better decision.

The SMMS can send context-aware reminders to the patient on multiple devices such as display screens, cell phones, and computers. These can be in the form of au-dible, visual, or vibratory alarms as well as voice and text messages. The reminders can include supplementary in-formation about the nature of the medications as well as guidance on how to take them.

Figure 2. An SMMS can provide context-sensitive reminders, dispense prescription medications, monitor vital signs, and support communication with healthcare professionals and other caregivers to improve scheduling and motivate medication adherence. The system can link to both wired networks and wireless local area and wide area networks.

Figure 3. Basic SMMS functions. The system’s flexibility ensures highly personalized service to the patient.

Smartmedication

managementsystem

Multinetwork access(wired and wireless)

Notification/alertsto designated people

Reminders/alarms

Notificationsand displays

Medication adherence goals

Information on medications

Monitoring ofmedication useand vital signs

Dispensing of medications

Receive information on medications, doses, adherence goals, type of interventions, and so on

Receive information on the type and durationof medication and health monitoring

Determine when and how (voice, video monitor, displayboard, and so on) to generate reminders to patients

Designate people who need to be notifiedif certain undesirable events occur

Communicate to healthcare professionals and wait for directions

Manage and monitor the dispensingof medications at certain times

Is dose taken?Yes

No

Yes

No

Generate context-aware reminder(s)

Is it safe to takedose now?



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






RESE ARCH FE ATURE

74 COMPUTER

In addition, the data provided by an SMMS enables caregivers to review a patient’s prescription drug use before appointments. This is especially valuable in the case of patients who might have trouble remembering or communicating.

BEHAVIORAL MONITORINGMental illnesses, especially depression, are among the

leading causes of disability worldwide. Monitoring and interpreting the symptoms of such illnesses is complex and requires mapping multiple behavioral and physiological parameters.

Obtaining sufficient data to make an accurate diagno-sis is challenging. Physicians consider numerous factors including the patient’s psychiatric and medical history, substance abuse history, and family and personal history as well as the results of medical and psychiatric evalu-ations. Much of this data is subjective and difficult to measure.

Sensors deployed on or near patients in their every-day surroundings would make it possible to remotely acquire a wide range of useful data over long periods of time—for example, sleep and activity patterns, weight loss or gain, and changes in behavior—that doctors could use to help diagnose and monitor the treatment of mental disorders.

An SMMS could assist in this process by observing patients’ patterns of prescription drug use. Increases or decreases in medication adherence could be an indicator of the improvement or decline, respectively, in a patient’s mental state. Likewise, irregular medication use could suggest an abnormal cognitive or emotional state. In conjunction with other variables, such data could guide appropriate intervention by healthcare professionals.

MODELING AND ASSESSMENTI used analytical modeling to evaluate the im-

pact of the three major interventions in medication management—motivation and support, reminders, and improved scheduling—and how these might be combined to improve medication adherence.12 No metrics were avail-able given the lack of comparable data in the literature.

Modeling medication adherenceI defined medication adherence during an observed

period as (DTaken/DPres) × 100 = (DTaken/(DTaken + DLeft)) × 100, where DPres is the number of prescribed doses and DTaken is the number of doses taken by the patient. In the case of codependent medications, where if a dose of medication X is not taken then the doses of Y and Z should not be taken either, DLeft is the sum of normally missed (unintended) and forcibly missed (unintended) doses.

Adherence is dependent on the patient’s ability to take the medication; the SMMS’s ability to detect, process, and

sedatives and tranquilizers, and stimulants—can lead to addiction. At the same time, prescription drug overdoses have risen dramatically in the US. For example, the Cen-ters for Disease Control and Prevention reports that the misuse or abuse of prescription painkillers led to 15,000 deaths in 2008, triple the rate of 20 years ago, as well as half a million emergency room visits.14 Prescription drug addiction treatment and overdoses significantly contribute to healthcare costs.

An SMMS can help prevent these problems by enabling healthcare professionals to specify the criteria or thresh-olds for overdose or addiction, such as five doses within 24 hours or two more than the required doses over 21 days. The system can process the number of doses of multiple drugs taken at a time as well as the time between doses to determine if the patient has exceeded any predefined limits and report undesirable patterns to healthcare pro-fessionals. This could lead to changing medications to less-addictive versions, discontinuing medications, or, in the case of consistent abuse, notifying law enforcement and insurance companies.

SMMSs alone cannot stop the epidemic of prescrip-tion drug abuse, but they can help detect and reduce it. If SMMSs were widely deployed, physicians could more easily identify real or potential abusers—not an easy task when a physician is managing hundreds of patients—and respond proactively. For example, a physician could limit future prescriptions to a few days at a time or refer the pa-tient to a substance abuse clinic. The physician also could determine if a patient is diverting prescriptions or obtain-ing and filling numerous prescriptions through “doctor shopping” or polypharmacy.

Figure 4. An SMMS generates context-aware reminders using rule-based protocols, in which logical connections among a wide range of data inputs weighted for the patient determine the rules.

Context-awarereminders

Special instructions

Filtering,informationintegration,

and processing(SMMS)

Prescribed medicationsPast adherence history

Adherence goalsDoses consumed

ThresholdsRate of changePrevious values

Vitalsigns

Currentactivities

SittingWalkingRunningSleeping

Patient’s medical history(electronic medical records)



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 75

transmit a set of actions; and reminder effectiveness. Over-all medication adherence can thus be given as

1/ T × C( )k = 1

T

∑

Pjk + 1 – Pjk( ) Aijk × Eijk( ) 1 – Aijk × Eijk( )r = 1

i – 1

∏⎛

⎝⎜⎞

⎠⎟j = 1

N

∑⎛

⎝⎜

⎞

⎠⎟

j = 1

C

∑ ,

where Pjk is the probability that on the kth day the patient

took the jth medication as scheduled; Aijk is the probability

that on the kth day the system correctly determined that the patient had not taken the jth medication on the ith re-minder attempt; and E

ijk is the probability that on the kth

day, for the jth medication the ith reminder is effective. Tis the duration of medication adherence in days, C is the number of medications prescribed per day for the patient, and N is the maximum number of reminders to take a given medication.

Pjk is dependent on the patient’s cognitive abilities, the

number and availability of medications, the correct dis-pensing of each medication, the cognitive load (level of difficulty) involved in taking a medication, and the patient’s motivation to get better, including both self-motivation and support from family, friends, and healthcare professionals. A

ijk is dependent on the system’s ability to detect whether

a patient took a medication and to transmit reminders. Eijk

is dependent on the system’s effectiveness at communicat-ing a reminder.

The desirable number of reminders can be computed as NDesirable = Min (NMin-Adher, NTolerable), where NMin-Adher is the minimum number of reminders needed to achieve a cer-tain medication adherence and NTolerable is the number of

reminders needed for patients with cognitive deficiencies such as the elderly or those with mental health problems.

Intervention assessmentUsing this model, I computed results for both single and

composite interventions to improve medication adherence. Figure 5a compares the level of adherence achieved,

for up to six medications, for each basic type of interven-tion. As expected, as the number of medications per day increases, the adherence level decreases. All three strate-gies lead to a higher level of adherence, but SMMS-issued context-aware reminders are the most effective.

Figure 5b compares the impact on medication adherence of different combinations of these interventions, which include

low motivation, low reminder persistence, and best scheduling;medium motivation, low reminder persistence, and best scheduling;high motivation, low reminder persistence, and best scheduling;low motivation, high reminder persistence, and best scheduling;medium motivation, high reminder persistence, and best scheduling; andhigh motivation, high reminder persistence, and best scheduling.

To quantify these interventions, I used additional motivation (M) of 5 (low), 10 (medium), and 15 (high) percent on top of self-motivation for a patient whose adherence ranges from 50 to 80 percent. I set the number of reminders (N) to 1 to

Figure 5. Impact on medication adherence of (a) individual interventions and (b) composite interventions. As the number of medi-cations increases, the level of adherence decreases. Increasing the number of SMMS-issued context-aware reminders achieves the best results. M = motivation; N = number of reminders.

0.80

0.82

0.84

0.86

0.88

0.90

0.92

0.94

0.96

0.98

1.00

1 2 4 6Number of medications

Med

icatio

n adh

eren

ce

M= .05, N= 1M= .10, N= 1M= .15, N= 1M= .05, N= 2M= .10, N= 2M= .15, N= 2

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1 2 4 6Number of medications(a) (b)

Med

icatio

n adh

eren

ce

Motivation and supportImproved schedulingReminders to take medications



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






RESE ARCH FE ATURE

76 COMPUTER

References1. W.S. Fenton, C.R. Blyler, and R.K. Heinssen, “Determinants

of Medication Compliance in Schizophrenia: Empirical and Clinical Findings,” Schizophrenia Bull., vol. 23, no. 4, 1997, pp. 637-651.

2. U. Varshney, Pervasive Healthcare Computing: EMR/EHR, Wireless and Health Monitoring, Springer, 2009.

3. L. Osterberg and T. Blaschke, “Adherence to Medication,” The New England J. Medicine, 4 Aug. 2005, pp. 487-497.

4. D. Wan, “Magic Medicine Cabinet: A Situated Portal for Consumer Healthcare,” Proc. 1st Int’l Symp. Handheld and Ubiquitous Computing (HUC 99), Springer, 1999, pp. 352-355.

5. L. Palen and S.A. Ballegaard, “Of Pill Boxes and Piano Benches: ‘Home-Made’ Methods for Managing Medication,” Proc. ACM Conf. Computer Supported Collaborative Work(CSCW 06), ACM, 2006, pp. 79-88.

6. F. Siegemund and C. Flörkemeier, “Interaction in Perva-sive Computing Settings Using Bluetooth-Enabled Active Tags and Passive RFID Technology Together with Mobile Phones,” Proc. 1st IEEE Conf. Pervasive Computing and Comm. (PerCom 03), IEEE CS, 2003, pp. 378-387.

7. J. Dunbar-Jacob and M.K. Mortimer-Stephens, “Treatment Adherence in Chronic Disease,” J. Clinical Epidemiology,vol. 54, 2001, pp. 57-60.

8. D.H. Stefanov, Z. Bien, and W.-C. Bang, “The Smart House for Older Persons and Persons with Physical Disabilities: Structure, Technology, Arrangements, and Perspectives,” IEEE Trans. Neural Systems and Rehabilitation Eng., vol. 12, no. 2, 2004, pp. 228-250.

9. S. Helal et al., “The Gator Tech Smart House: A Program-mable Pervasive Space,” Computer, Mar. 2005, pp. 64-74.

10. I. Korhonen, J. Parkka, and M. Van Gils, “Health Monitoring in the Home of the Future,” IEEE Eng. in Medicine and Biol-ogy Magazine, vol. 22, no. 3, 2003, pp. 66-73.

11. M.B. Skov and R.T. Høegh, “Supporting Information Access in a Hospital Ward by a Context-Aware Mobile Electronic Patient Record,” Pervasive and Ubiquitous Computing,vol. 10, no. 4, 2006, pp. 205-214.

12. U. Varshney, “Wireless Medication Management System: Design and Performance Evaluation,” Proc. Wireless Tele-communications Symp. (WTS 11), IEEE, 2011; doi:10.1109/WTS.2011.5960858.

13. Nat’l Inst. Drug Abuse, “Prescription Drugs: Abuse and Addiction,” NIH publication no. 11-4881, Oct. 2011, Nat’l Institutes of Health; www.drugabuse.gov/publications/research-reports/prescription-drugs.

14. Nat’l Center for Injury Prevention and Control, “Saving Lives and Protecting People: Preventing Prescription Painkiller Overdoses,” Centers for Disease Control and Prevention; www.cdc.gov/injury/pdfs/som_focusarea/NCIPC_FactSheets_PPO_v7.pdf.

Upkar Varshney is an associate professor in the Department of Computer Information Systems at Georgia State University. His research interests include mobile and pervasive comput-ing, especially in the health domain, and wireless networks. Varshney received a PhD in telecommunications and computer networking from the University of Missouri–Kansas City. Con-tact him at [email protected].

indicate low persistence and 2 to indicate high persistence. Both motivation and reminder persistence improve the adherence level, but increasing the number of reminders achieves the best results.

Smart medication management systems can exploit both wired and wireless connections to help patients safely and correctly take prescription drugs. They can

monitor and regulate medication use, support communi-cation with caregivers to modify scheduling and increase motivation, and generate context-aware reminders. As my research indicates, such reminders are particularly effective at improving adherence, even when the patient must take multiple medications per day. SMMSs can also be integrated with other smart technologies to monitor overall health, and they provide a cost-effective solution when patients live independently and in remote areas.

Future research will explore more fine-grained models to assess the optimal combination of SMMS-based interven-tions to improve medication adherence—for example, how many reminders the system can send without exceeding a patient’s cognitive load limits. More work is also needed to test the cost, complexity, and flexibility of different SMMS implementations for various types of patients in a wide range of environments.

IEEE TRANSACTIONS ON

EMERGING TOPICS IN COMPUTING

IEEE Transactions on Emerging Topics in Computing publishes papers on emerging aspects of computer science, engineering, technology, and applications not currently covered by other IEEE Computer Society Transactions. TETC is an open access journal which allows for wider dissemination of information.

SUBMIT TODAY!Publishing in 2013

Submit your manuscript at: www.computer.org/tetc. TETC aggressively seeks proposals for Special Sections and Issues focusing on emerging topics. Prospective Guest Editors should contact the EIC of TETC (Dr. Fabrizio Lombardi, [email protected]) for further details.Submissions are welcomed on any topic within the scope of TETC. Some examples of emerging topics in computing include:

and systems learning

objects, tools, and techniques

computing and storage with new technologies



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___________

____________

___________________

______________________

http://www.qmags.com/clickthrough.asp?url=www.cdc.gov/injury/pdfs/som_focusarea/NCIPC_FactSheets_PPO_v7.pdf&id=17866&adid=P76E2

http://www.qmags.com/clickthrough.asp?url=www.drugabuse.gov/publications/research-reports/prescription-drugs&id=17866&adid=P76E1

http://www.qmags.com/clickthrough.asp?url=www.computer.org/tetc&id=17866&adid=P76A1








CLOUD COVER

an overall greener computing environment.

This new computing model is drawing interest not only from end users, businesses, and the IT industry but also from governments, researchers, and professional and industry associations. Embracing cloud computing, however, requires a new mindset—a novel way of thinking about data storage and the deployment of software and other computing functions.

In the cloud computing environment, organizations must alter their approach to addressing issues such as information security and privacy, control, and ownership. There are also some perceived—and some real—risks associated with cloud computing, including performance, security, regulatory compliance, outage and service interruptions, and data ownership.

However, for most applications and users, the benefits of cloud computing outweigh the limitations, particularly when the risks are properly assessed and addressed. As a result, cloud computing is

UP IN THE SKYDriven by cloud technology,

mobile devices such as smartphones and tablets, and applications supported by ubiquitous broadband Internet access, the computing landscape—both personal and enterprise—is changing once again. There’s an accompanying paradigm shift in the way we deliver and use IT.

The cloud computing computational model provides applications, data, computing resources, and IT management functions to clients as a service through the Internet or a dedicated network. It offers many promising benefits including scalability, flexibility, freedom from system maintenance, and lower costs, especially up front—a continuous pricing structure avoids a huge capital expenditure at the time of purchase.

Cloud computing is also helping to close the global digital (information) divide, particularly in developing economies. It might even help save our planet by providing

S everal converging and complementary factors have led to cloud computing’s emergence

as a new IT service delivery model that appeals not only to the IT industry but also to individual users, businesses, educational institutions, governments, and community organizations.

This new bimonthly column will examine cloud computing’s impact now and in the future. Contributions from researchers, developers, service providers, adopters, and others will address questions related to technological trends and developments, research opportunities, best practices, standards, and compliance requirements and regulations from various perspectives.

To trigger an ongoing discussion, this inaugural article highlights the cloud’s current status and importance, examines the key questions that cloud adopters and providers face, and explores the new paradigm’s transformational potential.

Cloud computing is more than just another IT platform shift—it will transform not only the IT industry, but every sector of society.

San Murugesan

BRITE Professional Services, Australia

Cloud Computing: The New Normal?



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






78 COMPUTER

CLOUD COVER

questions. This column will address these issues and provide a holistic examination of cloud computing in all phases of its evolution.

CLOUD OUTLOOKCloud computing is more than

just another IT platform shift—it will transform not only the IT industry, but every sector of society.

Advances in cloud computing are occurring at a staggering pace. According to a 2012 report published jointly by the Cloud Security Alliance and the Information Systems Audit and Control Association (Cloud Computing Market Maturity: Study Results,ISACA, 2012), cloud computing is “at the point of advancing from infancy to growth and is reaching a level of maturity at which enterprises can benefit greatly by adopting cloud infrastructure, platform or software service offerings.” The report also provides guidance for understanding cloud market maturity and how we can address the factors that inhibit our ability to leverage cloud computing’s benefits.

Having a cloud strategy is becoming an IT essential in many business sectors, particularly for small- and medium-size businesses (SMBs) and start-ups, as well as emerging economies. SMBs that previously couldn’t afford to invest in IT resources can become competitive by taking advantage of cloud computing services. Having access to a range of high-end IT applications and services on a scalable, as-needed basis gives SMBs tremendous opportunities and offers a significant advantage as they compete with larger businesses.

Corporations are eagerly investing in promising cloud computing technologies and services not only in developed economies but also increasingly in emerging economies—including India, China, Singapore, Taiwan, the Philippines, and South Africa—to

Questions” sidebar provides a list of key queries that arise among the various stakeholders—IT professionals and executives, end users, researchers, and policymakers and regulators.

Depending on their goals and experience with the cloud, different people are likely to give varying, even conflicting, responses to these

marching toward mainstream adoption and is being offered in many shapes and sizes.

EXPLORING THE CLOUDAs we embrace cloud computing

and leverage its features for diverse applications, several questions arise that will have an impact on the cloud’s future. The “Cloud

CLOUD QUESTIONS

T he Cloud Cover column will cover a wide range of questions pertaining to this

emerging computing paradigm, including the following:

What is cloud computing’s real poten-tial? How can we leverage it? What have we learned so far from our experience with cloud computing? What are the success factors?What opportunities does cloud comput-ing present to businesses, IT developers, and individuals? How can we exploit those opportunities? What are the key challenges in embrac-ing cloud computing and how can we address them? Is cloud computing viable for business-critical enterprise applications? How compelling are the economics of cloud computing compared to a well-managed enterprise datacenter or IT system?What types of cloud application plat-forms or frameworks will prove popular in a marketplace?How dominant will “private clouds” be within an enterprise, or a group of enterprises? For what sort of enterprises or applications?How will vertical clouds, also called community clouds, meet the special needs of the specific groups they target, such as healthcare, finance, law, and education?How dependable are current cloud ser-vices? How can providers enhance the reliability of their services and gain users’ trust? What have we learned from major cloud outages and failures? What can we do to minimize their occurrence and their impact on users?How can we manage and secure infor-mation and applications in the cloud?

How can we better monitor and manage user access to cloud applica-tions and data? What are the emerging standards and practices for cloud access management?What are the integration, interoperabil-ity, and portability issues that many cloud services and applications pres-ent? How can we address them?Could multiple clouds emerge into a “mega cloud” or “cloud of clouds?” How might this happen?How might cloud aggregation occur in the near future? What business models would be suitable for cloud aggregators?How might cloud regulations evolve and compliance requirements be established? How should enterprises manage their cloud computing applications? What are the implications for enterprise departments and IT staff? Is a cloud computing strategy now an “IT essential” for an enterprise? What are the best practices for moving exist-ing applications and data to a cloud?What are the current cloud standards and how will they evolve?What are the pros and cons of closed clouds (single-vendor solutions) versus open clouds (those built upon founda-tions that support and encourage freedom of choice and integration)?What initiatives and developments are shaping the cloud? How can profes-sional societies and industry associations help advance cloud stan-dards, policies, and regulations?How will the nexus between cloud com-puting and big data, analytics, and mobile applications give rise to new kinds of applications that previously weren’t feasible or even imagined? How will cloud computing shape the future of the IT industry, business, and society as a whole?



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 79

tralia, and an adjunct professor at the University of Western Sydney.He’s the coeditor of Harnessing Green IT: Principles and Practices(Wiley-IEEE, 2012), associate editor in chief of IT Professional, and the leader of IEEE CS Cloud Computing STC’s publishing group. Contact him at [email protected] or follow him on Twitter @santweets.

innovations and demand, will be developed in due course.

But before jumping onto the cloud bandwagon, it’s important to understand exactly what that means. The articles published in this column are intended to motivate researchers, cloud service providers, IT professionals, and industry and professional associations to address the cloud’s limitations and develop solutions to the real problems facing businesses and communities in both developed and emerging regions.

San Murugesan, Cloud Cover column editor, is the director of BRITE Professional Services, Aus-

address a region’s specific needs (S. Murugesan, “Cloud Computing Gives Emerging Markets a Lift,” IT Professional, Nov./Dec. 2011, pp. 60-62). Companies can leverage the cloud’s flexibility and versatility to enter new markets quickly with little overhead costs and reduced financial risks. Innovations in technology, design, service delivery, and business models are needed to make further inroads and embrace the cloud’s untapped potential.

W ithin the next five years, individuals and enterprises alike

will be computing in one or more clouds, public or private, and cloud computing will become the new normal. Although there are major stumbling blocks to migrating enterprise applications into the cloud—including concerns about reliability, performance, bandwidth requirements, trust, and security issues—they’re gradually being addressed. Government regulations and other compliance requirements, which often lag behind market

ESSENTIAL INDUSTRIAL IMPLEMENTATIONS OF FLOATING-POINT UNITS

DURING THE LAST DECADE:

VOLUMES 1 & 2

Edited by TC AE Elisardo Antelo, this EssentialSet surveys the industrial design of fl oating-point units during the last decade. This EssentialSet is broken into two volumes, sold separately.

Order Online: computer.org/store.

{EssentialSets} Available:NEW

Column contributions We welcome short articles

(1,500 to 2,000 words) for publi-cation in the Cloud Cover column that address the cloud questions outlined here. Submit your ideas for advancing the technology or share your experiences in har-nessing the cloud at [email protected].

The CS Cloud Computing Special Technical Community (CS CCSTC) focuses on cloud activities across the IEEE Computer Society, involving both CS members and nonmembers. The STC’s work is complimentary to the IEEE Cloud Computing Initiative (IEEE CCI), a three-year project to promote cloud efforts across IEEE.

For details or to join the STC, visit computer.org/cc.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_________

________

_____________

http://www.qmags.com/clickthrough.asp?url=www.computer.org/store&id=17866&adid=P79A1

http://www.qmags.com/clickthrough.asp?url=www.computer.org/cc&id=17866&adid=P79E1







SCIENCE FICTION PROTOT YPING


to prototype the human, cultural, ethical, and legal implications of early-stage research and technology. This allows us to envision possible futures, based on science and research, and to explore their effects so that we can then build them.

When I said to my engineering colleagues at Intel, “Hey, I have a tool that I use that could help here: I use science fiction based on science fact to help build better technology solutions,” I pretty much figured they’d kick me out of the lab. But just the opposite happened. Their response was, “Cool! Can we see it? Let’s give it a try.” They saw that science fiction prototyping (SFP as we now call it) gave us a way to establish a more robust set of requirements so that we could develop better solutions. Plus, it was science fiction, and most engineers love science fiction.

That’s what brought me to NASA on a rainy day in Washington, DC, where I was hosted by Jim Adams, NASA’s Deputy Chief Technologist. In addition to being one of the

very pragmatic. It takes about 5 to 10 years to design, develop, manufacture, and release a new platform, so it’s of vital business importance for Intel to have a vision today for computing a decade from now.

I’m a principle engineer at the company, and ultimately I pro-vide a capabilities specification that then gets translated into the hardware, software, and marketing requirements for the platform. To build these models, I collaborate with a broad team at Intel, synthesizing data from social science, technical research, and statistical and economic data, as well as from interviews with global representatives from industry, academia, governments, the military, and passionate advocates.

A key part of this process is using science fiction to envision possible futures for the technologies we’re developing. About 15 years ago, before I came to Intel, I started developing a process for using science fiction based on science fact

W elcome to the Science Fiction Prototyping column. Over the next few

years, we’ll explore the intersection of science fiction and science fact and how we can use them not only to design better technology but also to envision our possible futures.

To get things started, I thought I’d give you a little background on myself and science fiction prototyping.

GOING BOLDLYI’m a geek. I was born a geek.

I was raised by geeks. This was back before the term was even used. To tell you the truth, I’m not sure my parents even know they’re geeks, but I really don’t have the heart to tell them. (Sorry, Mom and Dad) I love all things science and all things science fiction.

As a futurist at Intel, my job is to model how people will act and interact with technology 10 to 15 years from now. This might sound like science fiction, but it’s actually

Using science fiction based on science fact to prototype the human, cultural, ethical, and legal implications of early-stage research and technology allows us to envision possible futures, and to explore their effects so that we can then build them

Brian David Johnson, Intel

When Science Fiction and Science Fact Meet



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 81

out of the Frankenstein approach to robots and technology. With his engineer’s mind, Asimov said that if humans create and program robots, why would they allow them to harm humans? It didn’t make any sense. Asimov’s Three Laws of Robotics were born when he proposed three simple laws that could prevent most of the trouble:

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2. A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protec-tion does not conflict with the First or Second Law.

Asimov’s laws continue to influence both science fiction and science fact today. These rules and the robot stories that Asimov created based on them brought logic to science fiction. There was architecture to the technology. It acted and reacted to a program. When you read I, Robot or The Rest of the Robots, you see that Asimov was testing the limits and product readiness of his laws. He was stress testing and validating them in fiction.

Once Asimov provided a logic-based vision for the future, it wasn’t long before science fiction began to explore possible innovations in ways that technology couldn’t do by itself. In 2008, Gregory Benford, an astrophysicist and science fiction author, and Elisabeth Malartre took things a step further in their book, Beyond Human: Living with Robots and Cyborgs (Forge Books, 2008). According to these authors, “Science has often followed cultural anticipation, not led it. Fiction and film have meditated upon upcoming social issues of robots and cyborgs for centuries.”

In the book, Victor Frankenstein, in his relentless pursuit of knowledge, uses science (lightning and the stitched-together corpses of criminals) to create life. Once he’s successful, Frankenstein is repulsed and fearful of his creation, and this ultimately leads to his death and the deaths of everyone he holds dear.

Writing during the height of the first Industrial Revolution (1750-1850), Shelley was exploring the moral and ethical implications of rapidly advancing science and technology. Shelley’s world had been so rapidly changed in her lifetime that she, her husband, Romantic poet Percy Shelley, and their friend, the infamous Lord Byron, were trying to come to grips with the evolving modern world.

Frankenstein was such a powerful exploration of these ideas that it was the overarching science fiction narrative for more than a century. It goes something like this: Humans strive to create smarter and more powerful technology. Humans lose control of that technology. That technology kills humans.

This narrative continued well into the 20th century, when an American science fiction writer along with his superstar editor changed everything and brought a logic architecture to the genre. Isaac Asimov, a personal hero of mine, is the legendary writer of both science fiction and science fact. But back in 1940, Asimov was just a young upstart writer trying to make a name for himself. In December of that year, he was talking to his editor at the time, John W. Campbell, about how to break

nicest guys you could ever meet, Jim is genuinely excited about the possibilities of science and technology and is also a steadfast believer that it’s our human destiny to colonize space.

As I was being introduced around, I was overcome with a dizzying realization. When you’re talking with these crazy smart men and women about going to Mars or imagining what it would take to colonize space, you’re not just shooting the breeze with fellow geeks. These are the actual people who will be the ones going to Mars and colonizing space. I quickly turned to nerd jelly.

Just before the meeting kicked off, Jim and I were talking about using science fiction to envision futures that we could build. He scanned the room and said, “You know, pretty much everyone at this conference is here because of Star Trek.”

“Really?” I replied.“Yep,” he smiled. “Many of us are

here because of the Apollo mission, but if we were honest, most of us would tell you it was science fiction that brought us to work at NASA. It was really Star Trek.”

That’s the power of science fiction. Science fiction stories, movies, and comics are the playground of our logical mind and give us the inspiration to build things most people never thought possible. SFP taps into the power of that imagination, but applies it specifically as a tool to build better technologies.

SCIENCE FICTION AND SCIENCE FACT: A LITTLE HISTORY

Science fiction and science fact have a long and synergistic relationship. Many people cite Mary Shelley’s Frankenstein (1818) as the first science fiction novel. I’m in awe of this book. Written when Shelley was a teenager, the story still has a deep and lasting resonance today.

Science fiction has now become a way for science and technology to anticipate and explore the future.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






82 COMPUTER

SCIENCE FICTION PROTOT YPING

world have begun teaching SFP not only to engineering students but also to media theory, security, medical, and even business students. The business students say, “What’s a business plan if not a work of science fiction?”

In 2011, I authored a textbook on the process called Science Fiction Prototyping: Designing the Future with Science Fiction (Morgan and Claypool, 2011). Over the years, I’ve been excited and sometimes shocked at how students and professionals all over the world are using the SFP process. People’s visions for the future, based on solid science, never cease to amaze me!

In this column, we’ll explore the different ways people are using science fiction prototyping

to explore possible futures and the human impact of science and technology. I’ll introduce you to the engineers, scientists, students, and writers who are using SFP to design a better future.

I also want to encourage you to create your own science fiction futures based on the work you’re doing. This will be a great place to talk about it, with more on the Web as well. This is an important part of what we do.

As a futurist, I believe the future is made every day by the actions of people. We all build the future. We have to ask ourselves what kind of futures we want and what kind of futures we want to avoid. SFP is a way for us all to become active participants in the futures we’re building.

Brian David Johnson, Science Fic-tion Prototyping column editor, is Intel’s Director of Future Casting. As the company’s first and only futurist, Johnson is charged with developing an actionable vision for computing in 2021. Contact him at [email protected] or follow him on Twitter @IntelFuturist.

human, our acceptance shoots back up. This fluctuation in Mori’s graph is known as the uncanny valley.

In Fake Plastic Love, I explored the human impact of the uncanny valley if we designed robots based on media celebrities, not “regular humans.” The idea was that because people don’t see celebrities as being human, their acceptance of the robot could be different. The novel also investigates some of the legal and business implications of such a robot.

The success of that science fiction novel led me to a group of international engineers, roboticists, and artificial intelligence developers who wanted to continue to explore their research in works of science fiction (more on them in an upcoming column). For the past five years, we’ve been using SFP to develop new kinds of robots and AI. As a part of this collaboration, several universities around the

This is an important step. Science fiction has now become a way for science and technology to anticipate and explore the future. It was this exploration that brought so many of us to our current field and how Star Trek brought so many people to work at NASA. But I thought we could do more.

In 2007, I wrote Fake Plastic Love(iUniverse, 2007), a book in which I used science fiction to explore a new way to jump the “uncanny valley.” This concept was Masahiro Mori’s 1970 idea that uses a graph to explain how humans become more sympathetic to robots as they begin to look more human. There’s a point when the robot looks almost human, but not human enough; then, our acceptance of that robot drops to repulsion. When it’s too close to the appearance of a human, but just a little off, we feel that it looks like a zombie or the undead. But as the robot continues to look more

Calls for PapersIEEE Micro seeks general-interest submissions for publication in upcoming issues. These works should discuss the design, performance, or application of microcomputer and microprocessor systems. Of special interest are articles on performance evaluation and workload characterization. Summaries of work in progress and descriptions of recently completed works are most welcome, as are tutorials. IEEE Micro does not accept previously published material.

Visit our author center (www.computer.org/mc/micro/author.htm) for word, figure, and reference limits. All submissions pass through peer review consistent with other professional-level technical publications,and editing for clarity, readability, and conciseness. Contact IEEE Micro at [email protected] with any questions.

www.computer.org/micro/cfp



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


________________

____________

________

__

http://www.qmags.com/clickthrough.asp?url=www.computer.org/micro/cfp&id=17866&adid=P82A2

http://www.qmags.com/clickthrough.asp?url=www.computer.org/mc/micro/author.htm&id=17866&adid=P82A1








GREEN IT

storage technologies, pumped stor-age hydropower (PSH) currently accounts for more than 99 percent of the energy storage market (“Packing Some Power,” The Economist, 3 Mar. 2012; www.economist.com/node/21548495). PSH literally involves piping water into an ele-vated reservoir, converting excess electrical energy into gravitational potential energy. Water is later released from the upper reservoir to drive turbines to generate additional power when needed. In other words, PSH pushes water uphill.

If pushing water uphill, a meta-phor applicable to any futile task, doesn’t seem odd enough, other, equally strange, new ideas are being funded in the energy storage realm. Some start-ups have suggested compressing air in underground storage locations such as abandoned mines; others have proposed moving heavy objects such as railcars uphill. In both cases, the conversion of potential energy generates power on demand.

The push toward alternate, green sources of power further drives the need for energy storage. Wind power tends to be generated at night, when demand is low. Solar power fluctuates with the weather. Thus, wind and solar technologies

in a region, utilities must agree to provide a guaranteed amount of power at all times. There are huge potential fines and penalties if the agreed-upon electrical capacity isn’t available when needed.

Consider this: power companies are most efficient when the demand for power is steady. This alleviates the need for keeping massive genera-tors on line or supplemental power from other providers—usually at a premium cost. Unfortunately, consumer demand is anything but steady. Demand fluctuates sig-nificantly and unpredictably with weather, the economy, and both regional and global events. For example, a heat wave simultaneously covering Chicago and New York City would put tremendous stress on the national power grid. When mild tem-peratures return, a precipitous drop in demand will occur.

Regulations thus require power companies to effectively generate more power than is needed to guard against shortfalls between supply and demand.

PUSHING WATER UPHILLOverprovisioning power is

wasteful unless energy can be stored. Despite the tremendous amount of investment in power-

W hen I started researching green computing more than a decade ago,

market forces were driving perfor-mance improvements without duly considering the associated costs in power use. In recent years, how-ever, there has been a sea-change in the energy world, with societies demanding reductions in energy usage and waste. This has created some oddities in the market, though upon closer examination they’re not as strange as they appear.

NOT-SO-FREE ENERGY MARKETS

Like all business enterprises, the energy market is subject to the laws of supply and demand. On the supply side, utility providers generate the power we use. On the demand side, people and businesses consume the power created. In a free market, prices would fluctuate as supply and demand ebb and flow. It logically follows that if demand increases, selling more power would lead to higher profits for energy providers’ shareholders.

However, power companies are also subject to an array of conflicting regulations and constraints not found in other markets. To operate

Reflecting on power as a commodity evokes a seemingly inevitable conclusion: the energy world is ripe with oddities.

Kirk W. Cameron, Virginia Tech

Energy Oddities, Part 1: Why the Energy World Is Odd



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


__________

http://www.qmags.com/clickthrough.asp?url=www.economist.com/node/21548495&id=17866&adid=P83E1





84 COMPUTER

GREEN IT

customers, in their calculations. This cost is fixed, and the profit margin on overhead is small or nonexistent for the power company. Thus, customers without usage are simply unwanted overhead, which, in the extreme case, lowers the overall profit margins a utility company could annually report. And when profits and margins fail to improve year upon year, CEOs get fired.

Utility companies therefore have tremendous incentive to maintain margins on each customer, and will encourage reduction in energy waste and usage to that point. However, consumers who are too green coun-teract the utility companies’ driving motivation to increase profits. Thus, in some regions such as the UK, utility companies are now charging people more for using less energy (J. Insley, “Low Users Paying More for Energy,” The Guardian, 18 Sep. 2012; www.guardian.co.uk/money/2012/sep/19/low-users-pay-more-energy).

Overprovisioning, energy storage issues, revenue decoupling, and fixed

overhead conspire to create seeming oddities in the energy market. But while utility companies appear to be illogical or unable to make up their minds, they’re ultimately driven by the desire to increase revenue and are quite predictable in their behavior.

Because utility companies and federal governments are increas-ingly acknowledging computing as a growing contributor to energy consumption and potential waste, we must consider the implications of energy market trends on computing.

Part 2 will show how energy oddities are prolific in computing as well and how market forces provide explanations for seemingly weird behaviors.

Kirk Cameron, Green IT column editor, is a professor in the Depart-ment of Computer Science at Virginia Tech. Contact him at [email protected].

what they’re selling. Although their marketing efforts might imply that they’re motivated by environmental altruism, they’re mainly driven by profits—as most companies are. Any power that they’ve earmarked in a particular year but don’t have to provide to us, because we’re not using it, is pure profit for them.

BEING GREEN COULD COST YOU

We can see that utility companies want us to use less energy so that they can increase their profits. Revenue decoupling explains this to some extent. Governments have succeeded in providing incentives for utilities to waste less power and encourage their citizens to do the same. But just how far are utilities willing to go?

Let’s apply Amdahl’s law, some-times referred to as the law of diminishing returns, to our util-ity profit formula. Amdahl’s law encourages us to focus on improving a single aspect of the equation as much as possible to determine what the maximum improvement will be in the overall system. When con-sumers use less energy, they reduce the usage term of the utility profit formula. As usage approaches 0, the formula reduces to

utility profit = number of cus-tomers × overhead.

When governments and utility companies come together to discuss allowable profits for the following year, they consider the overhead, or the average cost to deliver power to

exacerbate the energy supply-demand gap and make storage technologies even more important for the coming years.

Because this will affect their ability to generate future profits, utility providers will handsomely reward companies that enable them to best match their energy supply to fluctuating demand.

PLEASE USE LESS OF WHAT WE SELL

Under free market conditions, a utility company would have incen-tive to encourage consumers to use more energy, thus increasing profits. However, because govern-ments want to curb energy use, they implement regulatory mechanisms that tie energy utilities’ profits to the number of customers served rather than the amount of power consumed.

Where such revenue decoupling is practiced, governments work with utilities to determine how much profit they are allowed to generate annually. The profit for the utility might be as simple as

utility profit = number of cus-tomers × (overhead + usage),

where overhead is the cost for the utility to bring power to the customer and usage is the amount of power consumed.

Because it regulates total profit, revenue decoupling could lead to the following formula for profitability: if utility profit is greater than a gov-ernment-authorized amount, then utility companies must refund cus-tomers; if utility profit is less than government authorized amount, then utility companies keep the difference.

This creates another curiosity in the energy world: utility companies offer us incentives to upgrade our water heaters, replace our incandescent light bulbs, and turn off our computers at night. In other words, they want us to use less of

Wind and solar tech-nologies exacerbate the energy supply-demand gap and make storage technologies even more important for the coming years.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


________

______

_________________________________________

____

http://www.qmags.com/clickthrough.asp?url=www.guardian.co.uk/money/2012/sep/19/low-users-pay-more-energy&id=17866&adid=P84E1







OUT OF BAND

“Estupidísima” ideas are a special class of a priori bad ideas. They enjoy a status in the world of suboptimal decision making. Examples include placement of the Ford Pinto fuel tank close to the rear bumper, the installation of untested blowout preventers in deep water oil exploration, and building bridges in high-wind areas while failing to embrace the notion of aeroelastic flutter. Conceptually flawed a priori bad ideas usually produce questions like, “What were they thinking?” or, perhaps, “What were they smoking?” These are the superfund sites of stupid. They may be identified by one or more of the following red flags:

Industry associations and over-sight groups either fail to warm up to them, or are critical of them, early on in their gestation.The wisest of investors tend to shy away from them. Rollouts are frequently unpredictable and rocky.Criticism, embarrassment, litigation, or hacks evolve in parallel with the implementation.

But those of us who remember old time radio are indebted to one inconvertible law of physics: radio frequency signals don’t obey property lines. Hold that thought for a moment.

NOT EVERYTHING WE CAN DO IS WORTH DOING

There has never been a shortage of bad ideas. But there are bad ideas, worse ideas, and what I’ll call the “estupidísima.” Some purposes for which we’ve used radio frequency identification (RFID) fall into the latter category.

A posteriori bad ideas are those that seemed reasonable enough at the time but failed at the level of implementation. Examples might include New Coke, the Microsoft Bob operating system, the IBM PCjr, and the Edsel. They just didn’t catch on—not because of some fundamental flaw, but rather because they targeted a nonexistent need or misjudged a market. A posteriori bad ideas produce responses like, “I’m not actually seeing this” or “This isn’t ready for prime time.”

O ne of the social glues that bonds baby boomers together is AM radio. Not the

current AM talk show radio babble, but the good stuff of yesteryear: Chicago bluesmen, rockabilly, and iconic American rock-and-roll bands like the Zombies, Traffic, and the Spencer Davis Group.

Howard Duff as Sam Spade, the Lux Radio Theater, and the Cisco Kid brought us together with shared experiences that were staples of daily life in that bygone era. Yours Truly, Johnny Dollar; Boston Blackie;Sonny Boy Williamson II on the King Biscuit Time; and the Grand Ole Opry on Nashville’s WSM 650 did it for me. We could get a taste of life in the distant lands and exotic places on the other end of the signal.

For a variety of reasons, this format only lasted a few years—from radio’s inception in the early 20th century through the 1950s, when modern transportation removed much of the curiosity inherent in the experience, and a displacing technology called TV caught on.

Not content with pedestrian applications, some manufacturers extended RFID far beyond the level justified by good taste and common sense.

Hal Berghel

University of Nevada, Las Vegas

RFIDiocy: It’s Déjà Vu All over Again



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






86 COMPUTER

OUT OF BAND

presented to Harriman by a Soviet “good will” organization. The plaque continued to broadcast conversa-tions from the ambassador’s office until 1952, when it was discovered accidentally by a British amateur radio operator who overheard some office conversations.

There’s a second lesson here, folks: RF not only doesn’t obey property lines, it also doesn’t respond well to authority.

Flash forward 50 years: the Thing has evolved into an inexpensive and more capable alternative to bar code recognition for asset management, inventory control, point-of-sale sys-tems, pet identification, high-value chip control in the gaming industry, firearms, and the list goes on and on. As the “Current Applications of RFID Technology” sidebar indi-cates, if there’s a need to know what something is, or where it is, RFID technology is ready and waiting.

Not content with pedestrian applications, some manufacturers extended RFID far beyond the level justified by good taste and common sense. RFID has now found its way into the holy trinity of security: authentication, validation and verifi-cation. They just couldn’t leave good enough alone. Over the past decade or so, RFID tags have become nearly as ubiquitous as name tags.

RFIDIOCYSimply put, RFID isn’t a

great choice for single-token verification/authentication mechanisms—they’re both too noisy and too easy to hack. The reasons are obvious and subtle. I’ll illustrate with two examples, both applications being spectacular in their foolishness.

Keyless entry and transit passes

Most of us are familiar with using transit pass and keyless entry applications while commuting or for office building access. This

transponder) and an interrogator (aka, reader) via middleware that supports the interface between the RFID hardware and the applications software.

The RFID concept isn’t new. Leon Theremin of electronic musi-cal instrument fame invented one progenitor in the 1940s. This device, called “the Thing,” was a passive cavity resonator that derived its power from an RF signal provided by an external transmitter.

Requiring no internal power source, the Thing was easy to con-ceal and difficult to detect, thus it became useful in spying. In fact, the Russians used this technique to bug the Moscow office of US Ambassa-dor W. Averell Harriman. The Thing was embedded in a wooden plaque of the Great Seal of the United States

Eventually, they become part of the literature on ecological nightmares, engineering disasters, and the like, and, if dumb enough, will eventually be featured in eponymous documentaries.They tend to be career stoppers for the principals involved.

A recent example of an exceedingly bad idea is the use of RFID in security-challenging applications. The operative part of RFID is RF—the very phenomenon, you might recall, that doesn’t obey property lines.

RFIDRFID technology uses radio fre-

quency transmissions to exchange information between a “tag” (aka,

CURRENT APPLICATIONS OF RFID TECHNOLOGY

R FID technology is used for automatic identification and tracking in in a wide

variety of applications, including the following:

Automotive industry

Vehicle immobilizersInventory managementAgile and flexible manufacturingProduct life-cycle management

Cattle ranching and animal tracking

ID tagsTiming pigeon races

Healthcare

Patient trackingTracking of high-value pharmaceuticalsResources managementInternal appliance IDHuman implants using VeriChip

Manufacturing

Supply-chain managementWarehousingAsset managementInventory control

Defense

Logistics and inventory controlField combat

Marking of high-value assets as well as targetsIdentification, friend or foe (IFF) aircraft detectionReconnaissance

Retailing

Inventory and shelf managementTracking point-of-sale informationInformation kiosk and customer serviceLoss preventionCustomer loyalty programs

Transportation

Electronic toll collectionAutomatic vehicle identificationFleet managementCar parking and access controlElectronic vehicle registration

Marine terminal operation

Container tracking and handling

Other applications

Payment transactionsCasino chip trackingLibrary managementIDs such as enhanced driver’s licenses and passports

Source: www.berghel.net/publication/rfid/rfid.pdf



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____

http://www.qmags.com/clickthrough.asp?url=www.berghel.net/publications/rfid/rfid.pdf&id=17866&adid=P86E1





JANUARY 2013 87

the use of RFID in PASS cards followed quickly thereafter (www.smartcardalliance.org/articles/2006/06/08/smart-card-alliance-challenges-dhs-stand-on-deploying-rfid-for-whti-pass-card)—three years before the cards were put into service. A short DHS description and video showing the PASS card’s intended use are available at www.getyouhome.gov/html/rfid/rfid_how_to.html. A comparison of this video with the one at www.youtube.com/watch?v=NW3RGbQTLhE should prove illuminating.

The problem is twofold. From a privacy perspective, even if it’s encrypted, it’s not the best idea to broadcast data that is used in identification. From a security perspective, this is an invitation for RFID spoofing—hacking the system to produce bogus credentials so the bad guys look like good guys. What every narcotics trafficker and terrorist needs is a bogus RFID tag that takes on a persona with saint-like qualities.

RFID spoofing is as old as RFID itself. Spoofing wasn’t perceived as a problem in the earliest RFID applications because so little was at stake. After all, what was the likelihood that someone would spoof RFID tags to mess up a grocer’s inventory control system?

PASS cards As ill-conceived as DESFire was,

it pales in comparison to the people access security service (PASS) card. L-1 Identity Solutions, which French defense contractor Safran acquired in 2011, manufactures PASS cards, which are designed to provide a single document verifying both identity and citizenship as now required by US law. This was a mistake carried through to digital perfection if ever there was one.

The concept is simple enough. Millions of people cross US borders each year. Wouldn’t it be nice if we could speed up the process and detect potential threats as far away from the turnstile as possible? I’m sure you see where this is headed. That’s right, the Department of Homeland Security (DHS) selected RFID as the solution of choice. Immediately following the announcement, trade groups such as the Smart Card Alliance pointed out that RFID was not the best fit because its use raised security and privacy concerns.

The original State Department RFID for the PASS card system was released in 2006 (www.homelandsecuritynewswire.com/state-department-issues-rfi-whti-pass-card-system.) A Smart Card Alliance press release critical of

application was wrapped around the concept of convenience, pure and simple—for users, managers who feel more comfortable with a steady stream of exception reports, and the people who maintain access logs. But it wasn’t wrapped around the security concept.

Imagine the appeal of board-ing a bus or accessing a building without so much as a card swipe. One such solution is the NXP Mifare DESFire RFID smart card. Mifare is the encryption standard used, and NXP is the Philips Electronics sub-sidiary that makes the card. This technology was exceedingly popular for nearly a decade—at least until 2011, when virtually everyone with any interest knew how to hack it. However, this isn’t an a posteriori bad idea—it’s a serious contender for estupidísima status. Why? Because the system was built around a known vulnerability that was under-stood as far back as 1999.

There’s no shortage of online resources for information about cracking Mifare RFID cards in a variety of settings, from transpor-tation tokens to key vaults. In his blog at www.schneier.com/blog/archives/2008/08/hacking_mifare.html, security expert Bruce Schneier referred to Mifare Classic security as “kindergarten cryptography” (www.schneier.com/blog/archives/2008/08/hacking_mifare.html).

The nail in NXP DESFire’s coffin came from a “template attack,” a specific type of side channel attack, which showed that fur-ther resistance to hack attacks was futile. The “Side Channel Attacks” sidebar provides more information about these tech-niques for breaking cryptographic systems. A discussion of template attacks, including links to source documents, can be found at http://arstechnica.com/business/2011/10/researchers-hack-crypto-on-rfid-smart-cards-used-for-keyless-entry-and-transit-pass.

SIDE CHANNEL ATTACKS

W orking with his colleagues at Cryptography Research in the late

1990s, Paul Kocher, one of the SSL 3.0 architects, developed techniques for breaking cryptographic systems, called side channel attacks (www.cryptography.com). The basic idea was to use the system’s physical characteristics against itself. Kocher observed that by monitoring power consumption, timing frequencies, electromagnetic propagation, acoustic signals, and so on, it’s possible to gain enough information about processor operation to recover keys and messages.

The earliest side channel attacks like simple power analysis required some under-

standing of the circuits involved. More powerful side channel attacks such as differ-ential power analysis and high-order differential power analysis use advanced sta-tistics and are largely circuit insensitive.

Kocher’s research went viral, and subse-quent researchers have proven the viability of his concept in scores of professional publi-cations. The technique of using “compromising emanations” to gain intelli-gence from electronics was the stuff of which the National Security Agency’s Tempest proj-ect in the 1970s was made (www.wired.com/threatlevel/2008/04/nsa-releases-se).



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________________

____________________

______________________

______________________

_____

____

_________________

______________

______________________

___

____

______________________

____________________

____________________________

_____________________

____ ____

http://www.qmags.com/clickthrough.asp?url=www.homelandsecuritynewswire.com/state-department-issues-rfi-whti-pass-card-system&id=17866&adid=P87E9

http://www.qmags.com/clickthrough.asp?url=http://arstechnica.com/business/2011/10/researchers-hack-crypto-on-rfid-smart-cards-used-for-keyless-entry-and-transit-pass&id=17866&adid=P87E8

http://www.qmags.com/clickthrough.asp?url=www.schneier.com/blog/archives/2008/08/hacking_mifare.html&id=17866&adid=P87E7

http://www.qmags.com/clickthrough.asp?url=www.schneier.com/blog/archives/2008/08/hacking_mifare.html&id=17866&adid=P87E6

http://www.qmags.com/clickthrough.asp?url=www.youtube.com/watch?v=NW3RGbQTLhE&id=17866&adid=P87E5

http://www.qmags.com/clickthrough.asp?url=www.getyouhome.gov/html/rfid/rfid_how_to.html&id=17866&adid=P87E4

http://www.qmags.com/clickthrough.asp?url=www.smartcardalliance.org/articles/2006/06/08/smart-card-alliance-challenges-dhs-stand-on-deploying-rfid-for-whti-pass-card&id=17866&adid=P87E3

http://www.qmags.com/clickthrough.asp?url=www.wired.com/threatlevel/2008/04/nsa-releases-se&id=17866&adid=P87E2

http://www.qmags.com/clickthrough.asp?url=www.cryptography.com&id=17866&adid=P87E1





OUT OF BAND

88 COMPUTER

of vulnerabilities by trying to suppress technical publications and presentations, and even TV shows—for example, Adam Savage of MythBusters fame refers to such RFID censorship at www.youtube.com/watch?v=-St_ltH90Oc.

For an overview of RFID and concomitant security issues, see www.berghel.net/publications/rfid/rfid.pdf.

Hal Berghel, Out of Band column editor, is a professor of computer sci-ence at the University of Nevada, Las Vegas, where he is the director of the Identity Theft and Financial Fraud Research and Operations Center (itffroc.org). Contact him at [email protected].

motional video produced just a few months earlier (www.youtube.com/watch?v=teKBR0BvulU).

The PASS card remains in use, but it’s my understanding that DHS is no longer confident in it as a source of trusted identity. By the way, the DHS solution to the RFID spoofing problem was to place the RFID card in a metallic sleeve. Of course, this eliminates the advantage of RFID over more secure options like smart cards—which is pretty much what the Smart Card Alliance pointed out to Congress before the rollout.

The unsuitability of RFID for secure applications has been understood as

long as the technology has been available. Although it’s obvious that using RFID in secure applications isn’t appropriate, the RFID industry continues to squelch disclosures

However, the PASS card presented an opportunity to put RFID spoofing to important use.

Of course, the proponents of this ridiculous use of RFID pointed to DESFire EV1, the uber-secure, 20-year-old RFID security standard embedded in the Mifare Classic cards. But before the first batch of PASS cards was even manu-factured in spring 2008, at least one hack was presented at the Chaos Communication Congress in December 2007 (www.linux-magazine.com/Issues/2007/77/Chaos-Communication-Congress). In addition, an actual attack was demonstrated on YouTube (www.youtube.com/watch?v=NW3RGbQTLhE) in Febru-ary 2008—several months before the PASS card became an official standard. It’s useful to compare this YouTube video with the NXP pro-

Selected CS articles and columns are available for free at

http://ComputingNow.computer.org.

NEW from

EMERGING ITO AND BPO MARKETSRural Sourcing and Impact SourcingMary Lacity, Joseph Rottman, and Erran Carmel

Discusses challenges, best practices, and lessons for providers seeking to build ITO and BPO capabilities in remote areas or with unique populations and for clients seeking to buy their services.

Order .PDF ($15 / $12 members):http://bit.ly/R477Cg

Order Paperback ($19):http://bit.ly/X1rmDi



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___

_________________________

_______________

_____

______________

_______________

______

_______

http://www.qmags.com/clickthrough.asp?url=http://bit.ly/X1rmDi&id=17866&adid=P88A2

http://www.qmags.com/clickthrough.asp?url=http://bit.ly/R477Cg&id=17866&adid=P88A1


http://www.qmags.com/clickthrough.asp?url=www.berghel.net/publications/rfid/rfid.pdf&id=17866&adid=P88E5

http://www.qmags.com/clickthrough.asp?url=www.youtube.com/watch?v=-St_ltH90Oc&id=17866&adid=P88E4

http://www.qmags.com/clickthrough.asp?url=www.youtube.com/watch?v=teKBR0BvulU&id=17866&adid=P88E3

http://www.qmags.com/clickthrough.asp?url=www.youtube.com/watch?v=NW3RGbQTLhE&id=17866&adid=P88E2

http://www.qmags.com/clickthrough.asp?url=www.linux-magazine.com/Issues/2007/77/Chaos-Communication-Congress&id=17866&adid=P88E1







http://www.qmags.com/clickthrough.asp?url=www.itffroc.org&id=17866&adid=P88E6


SOCIAL COMPUTING

of treatment options, outcomes, and side effects, and by asking questions on forums organized around their ailment. Figure 1 shows Heywood’s own profile. The site currently offers data on more than 160,000 patients and 1,300 life-changing health conditions.

PatientsLikeMe’s goals go far beyond helping people in desperate circumstances maintain their social connections. As a for-profit company, the site aims to revolutionize medical research by providing scientists, for a fee, with direct access to the detailed histories of its large user base. Many diseases have been frustratingly difficult to study, with decades of work leading to little understanding of the fundamental causes and resulting in inadequate treatments. By bundling patient data and selling it to researchers, PatientsLikeMe hopes to contribute to a deeper understanding of disease processes and the development of more efficacious remedies.

PATIENTSLIKEMEIn 1998, Stephen Heywood, a

29-year-old builder and architect living in Newton, Massachusetts, received a devastating diagnosis: he had amyotrophic lateral sclerosis (ALS). Commonly known in the US and Canada as Lou Gehrig’s disease, ALS is a severe neurological illness that kills most people it afflicts within five years. As his condition worsened, Heywood learned that scientists knew little about ALS and that few treatments were available. In 2004, his two brothers and one long- time friend—all MIT engineers— decided to do something about this state of affairs and created PatientsLikeMe (www.patientslikeme.com).

PatientsLikeMe provides a platform for people to report on the symptoms and progress of their disease, the treatments they’re undergoing, and the effectiveness of those treatments. Users can get a glimpse into their future by learning about others’ experiences in terms

W hen a life-changing illness like cancer or Lou Gehrig’s disease grabs hold of us,

we cling to family and friends to support us in our work, help care for our children, and, as the symptoms become more severe, minister to our failing bodies. Unfortunately, it’s at such times that we’re least able to maintain our social network. We don’t have the time or energy to keep others up to date on our condition, let alone ask for the assistance we might need.

Social websites have sprung up to fill this need. For example, CaringBridge (www.caringbridge.org) is a nonprofit organization that provides a platform for people to share information with others about the progression and treatment of a serious medical issue. Patients use CaringBridge and similar sites to ask for support, discuss symptoms and treatment options, and generally just stay in touch with family members, friends, and healthcare professionals.

Researchers can use data from social websites dedicated to patients to develop a deeper understanding of disease processes and devise more effective treatments.

John Riedl, University of Minnesota

Eric Riedl, Harvard University

Crowdsourcing Medical Research



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


__

_____________

____

http://www.qmags.com/clickthrough.asp?url=www.patientslikeme.com&id=17866&adid=P89E2

http://www.qmags.com/clickthrough.asp?url=www.caringbridge.org&id=17866&adid=P89E1





90 COMPUTER

SOCIAL COMPUTING

PARKINSON’S DISEASE AND GAMBLING

Researchers wondered whether PatientsLikeMe might help explain the curious fact that Parkinson’s patients are much more likely than the general population to develop a gambling addiction. Leading hypotheses include: (1) some drugs used to treat Parkinson’s modify the way dopamine works in the brain, which in turn modifies risk or reward pathways; (2) people with a degenerative brain condition might be depressed and turn to gambling to “self-medicate” the depression; or (3) Parkinson’s disease itself changes the brain in ways that make gambling addiction more probable. Understanding the reasons for this link could reveal more about the disease’s underlying causes and lead to better treatment options.

To explore these hypotheses, researchers solicited more than 1,000 people with Parkinson’s and, for comparative purposes, ALS, through the PatientsLikeMe site to seek information about their experiences with the disease and with gambling (P. Wicks and G.J. MacPhee, “Pathological Gambling amongst Parkinson’s Disease and ALS Patients in an Online Community [PatientsLikeMe.com],” Movement Disorders, 15 May 2009, pp. 1085-1088).

More than 200 patients with each condition completed detailed surveys, and the results were striking. While only about 1 per-cent of people in the general population have a serious gambling problem, 3 percent of the ALS patients and 13 percent of the Parkinson’s patients reported a problem. Hypothesis 2 would not predict this outcome given the great difference between ALS and Parkinson’s patients. The researchers also found that taking a dopamine-affecting drug did not predict a gambling disorder, which is evidence against hypothesis 1.

Figure 1. Stephen Heywood’s profile on PatientsLikeMe. The site provides a platform for patients to report on the symptoms and progress of their disease, the treatments they’re undergoing, and the effectiveness of those treatments. (Source: http://upload.wikimedia.org/wikipedia/commons/1/12/StephenProfile2011.jpg.)



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


http://www.qmags.com/clickthrough.asp?url=http://upload.wikimedia.org/wikipedia/commons/1/12/StephenProfile2011.jpg&id=17866&adid=P90E1





JANUARY 2013 91

Two researchers from the California Institute of Technology conducted a rapid study using data from PatientsLikeMe (K. Felzer and H. Macedo, “The Lithium ALS World-wide Study: Six Month Update,” 16 Nov. 2008; www.its.caltech.edu/~kfelzer/SixMonthUpdate.pdf). Over a six-month period, they tracked a group of patients who convinced their doctors to prescribe lithium, and compared their prog-ress to a group of similar patients who chose not to take the drug.

Delays Progression of Amyotrophic Lateral Sclerosis,” Proc. Nat’l Acad-emy of Sciences, 12 Feb. 2008, pp. 2052-2057).

Lithium is an approved treat-ment for psychological disorders, so doctors could legally prescribe it to their ALS patients. Physicians and patients alike wanted to find out as soon as possible whether the drug’s effectiveness at inhibit-ing ALS was worth its sometimes nasty side effects, but a rigorous scientific study takes years.

The study thus offered support for the theory that Parkinson’s itself might modify the brain in a way that makes gambling more likely.

ALS AND LIMB DOMINANCESince 1941, when baseball

legend Lou Gehrig succumbed to the disease named for him, medical experts have debated why ALS seems to afflict athletes more than the general population. One theory is that exercising an arm or leg vigorously can initiate the series of neuromuscular events that eventually leads to ALS.

To determine whether the arms or legs that people use most often are more likely to be the ones first affected by ALS, researchers surveyed PatientsLikeMe members (M.R. Turner et al., “Concordance between Site of Onset and Limb Dominance in Amyotrophic Lateral Sclerosis,” J. Neurology, Neurosurgery, & Psychiatry, vol. 82, pp. 853-854). About two-thirds of the 500 or so respondents provided information useful to the study.

An ALS patient’s dominant hand was more likely to be the limb first affected, while there was no statistically significant correlation between “footedness” and the site of the disease’s onset. These results are consistent with the hypothesis that limb usage increases the probability of developing ALS, since people use their dominant hand more often but use their feet about equally.

This sort of basic understand-ing about the possible origins of a disease can lead to key breakthroughs that save lives.

IMPACT OF LITHIUM USE ON ALS

In early 2008, ALS forums, including those on PatientsLikeMe, were buzzing about a study by a group of Italian researchers suggesting that lithium might be helpful in slowing the progression of ALS (F. Fornai et al., “Lithium

COMMUNITY-BASED HEALTH APPLICATIONS

C rowdsourcing has had success in the public health domain as well as in

medical research. Consider a major public health risk such as an infectious disease outbreak or an environmental disaster, such as the leakage of radiation from the Fukushima Daiichi nuclear power plant in 2011. Combining thousands of individual users’ communications about the event can provide rapid information to help clarify what happened, when it happened, where it happened, and who is at risk.

For example, HealthMap’s Outbreaks Near Me (www.healthmap.org/outbreaksnearme) is an iPhone and Android application that lets people share information about illness in their community, which the system then aggregates into a bigger picture of a dis-ease’s epidemiology. During the fall 2009

H1N1 (avian flu) pandemic, thousands of mobile users with the Outbreaks Near Me app entered reports about the disease from media or personal observations or recorded their own illness. The system then analyzed these reports to track the disease’s spread.

Figure A shows the Outbreaks Near Me data overlaid on federal Centers for Disease Control and Prevention data (C.C. Freifeld et al., “Participatory Epidemiology: Use of Mobile Phones for Community-Based Health Reporting,” PLOS Medicine, vol. 7, no. 12, 2010; doi:10.1371/journal.pmed.1000376). The crowdsourced data gave a good sense of the disease progression, and was available within a few hours of receipt of users’ reports. In contrast, CDC’s data was of higher quality but wasn’t released until after a week of analysis.

US iP

hone

repo

rts of

influ

enza

-like

illne

ss pe

r new

user

CDC I

LINet

perce

ntag

e of in

fluen

za-li

ke ill

ness

visits

CDC ILINet percentage US iPhone reports per new user

9 0.045

0.040

0.035

0.030

0.025

0.020

0.015

0.010

0.005

0

8

7

6

5

4

3

2

1

0

Figure A. Adjusted volume of H1N1 reports from HealthMap users as compared to weighted Influenza-like Illness Surveillance Network (ILINet) visit data from the CDC, from 5 September 2009 to 30 January 2010. Pearson’s correlation = 0.74, p < 0.0001. (Source: www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.1000376.)



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________________

__

http://www.qmags.com/clickthrough.asp?url=www.its.caltech.edu/~kfelzer/SixMonthUpdate.pdf&id=17866&adid=P91E3

http://www.qmags.com/clickthrough.asp?url=www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.1000376&id=17866&adid=P91E2

http://www.qmags.com/clickthrough.asp?url=www.healthmap.org/outbreaksnearme&id=17866&adid=P91E1





SOCIAL COMPUTING

92 COMPUTER

and those of its patient community. For example, how will patients feel if they learn that their online conversations are being sold to drug companies to help fine-tune the companies’ marketing messages.

Will PatientsLikeMe be able to continue balancing its competing goals of making money with improving healthcare treatments for diseases—at the same time respecting the wishes of those who make it all possible? Wouldn’t medical research progress even faster if PatientsLikeMe made its data available to researchers for free? As valuable as crowdsourcing is for generating data, it might be even more valuable for analyzing that data.

Crowdsourced medical research offers enormous potential, but realizing that potential will require continued dedication and innovation.

PatientsLikeMe is a remarkable community: members coach, support,

and mourn one another, selfl essly doing whatever they can to help develop a cure for their disease. Although it was too late to save Stephen Heywood, who passed away in 2006, the passion that he brought to improving medical care after his ALS diagnosis will eventually prevent other patients—with ALS and thousands of other life-altering illnesses—from having to suffer and die as he did.

John Riedl, Social Computing column editor, is a professor in the Depart-ment of Computer Science and Engineering at the University of Minnesota. Contact him at [email protected].

Eric Riedl is a graduate student in the Department of Mathematics at Harvard University. Contact him at [email protected].

involve a diverse set of patients from all over the world.

There are, however, many challenges to this type of research.

Most crucial is the lack of random selection of participants. Because the patients choose what treatments they undergo, it’s always possible that some unseen hidden factor will distort the results. Rigorous stud-ies using random selection are still essential, and it’s difficult to imagine running such studies in a crowd-sourced manner.

Another challenge lies in self-reporting bias. Survey participants often try to help researchers obtain the results they seek. For example, questions in the survey used to investigate the effect of “handed-ness” on ALS might have tipped off patients to the study’s objective, leading them to misremember the history of their own disease.

A third problem is in the potential conflict between the goals of a for-profit company like PatientsLikeMe

Most of the lithium patients followed the doses recommended in the Ital-ian study.

As is often the case in ALS research, the news was discourag-ing: the study found no significant reduction in ALS symptoms from taking lithium. On the positive side, the results were made available to the ALS community relatively quickly, enabling patients to stop taking lithium and thus avoid its harmful side effects.

RESEARCH CHALLENGESThese example PatientsLikeMe

studies show some of the potential benefits of crowdsourcing medical research. The biggest advantage is the opportunity it provides research-ers to carry out studies much faster and less expensively. Crowdsourcing lets scientists develop, test, refine, and retest ideas at an unprecedented rate. Furthermore, patients don’t have to travel to participate in costly medical trials, and studies can

IEEE Computer Society’s Conference Publishing Services (CPS) is now offering conference program mobile apps! Let your attendees have their conference schedule, conference information, and paper listings in the palm of their hands.

The conference program mobile app works for Android devices, iPhone, iPad, and the Kindle Fire.

For more information please contact [email protected]

CONFERENCESin the Palm of Your Hand




qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_______

_________________

____

_________









To Order Order Onlinehttp://bit.ly/TCQPUAEnter code DZWE5FVEfor 15% off!*

Also available onhttp://amazon.com

*Note: discount code only valid on createspace.com

NEW TITLE FROM

In his new book, David Alan Grier tells the stories that technical papers omit. Moving beyond the stereo-types of nerds and social misfi ts, The Company We Keep explores the community of people who build, use, and govern modern comput-ing technology. The essays are both insightful and intimate, showing the impact of technology and the hu-man character behind it.

Save 15%



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


http://www.qmags.com/clickthrough.asp?url=www.createspace.com&id=17866&adid=P93A3

http://www.qmags.com/clickthrough.asp?url=http://amazon.com&id=17866&adid=P93A2

http://www.qmags.com/clickthrough.asp?url=http://bit.ly/TCQPUA&id=17866&adid=P93A1






CAREER OPPORTUNITIES

SENIOR PROGRAMMER ANALYST. Piscataway, NJ: Travel to multiple client locations nationwide to perform pro-gramming services including, require-ment gathering, defining functional specifications business use cases, cod-ing, integration, testing, deploying, trou-bleshooting using UML Case Diagrams, Sequence Diagrams, Class Diagrams, PCI guidelines, SOA, Code Review, DHTML, HTML, Java Script, CSS, Web Sphere in multiplatform environment. Support us-ers. Reply to: Infotech Global, Inc., 371 Hoes Court, Suite 300 A Piscataway, NJ 08854.

IT CONSULTANTS w/BackOffice Assoc., S. Harwich, MA. Data conversions/migra-tions & bus. process analysis. Multiple openings. Need Bach.’s + 2 yrs. exper., incl. w/SAP. Normally travel to/work @ client sites, telecommuting discretion-ary, but can live anywhere & travel as relocation not req’d. See www.boaweb.com/careers.htm to apply online & up-load resume. Ref. Job MS/PW.

COMPUTER PROFESSIONALS for (Tappan, NY) IT firm. Programmer Ana-lysts, Sr. Programmer Analysts, QA Ana-lyst, Sr.QA Analyst, Sr. Network Admin-istrators, AIX Administrators to develop,

UNIVERSITY OF NORTHERN IOWA, Assistant Professor of Computer Sci-ence. The Department of Computer Sci-ence at the University of Northern Iowa invites applications for a tenure-track assistant professor position to begin Au-gust 2013. Applicants must hold a Ph.D. in Computer Science or a closely-related discipline. The department seeks candi-dates able to participate widely in the CS curriculum and conduct a research program involving undergraduates. Detailed information about the posi-tion and the department are available at http://www.cs.uni.edu/To apply, visit http://jobs.uni.edu/. Applications re-ceived by January 15, 2013, will be given full consideration. EOE/AA. Pre-employ-ment background checks are required. UNI is a smoke-free campus.

SENIOR SOFTWARE ENGINEER. Gainesville, FL – Design, implement and test software solutions, algorithms, and data structures to solve transportation and logistics problems. Master’s degree in Computer Science, Engineering or a related field required. Salary commensu-rate with exp. 40 hrs/wk, 8 AM - 5 PM, M – F. Mail resume to: Innovative Scheduling, Inc., 2153 SE Hawthorne Road, Suite 128, Gainesville, FL 32641.

create,& modify general comp. applica-tions s/w or specialized utility programs, analyze user needs & develop S/w net-work solutions. Some positions required to travel & relocate. Apply w/2 copies of resume to H.R.D, Data Aixsys, Inc, 258 Oak Tree Rd, Tappan, NY 10983.

SERVICES ARCHITECT. (Islandia, NY & locs throughout US). Use App Servers, d/bases & OS. Dvlp & script Security Mgt Solutions. Troubleshoot tech integrated solution implmnt issues. Assess client’s infrastructure, bus. reqs & planned bud-get. REQS: Bachelors or foreign equiv in Comp Sci, Engg (any field) Bus or rel + 5 yrs progr exp in job &/or rel occup. Must have exp w/ architecting, scoping, assessing & implmtg Security Mngmnt products; CA Identity Mgr, Role & Com-pliance Mgr, SiteMinder, Enterprise Log Mngr. JBoss, Websphere, MS SQL/Oracle, Windows OS & Unix OS. C/C++/Java/VB.net, VB Script/Java Script/DOS. Resolving bus needs by understanding, assessing, scoping & communicating Security Mgt Solutions. Mentoring & ad-vising Jr Engrs on Security implmntions. Freq Travel Reg. Work from home benefit avail. Send resume: Althea Wilson, CA Technologies, One CA Plaza, Islandia, NY 11749, Refer to Requisition #29745.

President and Chief Operating Officer

FX Palo Alto Laboratory (FXPAL) is a research laboratory owned by Fuji Xerox Co. Ltd., a joint venture between Fujifilm Holdings and Xerox Corporations that provides

document printing and business services to the Asia Pacific market. The FXPAL mission is to invent new technologies and in conjunction with Fuji Xerox (FX) transfer them into products. FXPAL has 35 regular employees of which 18 are PhD research scientists. Research areas include multimedia computing, human computer interfaces, and communication & collaboration.

The President reports to the Chairman and Chief Executive Officer and oversees research, business development and technology transfer, f inance, human resources, and IT operations. He/she will have a collegial and collaborative style with good personal interaction skills. He/she must be a visionary with the ability to develop and articulate research plans and provide strategic leadership for the laboratory. The Chairman and President interact with senior FX executives to provide advice on emerging technologies, business strategy, products, and operations.

Requirements: The successful candidate must have a PhD in Computer Science, Engineering or related field, 15 years of research and management experience, and a demonstrated record of research, technology transfer, and business and product development success. He/she will have a sensitivity to and personal adeptness dealing with other cultures. He/she will have demonstrated trustworthiness and decisive decision-making, and adaptability to organizational change.

To apply, please email your resume to: [email protected] are an equal opportunity employer and value diversity in the workplace.

Juniper Networks is recruiting for our Sunnyvale, CAoffice:

Technical Support Engineer #22596: Support Secured Routing products, working directly with our customers and partners. Work with highly knowledgeable group of customers and as escalation point for other TAC groups within the organization.

International Trade and Compliance Specialist Staff #15494: Represent Trade Compliance functions in all design, development, implementation and test stages of all the Trade Compliance relevant business process automation projects.

Juniper Networks is recruiting for our Herndon, VAoffice:

Technical Support Engineer #22274: Deliver in-depth diagnostics and root-cause analysis for network impacting issues on company’s routing/switching products (M series, T series and MX series) to large ISP and/or enterprise customers. Understand customer’s network architecture, design, and layout and provide focused troubleshooting relevant to customer’s network.

Mail single-sided resume with job code # toAttn: MS A1.2.1.435,Juniper Networks,1194 N. Mathilda Avenue,Sunnyvale, CA 94089



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_______________

______________

http://www.qmags.com/clickthrough.asp?url=www.boaweb.com/careers.htm&id=17866&adid=P94E3

http://www.qmags.com/clickthrough.asp?url=http://jobs.uni.edu/&id=17866&adid=P94E2

http://www.qmags.com/clickthrough.asp?url=http://www.cs.uni.edu/&id=17866&adid=P94E1

maitlo:[email protected]





JANUARY 2013 95

Florida International University is a multi-campus public research university located in Miami, a vibrant, international city. FIU is recognized as a Carnegie engaged uni-versity. Its colleges and schools offer more than 180 bachelor’s, master’s and doctoral programs in fields such as computer science, engineering, international relations, ar-chitecture, law, and medicine. As one of South Florida’s anchor institutions, FIU is worlds ahead in its local and global engagement, finding solutions to the most chal-lenging problems of our time. FIU emphasizes research as a major component of its mission and enrolls 48,000 students in two campus and three centers including FIU Downtown on Brickell and the Miami Beach Urban Studios. More than 160,000 alumni live and work in South Florida. For more information about FIU, visit http://www.fiu.edu/.

The School of Computing and Information Sciences seeks exceptionally qualified candidates for multiple tenure-track and tenured faculty positions at all levels as well as non-tenure track faculty positions at the level of Instructor.

TENURE TRACK/TENURED POSITIONS (JOB ID# 505004)

We seek well-qualified candidates in all areas of Computer Science and researchers in the areas of programming languages, compilers, databases, information retrieval, computer architecture, scientific computing, big data, natural language processing, computational linguistics, health informatics, and robotics, are particularly encour-aged to apply. Preference will be given to candidates who will enhance or complement our existing research strengths.

Ideal candidates for junior positions should have a record of exceptional research in their early careers. Candidates for senior positions must have an active and proven record of excellence in funded research, publications, and professional service, as well as a demonstrated ability to develop and lead collaborative research projects. In addi-tion to developing or expanding a high-quality research program, all successful appli-cants must be committed to excellence in teaching at both graduate and undergradu-ate levels. An earned Ph.D. in Computer Science or related disciplines is required.

NON-TENURE TRACK INSTRUCTOR POSITIONS (JOB ID# 505000)

We seek well-qualified candidates in all areas of Computer Science and Information Technology. Ideal candidates must be committed to excellence in teaching a variety of courses at the undergraduate level. A graduate degree in Computer Science or related disciplines is required; significant prior teaching and industry experience and/or a Ph.D. in Computer Science is preferred.

Florida International University (FIU), the state university of Florida in Miami, is ranked by the Carnegie Foundation as a comprehensive doctoral research univer-sity with high research activity. The School of Computing and Information Sciences (SCIS) is a rapidly growing program of excellence at the University, with 36 faculty members and over 1,500 students, including 75 Ph.D. students. SCIS offers B.S., M.S., and Ph.D. degrees in Computer Science, an M.S. degree in Telecommunica-tions and Networking, and B.S., B.A., and M.S. degrees in Information Technology. SCIS has received approximately $17.5M in the last four years in external research funding, has six research centers/clusters with first-class computing infrastructure and support, and enjoys broad and dynamic industry and international partnerships.

HOW TO APPLY:

Applications, including a letter of interest, contact information, curriculum vi-tae, academic transcript, and the names of at least three references, should be submitted directly to the FIU Careers website at https://jobsearch.fiu.edu; refer to Job ID# 505004 for tenure-track or tenured positions and to Job ID# 505000for instructor positions. The application review process will begin on Janu-ary 7, 2013, and will continue until the position is filled. Further information can be obtained from the School website http://www.cis.fiu.edu, or by e-mail to [email protected].

FIU is a member of the State University System of Florida and is an Equal Opportu-nity, Equal Access Affirmative Action Employer.

SR SERVICES CONSULTANT. (Islandia, NY & locations throughout the US). Ar-chitect, dsgn & dvlp solutions & complex interfaces. Tune Clarity systems & install Clarity. REQS: Bach deg or foreign equiv in Comp Sci, Engg (any), Math, Bus or rel + 5 yrs prog exp in job &/or rel occup. Must have exp w/providing tech & func consult-ing to clients in CA Clarity incl Clarity XOG, Clarity GEL Scripts, Clarity Workflow & Clar-ity Portlets. Must be CA Clarity Certified. Freq travel req. Work from home benefit avail. Send resume to: Althea Wilson, CA Technologies, One CA Plaza, Islandia, NY 11749, Refer to Requisition #29933.

SOFTWARE ENGINEER. Dsgn, dvlp, enhance, integrate & implmt applics & systems based on bus. & end user reqmts utilizing knowl of .Net BizTalk Server, SQL Server, Oracle, Windows Server, UNIX, LINUX, HTML, XML, ADO.Net, IIS, Visual Studio, QTP, & Quality Center. Min. Req. Bachelor’s deg in Bus. Admin or Comp Sci (or foreign equiv. deg in either req. discipline) + 5 yrs work exp. in job off. Travel reqd to unanticipated client sites in USA. Res. to Job Loc: InfoLogitech, Inc., 50 Cragwood Rd, Ste 209, South Plainfield, NJ 07080.

YAHOO! INC. seeks Technical Product Manager in New York, New York to ensure platform roadmaps align with greater genome roadmap and act as a product technology expert for other business units and re-search new technologies and solutions. Req. MS or foreign equivalent in CS or rel + 3 yrs. exp., or BS + 5 yrs. exp. (TO APPLY): Submit your resume to the following URL: https://tas-yahoo.taleo.net/careersection/yahoo_pmt_cs/ jobsearch.f t l ? l a n g = e n & p o r t a l =34140220106. Must reference Req. ID#: 1250354. EOE.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


___

__________

________

http://www.qmags.com/clickthrough.asp?url=http://www.cis.fiu.edu&id=17866&adid=P95A4

http://www.qmags.com/clickthrough.asp?url=https://tas-yahoo.taleo.net/careersection/yahoo_pmt_cs/jobsearch.ftl?lang=en%26portal=34140220106&id=17866&adid=P95A2

http://www.qmags.com/clickthrough.asp?url=http://www.fiu.edu/&id=17866&adid=P95A1






http://www.qmags.com/clickthrough.asp?url=https://jobsearch.fiu.edu&id=17866&adid=P95A3

96 COMPUTER

CAREER OPPORTUNITIES

Composite Software, Inc.has the following

job opportunity available inSan Mateo, CA:

SENIOR QA DEVELOPER

Manage and lead the product server test automation group to expand, maintain and monitor the automated server test suites and their environment.

Mail resume to :Composite Software, Inc.

Attn: L. Dominguez2655 Campus Drive, Suite 200

San Mateo, CA 94403. Must reference job

code QAD-CA.

Advertisement Language

Software Testing Engineer - TSS Software Corporation, Annapolis, MD.

For software solutions developer/provider, perform various tests to validate &

verify that software products, programs & applications meet function, design, implementation, operational & effectiveness specs & standards of company & clients. Utilize VB Script & SQL & other modalities to perform tests, including smoke, black box, white box, coverage, integration, acceptance, performance & load testing. Responsible for validating persistent data before/after testing; verifying installations; testing & validating machine configurations; creating, updating, & managing test tools & automated testing scripts; managing software builds & configurations; generating test cases; & managing defect tracking & reporting. Reqs: U.S. Bachelor’s degree or foreign degree equivalent in Software Engineering, IT, or closely related field; 2yrs exp in job offered; & proficiency in VB Script & SQL. Send resume to [email protected].

UNIVERSITY OF TORONTO, The De-partment of Mathematical & Com-putational Sciences, University of Toronto Mississauga and the Gradu-ate Department of Computer Sci-ence. The Department of Mathematical & Computational Sciences, University of Toronto Mississauga and the Gradu-ate Department of Computer Science, University of Toronto invite applica-tions for a tenure-stream position in Computer Systems. The appointment is at the rank of Assistant Professor and will begin on July 1, 2013. Specific areas of interest include operating systems, networks, distributed systems, data-base systems, computer architecture, programming languages and software engineering. The University of Toronto is an international leader in computer science research and education, and the Department of Mathematical and Com-putational Sciences enjoys strong ties to other units within the University. The successful candidate will be expected to participate actively in the Graduate Department of Computer Science at the University of Toronto. Candidates should have (or be about to receive) a Ph.D. in computer science or related field. We seek outstanding applicants with an ability to pursue innovative re-search at the highest level and with a strong commitment to undergraduate and graduate teaching. Evidence of excellence in teaching and research is required. Salaries are competitive with our North American peers and will be determined according to experience and qualifications. Applicants should apply online at http://recruit.cs.toronto.edu, and include curriculum vitae, a list of publications, a research and teach-ing statement, and the names and email addresses of at least three references. Other supporting materials may also be included. We will not accept appli-cations submitted by post. Review of applications will begin on January 7, 2013 and continue until the position is filled. To ensure full consideration ap-plications should be received by Febru-ary 4, 2013. For more information about the Department of Mathematical and Computational Sciences please visit our home page: http://www.utm.utoronto.ca/math-cs-stats/. The University of To-ronto is strongly committed to diversity within its community and especially welcomes applications from visible mi-nority group members, women, Aborig-inal persons, persons with disabilities, members of sexual minority groups, and others who may contribute to the further diversification of ideas. All quali-fied candidates are encouraged to ap-ply; however, Canadians and permanent residents will be given priority.

UNIVERSIT Y OF TORONTO, Com-puter Science. The Department of Computer Science at the University of Toronto invites applications for a tenure-stream position in the area of Machine Learning. The appointment is at the rank of Assistant Professor and will begin on July 1, 2013. Candidates should have (or be about to receive) a Ph.D. in Computer Science or a related f ield. We seek outstanding applicants with demonstrated excellence in re-search at the highest level and with po-tential for excellence in undergradu-ate and graduate teaching. Salaries are competitive with our North American peers and will be determined accord-ing to experience and qualif ications. Toronto is a vibrant and cosmopolitan city, one of the most desirable in the world in which to work and live, and a major centre for advanced comput-er technologies. The Department of Computer Science is an international leader in research and teaching, with recognized strength in most areas of Computer Science. The department also has close interdisciplinary ties to other units within the University and strong interactions with the computer industry. Applicants should apply on-line at http://recruit.cs.toronto.edu, and include curriculum vitae, a list of publications, a research and teaching statement, and the names and email addresses of at least three references. Other supporting materials may also be included. We will not accept appli-cations submitted by post. If you have any questions regarding this position, please contact Sara Burns at [email protected]. Review of applications will commence on January 7, 2013 and continue until the position is f illed. To ensure full consideration, applica-tions should be received by February 4, 2013. For more information on the Department of Computer Science, see www.cs.toronto.edu. The Univer-sity of Toronto is strongly commit-ted to diversity within its community and especially welcomes applications from visible minority group members, women, Aboriginal persons, persons with disabilities, members of sexual minority groups, and others who may contribute to the further diversif ica-tion of ideas. All qualif ied candidates are encouraged to apply; however, Ca-nadians and permanent residents will be given priority.

TECHNICAL DATA DESIGNER. Posi-tion available in Enfield, CT. Design and analyze data and software systems. En-sure software adheres to best practices in line with Association for Cooperative Operations Research and Development



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________

______________

___________

___

http://www.qmags.com/clickthrough.asp?url=http://www.utm.utoronto.ca/math-cs-stats/&id=17866&adid=P96E4

http://www.qmags.com/clickthrough.asp?url=www.cs.toronto.edu&id=17866&adid=P96E3

http://www.qmags.com/clickthrough.asp?url=http://recruit.cs.toronto.edu&id=17866&adid=P96E2

http://www.qmags.com/clickthrough.asp?url=http://recruit.cs.toronto.edu&id=17866&adid=P96E1







JANUARY 2013 97

IEEE Pervasive Computing

seeks accessible, useful papers on the latest

peer-reviewed developments in pervasive,

mobile, and ubiquitous computing. Topics

include hardware technology, software

infrastructure, real-world sensing and

interaction, human-computer interaction,

and systems considerations, including

deployment, scalability, security, and privacy.

Call for Articles

Author guidelines:

www.computer.org/mc/

pervasive/author.htm

Further details:

[email protected]

www.computer.o

rg/pervasive

(ACORD) standards. Perform data map-ping utilizing applications including MapForce and database optimization using DB2 Universal Database (UDB). Send applications to: S. Flannery, Mas-sachusetts Mutual Life Insurance Com-pany, 1295 State Street, Springfield, MA 01111; Please Reference Job ID: 50450181.

CTI GROUP (HOLDINGS), INC. is seeking a Software Engineer for its In-dianapolis, Indiana location to develop, support, maintain, and test software products for telecommunications bill-ing, call management and VoIP. Must have a Bachelor’s Degree in Computer Science or related field plus five years development experience. Salary DOE. Send cover letter and resume to CTI Group (Holdings), Inc. HR – Job # SEGC1, 333 N. Alabama St., Suite 240, Indianapo-lis, IN 46204.

PROGR AMMER ANALYSTS: Design, develop, maintain, test, coding, re-viewing, programming, tuning, sup-porting software applications using PRPC, Java, J2EE, XML, Hibernate,

Spring, Struts, Weblogic, Web Services, Websphere, WSAD, Portals, Business Events, and General Interface with SQL Server & Oracle as backend reposito-ries on Windows & Unix platforms in Agile and traditional SDLC environ-ment. Req. 2 yrs exp. Send resume to Virgo, Inc. 333 N. Oxford Valley Road, Suite #405, Fairless Hills, PA 19030. Po-sition will involve working in unantici-pated locations.

SR. PROGR AMMER ANALYST (Union, NJ) Participate in application architec-ture design & implementation. Perform application documentation & provide technical support. Actively involved in testing the whole application & main-taining appropriate documentation. Masters with 1 year of experience in the related f ield. Apply with 2 copies of resumes to: HR Department, Glomark International, LLC, 1185 Morris Ave, Ste. 202, Union, NJ 07083.

SATYAM COMPUTER SERVICES, LTD. is seeking to fill numerous IT positions. Prgrm Mgrs to oversee & manage mult. IT projects, proj. planning, dvlpmnt,

implementation, acct & delivery mgmt; Proj. mgrs to oversee & manage IT teams w/dvlmpt of various sftwre apps. Sys. Analyst/Programmers/Sftwre Eng. to design, dvlp, & maintain comp soft-ware apps through all phases of sftwre dvlpmt life cycle (Sftwre Eng. may also lead a team on various projects). Sales Eng/Bus. Analyst for solutions/pre-sales activities w/specific industry domain knowledge. IT Bus. Dev. Mgrs to create new business, negotiate contracts & dvlp proposals for customized IT solu-tions. Rel. Mgrs to manage/outsource commercial IT/Eng. deals, monitor & maintain existing accts. All Tech/Mgrial. positions require a MS or BS degree or equiv. in CS, IT, Eng. or related field & relevant industry exp; Sales/Rel.Mgrs. require a MS or BS deg or equiv. in Bus. Admin, Bus. Mgmnt. or related field & relev. Industry exp. Positions are based out of corp. HQ in Parsippany, NJ & sub-ject to relocation &/or travel throughout the U.S. Qualified applicants pl. mail resume & position applied for w/JOB CODE: IE1231 to HR, Satyam Computer Services, Ltd., One Gatehall Dr., Ste. 301, Parsippany, NJ 07054.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


http://www.qmags.com/clickthrough.asp?url=www.computer.org/pervasive&id=17866&adid=P97A2

http://www.qmags.com/clickthrough.asp?url=www.computer.org/mc/pervasive/author.htm&id=17866&adid=P97A1
















http://www.qmags.com/clickthrough.asp?url=www.computer.org/mc/pervasive/author.htm&id=17866&adid=P97A1

98 COMPUTER

Apple has the following job opportunities in Cupertino, CA:

Design Verification Engineer [Req. #23640890] Ensure correct functioning of microprocessors to be used in mobile devices such as phones and tablet computers.

iOS Cellular SW Field Test Engineer [Req. #25169610] Test the telephony (phone, sms, data, etc) functionality of the iPhone by completing both documented and ad hoc testing to ensure high quality releases.

Procurement Analyst – Corporate Procurement [Req. #25173205] Perform business process reengineering and workflow studies to drive improvements in Procurement operations effectiveness and scalability.

Software Engineer [Req. #25149609] Participate in development of the next generation of networking technology.

Engineering Project Manager 4 [Req. #25153285] Work in conjunction with New Production Introduction group as well as a cross functional team of Design, Supplier Quality, Test, Service and Technical Program Engineers to ensure top quality products and services from various suppliers.

Reliability Engineer – Soft Goods [Req. #25146548] Responsible for preparing concise and detailed test plans and test reports.

Software Engineer [Req. #25306224] Work closely with internal development team, radio stack vendor, certification and field test and carrier support teams.

Sr. Software Engineer [Req. #23489232] Design, develop and distribute large scale enterprise systems using Java 2 Enterprise Edition, Oracle Database and Application Frameworks, XML,SQL/PL/SQL.

Senior Software Quality Assurance Engineer [Req. #23486691] Responsible for Accessory Protocol verification for Apple devices.

Senior iOS Wireless System Performance Engineer [Req. #25307142] Evaluate and optimize wireless system performances.

Senior Systems Engineer [Req. #25310280] Provide standard administration including account creation, system audits, and third-level troubleshooting.

Software Engineer [Req. #25309641] Gather requirements, plan architecture, design components, and write software including unit tests.

Software Development Engineer [Req. #25420419] Perform Field Testing required for carrier certification and internal quality assurance.

Software Development Engineer 4 [Req. #25421729] Design and Re-Architect platform for internal management tools to address new business requirements; improve code quality, scalability, and usability.

Bluetooth Software Quality Assurance Engineer (2 openings) [Req. #25513089] Responsible for testing Bluetooth feature and software updates.

Sr. Software Engineer, Audio/Music Department [Req #25509006] Write code for the interface part of one of our professional audio applications.

Sr. Systems Design Engineer [Req. #25465218] Lead and mentor field and Lab testing engineers in GPS performance metrics creation/evolution.

Software QA Engineer [Req. #25508635] Responsible for qualifying iOS (iPhone, iPad, iPod touch) synchronization with iCloud and other supported services such as Yahoo! and Google.

Electronic Design Engineer [Req. #25465434] Provide wireless system development and integration.

Modem DSP SW Engineer [Req. #25465345] Develop and implement software to improve modem performance on iOS devices (iPhone, iPad).

Supplier Quality Engineer [Req. #25450911] Interface with the supply chain to address on-going production issues and support synchronization of Incoming Quality at the OEM and Outgoing Quality at Tier 1 supplier.

Senior Project Manager [Req. #25465144] Responsible for overseeing software engineering and design deliverables for web applications.

Project Engineering Manager [Req. #25511794] Responsible for the technical management and support of engineering development projects within Apple’s iPod Engineering division.

Software Engineer [Req #25511940] Responsible for the research and development of a proprietary and large-scale internet application product.

RF Systems Engineer [Req. #25602618] Responsible for RF parametric qualification, evaluation, and adherence to carrier and industry standard requirements.

Some positions may require travel and may have direct reports. Mail resumes to 1 Infinite Loop M/S: 104-1GM, Attn: LJ, Cupertino, CA 95014. Principals only. EOE. Must include Req# to be considered.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






JANUARY 2013 99

IEEE TRANSACTIONS ON

EMERGING TOPICS IN COMPUTING

IEEE Transactions on Emerging Topics in Computing publishes papers on emerging aspects of computer science, computing technology, and computing applications not currently covered by other IEEE Computer Society Transactions. TETC is an open access journal which allows for wider dissemination of information.

Submit your manuscript at: www.computer.org/tetc. TETC aggressively seeks proposals for Special Sections and Issues focusing on emerging topics. Prospective Guest Editors should contact the EIC of TETC (Dr. Fabrizio Lombardi, [email protected]) for further details.

Submissions are welcomed on any topic within the scope of TETC. Some examples of emerging topics in computing include:

IT for Green Synthetic and organic computing structures and systemsAdvanced analyticsSocial/occupational computingLocation-based/client computer systemsMorphic computer designElectronic game systemsHealth-care ITComputer support for peer tutoring and learning via discovery or project work or field or lab workCreation and management of learning objects.

SUBMIT TODAY!Publishing in 2013



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


__________________

http://www.qmags.com/clickthrough.asp?url=www.computer.org/tetc&id=17866&adid=P99A1






IEEE Computer Society Membership—

Focused on Your FutureNow when you join or renew your IEEE Computer Society membership, you can choose the membership package focused specifi cally on advancing your career:

■ Software and Systems—includes IEEE SoftwareDigital Edition

■ Information and Communication Technologies (ICT)—includes IT Professional Digital Edition

■ Security and Privacy—includes IEEE Security & Privacy Digital Edition

■ Computer Engineering—includes IEEE Micro Digital Edition

In addition to your standard benefi ts, each package gives outstanding new benefi ts never offered before:

■ A digital edition of the most requested leading publication specifi c to your interest

■ A monthly digital newsletter developed EXCLUSIVELY for your focus area

■ Your choice of three FREE webinars from the extensive IEEE Computer Society selection

■ Downloads of 12 free articles of your choice from the IEEE Computer Society Digital Library (CSDL)

■ Discounts on training courses specifi c to your focus area

The Community for Technology Leaders



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM






IEEE Computer Society is expanding its services every day and we encourage you to take full advantage of all the opportunities that membership affords.

Check out the new “Welcome Webinar,” designed to orient members to the many benefi ts of Computer Society membership. Here, you may discover new opportunities for participation, continuing education, or mentoring that appeal to you. You can watch the Welcome Webinar at www.computer.org/membership.

In addition to your new benefi ts, Computer Society members still enjoy:

■ Computer magazine in print and digital editions

■ Our extensive eLearning Library made up of technical and management courses, videos, and books

■ Six technical and community newsletters

■ Discounts on publications, certifi cations, and conferences

■ A computer.org email alias ■ The opportunity to join technical committees ■ Membership in local chapters ■ Mentoring and networking

To choose one of the new, focused membership packages, go to www.computer.org/membership when you join or renew your membership. For just US$123, you receive thousands of dollars of value and much more.

JOIN OR RENEW TODAY!

Select your membership package at www.computer.org/membership or call us at 1-800-272-6657.

Grow with IEEE Computer Society



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_________

http://www.qmags.com/clickthrough.asp?url=www.computer.org/membership&id=17866&adid=P101A4



http://www.qmags.com/clickthrough.asp?url=www.computer.org&id=17866&adid=P101A1







qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


______________________________________

___________________





http://www.qmags.com/clickthrough.asp?url=www.computer.org/jobs&id=17866&adid=P102A1

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org/jobs&id=17866&adid=P102A2

COMPUTER SOCIET Y CONNECTION


EIC APPLICANTS SOUGHT FOR CS PERIODICALS, 2014-2015 TERMS

The IEEE Computer Society seeks editor in chief applicants to serve two-year terms, starting 1 January 2014, for three magazines and three transactions journals. Prospective candidates are asked to provide a complete curriculum vitae, a brief plan for the publication’s future, and a letter of support from their institu-tion or employer as pdf files by 1 March 2013.

For more information on the search process and to submit appli-cation materials for the following titles, please contact the following

: Robin Baldwin ([email protected])

: Robin Baldwin ([email protected])

: Brian Kirk ([email protected])

: Kathy Santa Maria ([email protected])

:Kathy Santa Maria ([email protected])

: Erin Espriu ([email protected])

EIC candidates should understand the industry, academic, and government aspects of the specific publication’s field and be able to attract respected experts to the editorial board. In addition, they must demonstrate the managerial skills necessary to process manuscripts through the editorial cycle in a timely fashion. Major responsibilities include

actively soliciting high-quality manuscripts from potential authors and, with support from publication staff, helping these authors get their manuscripts published;

identifying and appointing editorial board members, with the concurrence of the CS Publications Board;selecting competent manuscript reviewers, with the help of editorial board members, and managing timely reviews of manuscripts;directing editorial board members to seek special issue proposals and manuscripts in specific areas;providing a clear, broad focus through promotion of personal vision and guidance where appropriate; andresolving conflicts or problems as necessary.

Following the first two-year appointment, EICs are eligible for reappointment to a second two-year term. The IEEE Computer Society Publications Board invites comments on the tenures of the individual editors. Just one EIC is standing for reappointment to a second term in 2014-2015: Lars Heide, EIC of

.Please send any comments to Robin Baldwin ([email protected]).

NEW CHAIR NAMED FOR CS INDUSTRY ADVISORY BOARD

Shmuel Shottan, IAB chair, 2013-2014.

Shmuel Shottan, vice president of product operations and technology at Hitachi Data Systems, has been named chair of the IEEE Computer Society Industry Advisory Board.

Shottan, who leads the develop-ment of file and scale-out storage systems and solutions at Hitachi, will serve a two-year term begin-ning January 2013. He has been an IAB member since 2010 and replaces outgoing IAB chair Harold Javid,

director of External Research North America and Global Programs at Microsoft Research, who has been named to the CS Board of Governors.

The IAB comprises representa-tives from private industry and advises the society on useful pro-grams for computer professionals working in industry. Current IAB initiatives include the Corporate Affiliate Program, a new mem-bership model that extends CS education and training resources to a company’s employees, and the training webinar program, which connects company employees with Computer Society experts.

Other IAB board members include

Avijit Biswas, senior vice presi-dent of Atos Origin;Nick Bowen, vice president of technology at IBM;Deborah Dunie, executive vice president and CTO of CACI International;Steven J. Hillenius, executive vice president and executive director of the Semiconductor Research Corp.;Stephen D. Huffman, vice president and CTO, Mitre;Victor Manuel Cortes Galvan,program/project manager, Hewlett Packard;Cheryl McIntyre, director of software engineering at Lockheed Martin;Linda Rosenberg Wilbanks,director of IT Risk Management Group, US Department of Education;Ike Nassi, former executive vice president and chief scientist, SAP; and



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_______________

_________

________

_________________

______________

________________

________________

________________







maitlo:[email protected]





COMPUTER SOCIET Y CONNECTION

104 COMPUTER

www.computer.org/itpro

Neil Siegel, sector vice presi-dent, Technology and Advanced Systems, Northrop Grumman Information Systems.

The board currently has three vacancies. Technology company executives who are interested in being considered can contact Chris Ruoff, IEEE Computer Society senior manager of Sales and Channel Devel-opment at [email protected].

13 FOR 2013: CS EXPERTS FORECAST TECHNOLOGY TRENDS

In the coming year, the Internet of Things will change how consum-ers and enterprises use technology, interactive displays will become common in public spaces, robots will be used to rehabilitate patients, and visualization will help solve the big data problem.

Those are four of 13 technologi-cal advances that IEEE CS experts foresee in 2013. “The promise for the coming years is not just technology,” said incoming CS president David Alan Grier. “Technology and data—how we get data from the right sources to the right people in the right forms—that is the big issue that engages many of our members.”

The other major advances forecast for 2013 include

researchers will develop new approaches to cybersecurity,enterprises will deploy hybrid clouds and consumers will embrace personal clouds,mobile computing will meet the cloud,concern will increase over Internet censorship and control,the need for next-generation mobile computing will continue to increase, new multimedia applications will emerge for 3D printing, new approaches to securing safety-critical systems will emerge, reliability will become the biggest design challenge, and use of shared memory for mul-ticore processing will advance.

For details on each trend, along with related theme publica-tions planned during 2013, visit www.computer.org/portal/web/membership/13-Top-Trends-for-2013.

CS LAUNCHES OPEN ACCESS JOURNAL ON EMERGING TOPICS

The IEEE Computer Society has launched a new open access jour-nal,

. will publish papers on emerging aspects

of computer science, computer engineering, information technol-ogy, software, and applications not currently covered by other IEEE transactions. Example topics include green IT, synthetic and organic computing structures and systems, advanced analytics, social computing, and electronic game systems.

The online-only journal will publish preprints before the first issue appears in July 2013. For more information, see www.computer.org/portal/web/tetc.

is one of several IEEE open access journals being developed in response to the academic com-munity’s calls for articles to be freely available to the public. With open access, readers can download articles without paying subscription fees. Instead, authors pay a fee to publish their articles. The open access author fee will be $1,350 per article.

“The IEEE Computer Society is committed to providing the computer research community with the best possible venues for disseminating top work,” said Tom Conte, CS first vice president and VP for publications. “The reputation of the Computer Society’s publications is second to none, and will be no exception.”

Prospective authors also have the option of publishing open access papers in any of the CS transactions. Articles designated as being open access are automatically made publicly accessible at no charge to the user. Other articles in the journal remain available only to paid subscribers.

For more information on open access publishing, visit www.ieee.org/publications_standards/publications/authors/open_access.html.




qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


______________

________________________

__________

___

______________________

_____________________

___


http://www.qmags.com/clickthrough.asp?url=www.ieee.org/publications_standards/publications/authors/open_access.html&id=17866&adid=P104E3

http://www.qmags.com/clickthrough.asp?url=www.computer.org/portal/web/tetc&id=17866&adid=P104E1

http://www.qmags.com/clickthrough.asp?url=www.computer.org/itpro&id=17866&adid=P104A1

http://www.qmags.com/clickthrough.asp?url=www.computer.org/portal/web/membership/13-Top-Trends-for-2013&id=17866&adid=P104E2






CALL AND CALENDAR


CALLS FOR ARTICLES FOR COMPUTER

Computer seeks submissions for an October 2013 special issue on multicore memory coherence.

As we enter an era of large multicore systems, the question of efficiently supporting a shared memory model has become more important. Massively parallel archi-tectures lacking coherent shared memory have enjoyed great success in niche applications such as 3D rendering, but general program-ming developers still demand the convenience of a shared memory abstraction.

This special issue will focus on approaches to providing scalable, shared on-chip memory, paramount in a future where individual nodes will have on the order of 1,000 cores each.

Suggested topics include but are not limited to private and shared cache hierarchies; scalable memory coherence protocols, directory-based and otherwise; data layout and placement techniques; on-chip interconnects to support shared-memory abstractions; and hardware, software, and hybrid approaches.

Articles are due by 1 March 2013. Visit www.computer.org/computer/cfp10 to view the com-plete call for papers.

CALLS FOR ARTICLES FOR OTHER IEEE CS PUBLICATIONS

IEEE Security & Privacy plans a November/December 2013 special issue on health IT security and privacy.

The issue will look at how to protect privacy and strengthen the security of patients’ health information in an electronic and interconnected healthcare system. The goal of this special issue is to provide a forum for researchers, practitioners, and policymakers to delve deeply into these issues, pres-ent results, discuss experiences, and outline opinions.

The guest editors welcome case studies, experience reports, practices, research results, and standards reports.

Please email guest editors Kelly Caine ([email protected]) or Michael Lesk ([email protected]) a brief description of the article you plan to submit by 1 February 2013.

Articles are due to ScholarOne (https://mc.manuscriptcentral.com/cs-ieee) by 1 March 2013.

Visit http://www.computer.org/portal/web/computingnow/spcfp6 to view the complete call for papers.

IEEE Internet Computing plans a November/December 2013 special issue on smart cities.

Smart cities are currently the

focus of a broad research commu-nity as well as of many government and industry innovation agendas. The Internet plays a fundamental role in communication, information sharing and processing, data trans-fer and analysis, and distributed computing in many of today’s cities. The rise of the Internet of Things and the large-scale adoption of Web technologies in urban environments have proved that Internet-based solu-tions can successfully address smart cities’ multifaceted, cross-domain challenges.

This special issue seeks sub-missions about recent or ongoing research efforts and experiences in applying Internet technologies to realize the smart city vision.

Email the guest editors ([email protected]) a brief description of the article you plan to submit by 15 February 2013. Articles are due 1 March 2013. Visit www.computer.org/portal/web/computingnow/iccfp6 to view the complete call for papers.

IEEE Intelligent Systems plans a January/February 2014 special issue on the Web of Things.

The ability of mobile devices and sensors to observe and monitor their environments, increasing the coor-dination between things in the real world and their counterparts on the Web, introduces the new concept of the Web of Things. This Web of Things is expected to produce large volumes of data related to the

Submission InstructionsT he Call and Calendar section lists conferences, symposia, and workshops that the

IEEE Computer Society sponsors or cooperates in presenting.Visit www.computer.org/conferences for instructions on how to submit confer-

ence or call listings as well as a more complete listing of upcoming computer-related conferences.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________

_________

__________

_____

______________________

_____

__________

http://www.qmags.com/clickthrough.asp?url=www.computer.org/conferences&id=17866&adid=P105E6

http://www.qmags.com/clickthrough.asp?url=www.computer.org/portal/web/computingnow/iccfp6&id=17866&adid=P105E4

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org/portal/web/computingnow/spcfp6&id=17866&adid=P105E3

http://www.qmags.com/clickthrough.asp?url=https://mc.manuscriptcentral.com/cs-ieee&id=17866&adid=P105E2

http://www.qmags.com/clickthrough.asp?url=www.computer.org/computer/cfp10&id=17866&adid=P105E1








CALL AND CALENDAR

106 COMPUTER

physical world, and intelligent solutions are required to enable connectivity, internetworking, and relevance between the physical world and the corresponding digital world resources.

This special issue seeks inno-vative contributions to intelligent system and interaction design, infor-mation processing and knowledge engineering, and adaptive solutions

to assist in efficient utilization of the Web of Things. Submissions that include audio, video, and commu-nity content are encouraged.

Articles are due 1 March 2013.Visit http://www.computer.org/portal/web/computingnow/iscfp1 to view the complete call for papers.

IEEE Computer Graphics and Applications plans a November/December 2013 special issue on the computational aspects of fabrication: modeling, design, and 3D printing

The digital age in manufactur-ing is giving rise to output devices that allow rapid customization and rapid manufacturing, thereby revo-lutionizing how we design, develop, distribute, fabricate, and consume products. Given the digital nature of the process, a clear need exists for computational models that support a new way of production-related thinking and that facilitate intui-tive design, efficient representation, fast simulation, and visualization of physically realizable objects.

Articles are due 14 March 2013.Visit http://www.computer.org/portal/web/computingnow/cgacfp6 to view the complete call for papers.

IEEE Internet Computing plans a January/February 2014 special issue on the high-performance mobile Internet.

Every day, hundreds of millions of users access the Internet using handheld smartphones and tablet computers. With the increasing popularity of such devices and the availability of myriad applications specifically designed for them, it isn’t surprising that cellular data network traffic has experienced unprecedented growth. Stake-holders—including users, mobile application developers, network operators, content providers, and regulatory authorities—are thus interested in understanding the performance that cellular data net-works provide to their users.

Email the guest editors ([email protected]) a brief description of the article you plan to submit by 15 April 2013. Articles are due 1 May 2013. Visit http://www.computer.org/portal/web/computingnow/iccfp1 to view the complete call for papers.

IEEE Software plans a March/April 2014 special issue on next-generation mobile computing.

Ubiquitous, pervasive mobile computing is all around us. We use mobile computing not only when we interact with our smartphones, but also when we use a ticketing system on a bus or train, or watch videos and listen to music on our phones and portable music playing devices. Any computation system expected to move and interact with end users or other computational systems despite potential changes in net-work connectivity—including loss of connectivity or changes in type of connectivity or access point—participates in mobile computing. The number of such systems is expected to grow significantly each year over the coming decades. And mobile technology is expected to change, creating new challenges.

The guest editors seek articles that explore the next generation of mobile computing within the con-texts of mission-critical scenarios, quality-of-service differentiation, and resource constraints.

Articles are due 30 June 2013.Visit http://www.computer.org/portal/web/computingnow/swcfp2 to view the complete call for papers.

CALENDARFebruary 201323-27 Feb: HPCA 2013, IEEE Int’l Symp. on High-Performance Com-puter Architecture, Shenzhen, China; www.cs.utah.edu/~lizhang/HPCA19/index.html

March 20136-8 Mar: ISADS 2013, IEEE Int’l Symp. on High-Performance Com-

ICSE 2013T he International Conference on

Software Engineering is technically cosponsored by the IEEE Computer Society and the Association for Com-puting Machinery. ICSE 2013 will provide programs in which research-ers, practitioners, and educators present, discuss, and debate the most recent innovations, trends, experi-ences, and challenges in the field of software engineering. Contributors will be from academia, industry, and government.

ICSE 2013 will take place 18-26 May 2013 in San Francisco. Visit http://2013.icse-conferences.org for complete con-ference information.

Events in 2013FEBRUARY 201323-27 ...................................HPCA 2013

MARCH 20136-8 ....................................... ISADS 201316-23 ........................................ VR 201325-28 ................................... AINA-201325-28 ....................................SOSE 2013

APRIL 20131-5 ........................................LADC 20138-11 ......................................... STC 20138-12 ....................................... ICDE 201315-18 ............................... SYSCON 201329 Apr-1 May: ...................... NSW 2013

MAY 20136-10 ...................................... MSST 201318-26 ..................................... ICSE 201319 -22: ............................ IEEE S&P 2013



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_________

______

_____________________

_____________________

__________________________________________

__________________________________________

_____________

http://www.qmags.com/clickthrough.asp?url=www.cs.utah.edu/~lizhang/HPCA19/index.html&id=17866&adid=P106E6

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org/portal/web/computingnow/swcfp2&id=17866&adid=P106E5

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org/portal/web/computingnow/cgacfp6&id=17866&adid=P106E4

http://www.qmags.com/clickthrough.asp?url=http://2013.icse-conferences.org&id=17866&adid=P106E1

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org/portal/web/computingnow/iscfp1&id=17866&adid=P106E2

http://www.qmags.com/clickthrough.asp?url=http://www.computer.org/portal/web/computingnow/iccfp1&id=17866&adid=P106E3






Now available: A video introducing the IEEE Computer Society’s new OnlinePlus™ publication model for Transactions. Viewers will see an

overview of the great features and benefi ts included with an OnlinePlus™ subscription and will take a tour of the user-friendly interface included on the accompanying disc.

Go to www.computer.org/onlineplus to view the video and learn all about it today.

JANUARY 2013 107

puter Architecture, Mexico City; www.isadsmexico2013.mx/isads

16-23 Mar: VR 2013, IEEE Virtual Reality 2013, Orlando, Florida; http://ieeevr.org/2013

25-28 Mar: AINA-2013, 27th IEEE Int’l Conf. on Advanced Informa-tion Networking and Applications, Barcelona, Spain; http://voyager.ce.fit.ac.jp/conf/aina/2013//index.html

25-28 Mar: SOSE 2013: 7th Int’l Symp. on Service Oriented System Engineering, San Francisco; http://sei.pku.edu.cn/conference/sose2013/index.htm

April 20131-5 Apr: LADC 2013, 6th Latin-American Symp. on Dependable Computing, Rio de Janeiro, Brazil; www.ft.unicamp.br/ladc2013/

8-11 Apr: STC 2013, 25th Annual Software Technology Conf., Salt Lake City; http://sstc-online.org/home.cfm?pg=hm

8-12 Apr: ICDE 2013, 29th IEEE Int’l Conf. on Data Engineering, Bris-bane, Australia; www.icde2013.org/index.html

15-18 Apr: SYSCON 2013, IEEE Int’l Systems Conf., Orlando, Florida; http://ieeesyscon.org

29 Apr-1 May: NSW 2013, IEEE 2nd Intl Workshop on Network Science,West Point, New York; http://ieee-nsw.org

May 20136-10 May: MSST 2013, 29th IEEE Conference on Massive Data Stor-age, Lake Arrowhead, California; http://storageconference.org/index.html18-26 May: ICSE 2013, 35th Intl Conf. Software Engineering,San Francisco; http://2013.icse-conferences.org19-22 May: IEEE S&P 2013, IEEE Symp. Security and Privacy, San Francisco; www.ieee-security.org/TC/SP2013



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


____________

_______

__________

__________

_______

_____

_______

___

________________________

_______

____

_____________________

___

http://www.qmags.com/clickthrough.asp?url=www.computer.org/onlineplus&id=17866&adid=P107A1

http://www.qmags.com/clickthrough.asp?url=http://ieeesyscon.org&id=17866&adid=P107E5

http://www.qmags.com/clickthrough.asp?url=www.ieee-security.org/TC/SP2013&id=17866&adid=P107E9

http://www.qmags.com/clickthrough.asp?url=http://sei.pku.edu.cn/conference/sose2013/index.htm&id=17866&adid=P107E4

http://www.qmags.com/clickthrough.asp?url=www.icde2013.org/index.html&id=17866&adid=P107E6

http://www.qmags.com/clickthrough.asp?url=http://voyager.ce.fit.ac.jp/conf/aina/2013//index.html&id=17866&adid=P107E3

http://www.qmags.com/clickthrough.asp?url=http://storageconference.org/index.html&id=17866&adid=P107E11

http://www.qmags.com/clickthrough.asp?url=http://sstc-online.org/home.cfm?pg=hm&id=17866&adid=P107E7

http://www.qmags.com/clickthrough.asp?url=http://ieeevr.org/2013&id=17866&adid=P107E2

http://www.qmags.com/clickthrough.asp?url=www.ft.unicamp.br/ladc2013/&id=17866&adid=P107E8

http://www.qmags.com/clickthrough.asp?url=www.isadsmexico2013.mx/isads&id=17866&adid=P107E1





http://www.qmags.com/clickthrough.asp?url=http://ieee-nsw.org&id=17866&adid=P107E10

http://www.qmags.com/clickthrough.asp?url=http://2013.icse-conferences.org&id=17866&adid=P107E12


THE ERR ANT HASHTAG

The Here and Now

D uring a recent spate of unpleasant weather in New York City, I contacted my friends

Todd and Sharon to see how they were surviving the difficulties. Sharon quickly replied that neither time nor space was interfering with their operations.

“We pulled out of the city the day before the storm,” she replied, “We’re at our place upstate and are just as productive as if we were in Manhattan.”

The storm kept them away from the city for at least a week, as it damaged the power grid in their neighborhood and flooded the ground floors of many buildings around their offices. It also emphasized how much the Internet has helped us conquer the limitations of time and space.

Barely two decades ago, companies were tied to email and servers located in their own offices. A firm couldn’t abandon these servers to a hurricane without abandoning its work as well. Under such circumstances, a week in upstate New York wouldn’t be a productive period of work but a vacation marred by inclement weather.

The same technology that has caged some of the power of geography has also limited the ability of time to wither and narrow our lives. Thinking of my friend Sharon’s situation led me to a quick

Internet search that revealed more than a few remnants of her prior career as a punk rock musician: videos, lyrics, gig announcements, photos combining the high fashion of 90s punk with its implied threat of physical violence. If you didn’t look at the dates carefully, you might conclude that her band was still performing regularly at clubs that vanished from the city long ago.

The persistent memory of Sharon’s musical career comes from our efforts to make the Internet a permanent repository for the world’s knowledge. Almost from the start, the pioneers of the Internet argued that all network information should be permanent. “Pretty much the only good reason for a document to disappear from the Web,” claimed Tim Berners-Lee, “is that the company which owned the domain name went out of business or can no longer afford to keep the server running.”

Of course, documents disappear from the network for many good reasons—they might be no longer relevant or describe activities that have been overtaken by events. They could be preliminary information that has been replaced by a final version. But even if we didn’t encounter such challenges, we would still have tremendous problems keeping the network as a permanent repository of information. Software changes.

Organizations do go out of business. Ideas are reorganized into a more convenient form. “As is the case with handling change in general,” noted some Computer Society members in a recent issue of Internet Computing,“managing change in a linked data environment turns out to be hard.”

Equal to the problem of managing change in a linked data environment is the task of interpreting the information it generates. Documents carry dates that mark when they were created, but these dates might not reveal much about the context of the information. All we know is that they fall after the date for the information in the document. Sometimes it’s easy to put the documents in context: a picture of Sharon as a baby is certainly older than those of her as a business executive, but those of her as a punk rocker might inadvertently seem contemporary to those she recently posted for her company.

W ith computer systems offering the Internet’s organized knowledge to

our applications, we no longer live in an age where past is prologue. The past is now intermingled with the present and is sometimes indistinguishable from it.

David Alan Grier is the author of many books, including the recent The Company We Keep (IEEE CS, 2012). Contact him at [email protected].

David Alan Grier

With computer systems offering the Internet’s organized knowledge to our applications, we no longer live in an age where past is prologue.



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________

VI

DE

O









qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________________

____________





http://www.qmags.com/clickthrough.asp?url=http://www.servicescongress.org&id=17866&adid=PCOVER 3A1

http://www.qmags.com/clickthrough.asp?url=http://thecloudcomputing.org&id=17866&adid=PCOVER 3A2

http://www.qmags.com/clickthrough.asp?url=http://conferences.computer.org/scc/&id=17866&adid=PCOVER 3A3

http://www.qmags.com/clickthrough.asp?url=http://ieeebigdata.org/2013&id=17866&adid=PCOVER 3A4

http://www.qmags.com/clickthrough.asp?url=http://tab.computer.org/tcsc&id=17866&adid=PCOVER 3A5

http://www.qmags.com/clickthrough.asp?url=http://themobileservices.org/2013&id=17866&adid=PCOVER 3A6

http://www.qmags.com/clickthrough.asp?url=http://icws.org&id=17866&adid=PCOVER 3A7


IEEE Member Digital Library is an exclusive subscription available only to active IEEE members.

The IEEE Member Digital Library gives you the latest technology research—so you can connect ideas, hypothesize new theories and invent better solutions.

Get full-text access to the IEEE Xplore® digital library—at an exclusive price—with the only member subscription that includes any IEEE journal article or conference paper.

Choose from two great options designed to meet the needs of every IEEE member:

Fuel your imagination.

IEEE Member Digital Library BasicCreated for members who want to stayup-to-date with current research.

US$15 per month

NEW!

IEEE Member Digital LibraryDesigned for the power researcher who needs a more robust plan. Access all the IEEE content you need to explore ideasand develop better technology.

per month

Get the latest technology research.

www.ieee.org/go/freemonth



qqM

Mq

qM

MqM


qqM

Mq

qM

MqM


_____________________





http://www.qmags.com/clickthrough.asp?url=www.ieee.org/go/freemonth&id=17866&adid=PCOVER 4A1