1536-1268/05/$20.00 © 2005 IEEE ■ Published by the IEEE CS and IEEE ComSoc PERVASIVEcomputing 35

T H E S M A R T P H O N E

Supporting SocialInteraction with SmartPhones

The smart phone represents the cur-rent pinnacle of mobile phone devel-opment, coupling phone capabilitieswith the additional functionalitiesof a PDA. In this convergence

between phone and handheld computer, thephone has the dominant genes—smart phones

generally look more like phonesthan PDAs. The smart phone’sevolution from the mobilephone influences how userstend to think of these devices,

as reflected in the handset design. Smart phonesare predominately communication devices, withadditional computing power built in.

Compare this to PCs. Originally developedfrom larger computers that were pure calculat-ing machines, PCs were initially desktop systemsused for content creation and information stor-age. It was only with the advent of the Internetand browsers that they became strong commu-nication tools. Now highly capable at communi-cation, offering systems as diverse as email, Webpublishing, instant messaging, and voice over IP,the combination of computing power and com-munication lets us envisage pervasive environ-ments in which we can interact with and exploitthe advantages of digital systems. But it’s theadvent of the smart phone that will let us realizesome of these visions.

For smart phones to become successful perva-sive system components, they must support andenhance various user activities and offer useful,

effective functionality. Given their origins, clearlysmart phones must support interaction and com-munication between people. In a pervasive envi-ronment, phones exist in a social setting wherethe focus is communication, not computation.Smart phones should still perform computations,store information, and support other typicallycomputing-related tasks, but they’re likely to bemore successful if they do these things to aug-ment communication.

This perception has guided the research of agroup of students and I into the design of perva-sive computing systems. The systems we’ve builtfocus on augmenting and enhancing communi-cation; they’re communication devices first andcomputers second. The systems I report on heresupport social interaction between people—inparticular, interaction that would be difficult ifnot impossible to achieve without the smartphone technology. This dependency on the sys-tems endows them with a degree of utility thatdrives their effectiveness, which is key to increas-ing the use of pervasive systems.

Supporting social interactionsThe smart phone is a convenient, highly acces-

sible, and capable device that’s well suited to com-munication and yet can still create interesting con-tent, whether it be video, audio, or text. It’s atwo-way device, creating and consuming infor-mation, is highly personal, and is almost alwaysavailable, making it an ideal system for pervasive,supportive social computing.

The smart phone offers communication, connectivity, contentconsumption, and content creativity. Seven different systems exemplify its ability to support a wide range of social interactions, helping makepervasive computing a reality.

Russell Beale University of Birmingham

Each of our smart phone applicationsexamines one approach to augmentingsocial communication. We use a short-hand notation for interactions betweenpeople, loosely based on the numberinvolved:

• 1 represents an individual, with 1–1representing person-to-person com-munication

• N represents a group, with 1–N rep-resenting person-to-group communi-cation and N–N representing within-group communication

• � represents the world, with 1–� rep-resenting person-to-world communi-cations and N–� representing group-to-world communication

If we can support all of these commu-nication styles using the smart phone asthe main user device, we can determinethat it’s indeed a useful device for per-vasive computing, able to support a mul-titude of communication approaches.The collective effect is important; we canhardly call the smart phone a pervasivedevice if it only supports one form ofinteraction.

Localized dating serviceTo support interaction between indi-

viduals (1–1 interaction), a system shouldoffer people a new experience or at leastprovide them with an easier solution toan old problem. To identify areas fordevelopment, we matched the smartphone system’s features to interactionneeds that conventional systems weren’tadequately supporting. A smart phone’scharacteristics are easy to identify: theyhave screens with sufficient screen estateand resolution to show information(though nothing like that of notebooksor PCs), some processing power, somememory, and (free) short-range and (notfree) long-range connections. Signifi-cantly, they’re personal devices with pri-vate information stores, almost always

used by one individual. They’re text andimage compatible, and because of theirsize and battery life, they offer permanentavailability attached to the user.

Almost all smart phones support Blue-tooth communication, a wireless stan-dard offering free local (10–100 meter)connectivity. The Bluetooth protocol letsdevices discover each other and ex-change data with differing levels of userinput. It’s often used to communicatebetween the phone and a Bluetooth head-set, allowing hands-free operation of thephone, or to exchange data between thephone and a user’s PC.

Bluetooth technology maps well todating: It requires people to be in closephysical proximity and can be config-ured to allow information exchangewithout user input. Because smartphones are personal devices, they cancontain personal and private informa-tion and can selectively share this infor-mation with other devices.

We call our dating application Blue-dating. Users enter their interests anddesires using the interface shown in Fig-ure 1, as well as a profile of their desiredpartner. The system advertises this infor-mation (and only this information) overBluetooth. A second part of the applica-

tion continually searches for other pro-files over Bluetooth. When it finds one,it compares the discovered profile to thedesired profile. If the two profiles match,the system informs both users (usually byvibrating the phone) of the potentialmatch. The rest is up to the users.

The system exploits the Bluetooth pro-tocol’s local nature, alerting the user topeople nearby who are also looking fora date. This is quite different from Inter-net dating approaches, which use a sim-ilar profile system but might identify apotential partner who is many milesaway. A localized system lets users seetheir matches without having to travellong distances or arrange a formal dateand decide whether to pursue a rela-tionship. It also lets people keep theirpersonal information on their phone,unlike the Serendipity system, which hassimilar functionality but stores profileson a central server and requires a phoneconnection to the server to compare pro-files (see the related article in this issue,“Social Serendipity: Mobilizing SocialSoftware”). Bluedating also shares somefeatures with the profile-based Proemsystem1 and LoveGety,2 although bothof these systems use hardware specifi-cally designed for the dating applica-tion. Increasing the functionality of sys-tems that users already possess bringsus much closer to the pervasive com-puting ideal.

We ran trials of the Bluedating systemat the University of Birmingham. Weprovided the system to everyone whowanted it and gave questionnaires andperformed informal interviews withthose who agreed to try it. Reactionswere positive. Most users had initialreservations about such systems, as theydid about similar online services, butfound Bluedating more acceptablebecause it alerted them subtly, lettingthem interact as they chose. Because fewusers have worked with the system withany degree of seriousness, we can’t as yet

36 PERVASIVEcomputing www.computer.org/pervasive

T H E S M A R T P H O N E

Figure 1. Bluedating profile interfaceshowing the owner’s gender, age, andinterests.

report any statistically significant results.However, the positive feedback encour-aged us to redesign the Bluedating sys-tem to be more generic, thus letting ustarget a wider audience as we candevelop additional applications.

Community buildingThe new system, called BT Communi-

ties, provides a software framework thatcan run and manage multiple Bluetoothservices from a single application. Thisframework will let the device act as bothclient and server by offering services thatremote devices can discover and connectto and by letting the device search forservices offered by other devices. Theframework provides a platform that canbe easily extended with additional pack-ages and new user groups offering theirown functionalities.

To achieve this, we separated theapplication into three layers. The bot-tom layer—the framework layer—pro-vides the basic shared functionalityrequired throughout the application.This framework is completely indepen-dent of the subapplications built on topof it, letting us add new services simplyby adding new subapplications. Inte-grating the new services into the rest of the application, however, requiresanother layer. This integration layer sitsbetween the framework layer and thesubapplication layer and is composed ofa series of controller and processingclasses. All of the modifications requiredto add a new service to the applicationoccur in the integration layer.

The design of the BT Communitiesuser interface follows the style outlinedin the Nokia Series 60 UI Style Guidev1.0 (www.forum.nokia.com). BT Com-munities provides a simple menu-driveninterface that lets users easily navigatethrough the application. Because a searchcan take a long time to complete, depend-ing on the number of devices and servicesinvolved (a single device inquiry and ser-

vice search typically takes around 11 sec-onds), the system provides status screensshowing activity and cancel buttons. Tokeep the interface responsive at all times,the system performs lengthy or blockingoperations, such as searches and runningservices, in a thread separate from themain system and user interface threads.

We built a joke-sharing applicationcalled JokeSwap on top of this frame-

work, which lets people exchange jokesover Bluetooth. If someone has a joke intheir joke store, the system offers it toother devices. Devices detecting the offerexamine their joke stores for the joke,check their owners’ personal profiles tosee if it’s the sort of joke they like, and,if so, accept it and offer a joke inexchange. This form of informal infor-mation exchange supports local com-munity building. Mutual informationsharing is part of the glue that binds acommunity together; hence, sharingjokes and other informal information isintegral to supporting social systems.

We then rebuilt the Bluedating systemon top of BT Communities. We also pro-vided a chat facility, much like InstantMessenger, but local. The chat systemhas had extensive use, especially by stu-dents in lectures.

File sharingOur relative success using Bluetooth to

support individuals led us to extend theapproach to groups, thereby supportingN–N interactions. We wanted to createa system that supports groups of peopleworking together while still using thenotions of localized connectivity. This

aim is similar to that of Hummingbird,3

but we wanted to more fully exploitsmart phone capabilities and commonuses. Many users store and share docu-ments, data, videos, and photographs ontheir phones; we wanted our system toenhance this information sharing.

We thus created BT Share, whichimplements a peer-to-peer file-sharingsystem in which a user identifies a file-

store on their phone as being public andopen for sharing with the group. The sys-tem negotiates security protocols andlocking mechanisms, letting other autho-rized users access information in thepublic store. Users can access, modify,and spread information (documents,music, and so on) among their groupwithout needing a centralized server orexplicit communication. As they passeach other in their everyday lives, theirphones exchange and share relevantinformation.

BT Share is similar to peer-to-peer sys-tems common on PCs and the Internet(Kazaa, for example). Like BT Share,these systems offer a transparent simplesharing mechanism.

Because our system is Bluetooth-based, users must be near each other forthe devices to exchange information. Ittherefore encourages face-to-face inter-action. We’ve observed informally thatthe sharing mechanisms encourage usersto share images and content that theywould otherwise not explicitly send toother users. When users select an item,they provide a topic of conversation,thus enhancing interpersonal communi-cation. By encouraging and reinforcing

APRIL–JUNE 2005 PERVASIVEcomputing 37

Many users store and share documents, data,

videos, and photographs on their phones;

we wanted our system to enhance this

information sharing.

internal group relationships, BT Shareachieves our goal of supporting groupsand their interactions.

Shared spacesWe also designed two systems that sup-

port wider group communication: groupto group (N–N) or, more generally, groupto world (N–�). We’ve noticed that in thecoffee area outside our labs, where tables

and chairs encourage people to sit andchat, students congregate and shareinformation related to both work andwider social interests. Notice boards inthe area collect much of the durableinformation. Our goal was to find moreeffective ways of supporting this sharingand exchange of public information.

Existing projects in this domain shareambient and specific information in pub-lic spaces4–6 but don’t focus on pervasiveaccess. Mike Ananny and Carol Stro-hecker7 present a public space into whichusers can text-message the captions orpartial content of stories, but we focus onsupporting typical conversations, ratherthan structuring conversations aroundartifacts.8 We designed two systems thatlet users post messages to a shared spacethat we projected into the café area. Thisspace is digital but has a clear expressionin its projection into the public space, pro-viding a central locus around which weexpected interaction to occur. This letsusers without smart phones or otherdevices view the information and partic-ipate in the interactions. This feature sep-arates the approach from the BT Com-munities approach, which only connectstechnologically enabled participants.

Users can post text or image messagesfrom any mobile device.

The first system, PublicSpace, is basedon a client-server architecture, withclients—smart phones, PDAs, and lap-tops—communicating via Short MessageService (SMS), Multimedia MessagingService, 802.11b, or 802.11g technolo-gies. Clients send messages to the Java-based server, which collates the messages

into one display image and projects it. The second system, SharedSpace, uses

a peer-to-peer architecture, with eachdevice in the system forming a node inthe distributed digital space. Each nodeholds the same information. When newinformation is created, the system dis-tributes it to the nodes, each one passingmessages to the next (via SOAP9 andHTTP). Users can also view the sharedspace on the Internet using some smartphones or a PC and can view a text ver-sion of the messages using a WirelessMarkup Language interface for less pow-erful smart phones. Thus they can inter-act even when not physically colocated.

However, letting anyone post mes-sages meant that material soon over-whelmed the available display space,with small groups of users sharing infor-mation that interested few others. Inaddition to posting new material, userswanted to comment on major postings.We therefore created an alternativeshared space, public but organized bypersonal preferences. Our priority herewas to address the clutter in the space aswell as to create personalized views forsmall devices. Rather than provide a dig-ital version of an existing solution, we

took the conventional approach’s goodfeatures and enhanced the user experi-ence by applying the digital technology.

In the revised system, we analyzed eachmessage as it was posted via the clientand provided summaries of them on thedisplay. We adopted a simple algorithmthat removed common words from eachmessage and subjected the remainder toa word frequency count. We then pickedthe most common words and identifieda sentence or significant sentence frag-ment containing all of these words (orcame as close to this as possible), usingthis as the summarization. We also usedBayesian statistics to classify messages bycategory and user interests, letting usersdevelop personal profiles with the sorts ofmessages they most liked to see.

We provided two forms of display. Apublic view presented a news-orientedview of the data with summaries of themain stories appearing on the displayprojected into the communal space. Per-sonal views on the shared data presentedthe same material, but ordered by userpreference (based on the Bayesian analy-sis). So, some users would see newsmaterial as a priority, just as on the pub-lic display, whereas others would seesports first and then gossip. Dependingon the client, the display would includemore or less information. A laptopwould show all of the content of themost recent messages and summaries ofthe older ones, whereas the smart phoneclient would show summaries only untilthe user chose to drill down into a storyto read more or to comment.

We also used the Bayesian filter to auto-matically remove spam and extended thesystem by keeping the most active mes-sages (those most frequently commentedon or read in detail) near the top of thepile, rather than letting other postingsdisplace them.

We also fed information from news-groups and other digital sources intothe system, providing a shared reposi-

38 PERVASIVEcomputing www.computer.org/pervasive

T H E S M A R T P H O N E

We designed two systems that let users post

messages to a shared space that we projected into

the café area. This space is digital but has a clear

expression in its projection into the public space.

tory of material. Our intention was toprovide the raw material for gossip andsocial exchange, even if the users con-tributed nothing initially themselves.These items—“coffee-room” or “water-cooler” items—were to be the initialseeds for the informal conversations thattypically take place in these locations.

Information only remained active inthe space if people viewed and com-mented on it; uninteresting or stale mate-rial quickly disappeared from the dis-plays. People enjoyed the shared anddynamic nature of the space. Our obser-vations found that the space was used fordifferent sorts of information exchangethan the notice board that spawned theidea—rather than long-term notices,information moved through the systemover the course of a couple of days, witha few items having greater longevitybecause they provoked comment aftercomment. The system became somethingof a cross between a bulletin board andinstant messaging. However, by limitingthe uploading of comments or new itemsto only when users were in the coffee areaor its close vicinity, we retained a com-munity feel.

Integrating situatedinteraction with mobileawareness

Our next system, the Intelligent Mul-timedia Messaging System (IMMS),addresses person-to-group (1–N) inter-actions.10 A common problem in manylearning organizations is the lack of anyformal means to contact staff membersrapidly without potentially disclosingpersonal information, such as mobilephone numbers. One traditional methodof interaction between staff and studentsinvolves sticking notes on office doors.Lecturers wanting to leave messages ontheir office doors from remote locationsmust often rely on other staff members,such as a receptionist, to post a note ontheir behalf. Although this method has

worked for a long time, there are intrin-sic issues relating to a lack of securityand privacy, and the practical problemsof notes falling off doors or the lack oftimeliness in posting the note.

To address these issues, IMMS usesmobile and other technologies to increasethe interaction level between staff andstudents. We based the system on theconcept of situated interaction11,12 in an

attempt to bridge the gap between learnerand instructor and to increase mobilityand remote accessibility, necessary formodern learning situations.

IMMS involves placing several displayunits (iPAQs or smart phones) on theoffice doors of various staff members toact as information and messaging termi-nals for students. Because these devicesare placed at the locus of traditional inter-action between participants, they are sit-uated, and they afford communicationpossibilities because users arrive at thelocation with that express purpose inmind. A remote access Web-based man-agement system lets the unit’s owner setthe display contents, typically a messageor image. An SMS-based interface letsusers update the display by sending atext message from their phone to theIMMS server. For example, staff mem-bers can update their display to informscheduled visitors that they will be latebecause they’re stuck in traffic.

Student members of the departmentcan not only view the image and textualmessage but also send messages to theowner via the display unit interface. Astudent coming by to see a lecturer andfinding an empty office can use the dis-

play unit to compose a short message.IMMS will either send the message viatext message to the owner’s mobile phoneor store the message in the managementsystem for display the next time theIMMS user logs into the system, usingheuristics to determine which is appro-priate. This intelligent routing of mes-sages extends similar work on situateddoor displays by Keith Cheverst and col-

leagues,13 which used unintelligent SMSnotification. The system’s Web-basedcomponents offer configuration and mes-sage management and also allow remoteaccess to the screen display, so users onthe Internet can find lecturers’ statuseswithout having to go to their doors.

We evaluated the system with the helpof six students representing a cross-section of potential users. Using a scalefrom 0 (bad) to 10 (good), we askedthem to score the interface’s look andusability (9.33, standard deviation 0.67),the display’s content and appropriate-ness (7.83, s.d. 0.94), and the system’susefulness and functionality (9.00, s.d.0.42). The results indicate that the userssaw value in the system, which clearlymet our goals in providing a pervasiveand appropriate way for students andstaff to communicate with each other.

Mobile bloggingOur next system, SmartBlog, used

mobile blogging to support one-to-worldinteractions (1–�). Mobile devices in gen-eral, and smart phones in particular, haveexpanded beyond their communicationorigins to become accomplished photo-and video-creation tools, able to provide

APRIL–JUNE 2005 PERVASIVEcomputing 39

Mobile devices in general, and smart phones in

particular, have expanded beyond their

communication origins to become

accomplished photo- and video-creation tools.

immediate images and text regardless ofthe user’s location. Blogging is the processof publishing a personal diary or journalonline. The resulting weblogs, or blogs,are simple layouts of personal postsordered chronologically. Mobile blog-ging makes this process even moreimmediate, and given that we like pic-tures and photographs, the ability topost multimedia content makes a mobileblogging client desirable. A mobile blog-ging tool could be the best shortcutbetween bloggers and their weblogs if itovercomes the issues mobile applicationsface—for example, unreliable and inter-rupted network signals, limited interfacecapabilities, and limited processingpower.

Existing mobile blogging clients tendto make trade-offs that limit their use-fulness. For example, Azure (http://web.vee.net/projects/azure), a Java-basedblogging client, prioritizes targeting awide set of smart phones over interfaceusability, and doesn’t provide multi-media publishing. On the other hand,Nokia’s LifeBlog application (www.nokia.com/nokia/0,1522,,00.html?orig=/lifeblog) provides a slick interface withsome interesting photo-editing and syn-chronization options on a subset of Nokiaphones. However, LifeBlog doesn’t sup-port instant online publishing; instead, it

creates a local diary or gallery of anno-tated photos on the user’s PC, which canthen be uploaded—not much use if you’reaway from the PC for a long time.

We interviewed some users to deter-mine what features we needed to pro-vide. All interviewees were active blog-gers, and all felt that a decent mobileblogging system should be more than away to publish camera phone images ona Web site. Blogs are fairly immediate,thus requiring more frequent revisingthan crafted Web sites, so editing andmanagement abilities are important. Theusers also all wanted to use multimediaif it were simple, but 67 percent wereconcerned about the potential costs.

We therefore designed and imple-mented a client that targets the Symbian-based smart phone series. SmartBlogprovides an easy remote-managementtool for blogs because it’s designed tocommunicate with several bloggingaccounts (possibly heterogeneous) at thesame time. SmartBlog offers all the reg-ular blogging options for retrieving, cat-egorizing, publishing, and editing blogposts. In addition, it provides automatedmultimedia publishing. SmartBlog usesHTTP, communicating as a browserdoes; therefore, it works over any typeof Internet connectivity. The SmartBlogarchitecture is multithreaded, letting the

smart phone function as a mobile phonewithout compromising its behavior orperformance. No matter SmartBlog’sstate, the phone should continue toreceive and place calls.

The SmartBlog input dialog screenshotsin Figure 2 demonstrate the system’s abil-ity to manage different accounts and cre-ate and post information with minimaleffort. Typical blog postings take between20 and 50 seconds, depending on theimage size and amount of text. Oneadvantage of SmartBlog is that it cus-tomizes the multimedia produced fromthe phone to fit the limits imposed by pub-lishing systems. We designed SmartBlogto reduce captured images to sensible sizes(all customizable) for Internet display (seeFigure 3), thereby saving connection timewhen publishing and download time forthe users’ viewing.

Six users tested SmartBlog, and theyfound it fun and easy to use, leadingthem to post more images to their blogsthan usual. No one reported problemswith installation or the interface, thoughmost had problems setting up Bluetoothconnectivity (a known problem withNokia’s PC Suite router). Regardless, 67percent favored Bluetooth because it’sfree and fast. Moreover, all of the userswere satisfied with the system’s perfor-mance, especially because it didn’t blockany other phone functionality.

Pervasive computing will becomea reality when systems provideappropriate support and let usachieve our desires in new and

effective ways. Each of the systems we’vedescribed has demonstrated that smartphones support interactions betweenusers, whether one to one or many tomany. More importantly, the smartphone platform supports all seven of ourapplications.

40 PERVASIVEcomputing www.computer.org/pervasive

T H E S M A R T P H O N E

Figure 2. Three examples of user inputdialogs: (a) define settings, (b) createposts, and (c) import accounts.

(b) (c)(a)

The smart phone’s ability to support somany enhanced forms of communicationsuggests that it’s both pervasive and flex-ible. Furthermore, as our research hasdemonstrated, it can improve users’ socia-bility. By using the device’s technical char-acteristics, we can produce designs thatexploit the infrastructure and can thusdevelop new approaches to supportinginterpersonal communication.

ACKNOWLEDGMENTSI thank the students and colleagues who assisted indesigning, building, and evaluating the conceptsand applications described here: Nick Stoppard,Fatma Elsayed Meawad, Matthew Jones, and Nick-olas Pitsioras, and the many unknown but invalu-able users and students who participated in trials. Ialso thank the reviewers for their suggestions andcomments.

REFERENCES1. G. Kortuem, Z. Segall, and T.G.C. Thomp-

son, “Close Encounters: Supporting MobileCollaboration through Interchange of UserProfiles,” Proc. 1st Int’l Symp. Handheldand Ubiquitous Computing (HUC99), H.-W. Gellersen, ed., LNCS 1707, Springer-Verlag, 1999, pp. 171–185.

2. Y. Iwatani, “Love: Japanese Style,” WiredNews, 11 June 1998; www.wired.com/news/culture/0,1284,12899,00.html.

3. L. Holmquist, J. Falk, and J. Wigstroem,“Supporting Group Collaboration withInter-Personal Awareness Devices,” J. Per-sonal Technologies, vol. 3, no. 1–2, 1999,pp. 105–124.

4. D. Vogel and R. Balakrishnan, “InteractivePublic Ambient Displays: Transitioningfrom Implicit to Explicit, Public to Personal,Interaction with Multiple Users,” Proc.17th Ann. ACM Symp. User Interface Soft-ware and Technology, ACM Press, 2004,pp. 137–146.

5. S. Izadi et al., “Dynamo: A Public Interac-tive Surface Supporting the CooperativeSharing and Exchange of Media,” Proc.16th Ann. ACM Symp. User Interface Soft-ware and Technology, ACM Press, 2003,pp. 159–168.

6. S.K. Card, G.G. Robertson, and J.D.Mackinlay, “The Information Visualizer, an

Information Workspace,” Proc. ACMConf. Human Factors in Computing Sys-tems (CHI 91), ACM Press, 1991, pp.181–188.

7. M. Ananny and C. Strohecker, “DesigningPublic Spaces for Democratic Stories,”Proc. 1st ACM Workshop on Story Repre-sentation, Mechanism, and Context, ACMPress, 2004, pp. 47–50.

8. G.W. Fitzmaurice, “Situated InformationSpaces and Spatially Aware Palmtop Com-puters,” Comm. ACM, vol. 36, no. 7, 1993,pp. 39–49.

9. M. Gudgin et al., eds., “SOAP Version 1.2Part 1: Messaging Framework,” WorldWide Web Consortium, 2003, www.w3.org/TR/soap12-part1.

10. R. Beale and M. Jones, “Integrating SituatedInteraction with Mobile Awareness,” Proc.3rd European Conf. Mobile Learning(MLEARN), E. Murelli, G. Da Bormida, and C.Alborghetti, eds., Center for Research on theApplications of Telematics to Organizationsand Society (CRATOS), 2004.

11. M. Hirakawa and K.P. Hewagamage, “Sit-uated Computing: A Paradigm for theMobile User-Interaction with MultimediaSources,” Ann. Software Eng., vol. 12, no.1, 2001, pp. 213–239.

12. N.A. Streitz et al., “Situated Interactionwith Ambient Information: FacilitatingAwareness and Communication in Ubiqui-

tous Work Environments,” Proc. 10th Int’lConf. Human-Computer Interaction (HCIInt’l 2003), Lawrence Erlbaum Assoc.,2003.

13. K. Cheverst et al., “Exploring Situated Inter-action with Ubiquitous Office Door Display,”Proc. Int’l Workshop Situated Interaction,Conf. Computer-Supported CooperativeWork (CSCW 02), ACM Press, 2002.

For more information on this or any other comput-ing topic, please visit our Digital Library at www.computer.org/publications/dlib.

APRIL–JUNE 2005 PERVASIVEcomputing 41

Figure 3. A blog produced with SmartBlog.The system configures images to appearthis size automatically.

the AUTHOR

Russell Beale leads theAdvanced Interaction Groupin the School of ComputerScience at the University ofBirmingham, UK. His researchinterests include supportingsocial interaction, mobileand pervasive systems inter-

action, creative design, and information visual-ization and access. He received his DPhil incomputer science from the University of York.He‘s chair of the British Computer Society’sHCI Special Interest Group. Contact him atAdvanced Interaction Group, School of Com-puter Science, Univ. of Birmingham, Edgbas-ton, Birmingham B15 2TT UK; r.beale@cs.bham.ac.uk.