NASA-TM-109331

NASA D Aerospace Knowledge• ;,, . .- • _ - ,

DiffusmnResearch Project

: , _: -Paper Thi_Nmet - \

< • " The Roteof_er Networks in

Aerospace Engi_ng

Reno, Nevacla.,

AIAA94-0841

Paper Pi_seRted.at the 32nd Aerospace Sciences Meeting & ExhibitOf _e An_ "nc_i_'tute of Aeronautics and Astronautics (AIAA)RenD Hitton_R_

_ ? January11, i994_i" _

• . - ,

/ /

/

/" / /

N9_-14487

Unclas

. +

AnnP. Bishop. - .

, _ d,f/l/iL_'at Urbana-Champaign-__, Illinois .,,

. _ _ _ _ '_\ -

(NASA-ru-lO933 i ) _ASA/gt_

AEROSPACE KNOWLEDGE DIFFUSION

RESFAPCH PROJFCT. PAPER 39: THE

ROLE OF COMPUTER NETWORKS IN

AEROSPACE ENGINEERING (NASA} 16

G3/82 0190799

https://ntrs.nasa.gov/search.jsp?R=19940010014 2020-05-11T06:01:22+00:00Z

THE ROLE OF COMPUTER NETWORKS IN AEROSPACE ENGINEERING*

Ann P. BishopUniversityof Illinoisat Urbana-Champaign

Urbana, Illinois

Abstract

This paper presents selected results froman empirical investigation into the use of computernetworks in aersopace engineering. Suchnetworks allow aerospace engineers tocommunicate with people and access remoteresources through electronic mail, file transfer, andremote log-in. The study drew its subjects fromprivate sector, government and academicorganizations in the U.S. aerospace industry. Datapresented here were gathered in a mail survey,conducted in Spring 1993, that was distributed toaerospace engineers performing a wide variety ofjobs. Results from the mail survey provide asnapshot of the current use of computer networksin the aerospace industry, suggest factorsassociated with the use of networks, and identifyperceived impacts of networks on aerospaceengineering work and communication.

I. The Need for User-Based Studiesof Electronic Networkino

Both individual engineering organizations andthe federal government in the U.S. are makinglarge investments in computer networks (i.e.,telecommunications links that connect computersto each other or to other devices) in order to,among other things, increase R&D productivity,

*The research reported here was supported by theCouncil on Library Resources, NASA, DoD, and theIndiana Center for Survey Research; it was fundedunder the NASNDoD Aerospace Knowledge Diffu-sion Research Project. Portions were conductedwith the assistance of John M. Kennedy, Director,Indiana Center for Survey Research and with advicefrom Thomas E. PineUi, Assistant to the Chief ofthe Research Information and Applications Division,NASA Langley Research Center. SAE provided thesample used in this study.

facilitate technology transfer, and improve industrialcompetitiveness. Federal policy makers, networksystem designers and service providers, andworkplace managers are struggling to implementeffective systems and to develop appropriatepolicies to govern network implementation anduse. The success of institutional and nationalnetworking endeavors will depend on thedevelopment of network features, policies, andsupport programs that are based on solidknowledge of users' needs and habits andsubstantiated links between network use and

engineering outcomes. But little empiricalinformation has been gathered that can be used tohelp in understanding the impact of networkinginvestments, designs, and policies on engineeringwork. And little is known about the extent of

computer network use across different types ofengineering organizations. Thus, many majorinvestment, design, and policy decisions are beingmade solely on the basis of educated guessesabout the current use of networks and the

assumed contribution of networking to thescientific and technical enterprise.

In order to help remedy this situation, theauthor undertook an empirical investigation ofcomputer networking in engineering that collecteddata from the network user's point of view. Thestudy°s aim was to describe and explore the use ofelectronic networks by one particular group:aerospace engineers. It focused on the way thatnetworks are currently used by aerospaceengineers to facilitate communication andotherwise assist in the performance of work tasks.The study was guided by the following researchquestions:

1) What types of computer networks and networkapplications are currently used by aerospaceengineers?

2) What work tasks and communication activities

doaerospaceengineersusecomputernetworkstosupport?

3) What work-related factors are associated withthe use of computer networks by aerospaceengineers?

4) What are the impacts of network use onaerospace engineering work andcommunication?

In order to include subjects representing a widerange of work and communication activities and tolook at as many aspects of the aerospace industryas possible, "aerospace engineer" was interpretedvery broadly. It included people engaged in allphases of the development and production ofmilitary and commercial aeronautical or aerospaceequipment and processes.

IL._B=lt,kground: Comouter Networkinain Engineerir__ Settir_s

Engineers are employed to research, develop,design, test, and manufacture technology, whichmay exist in the form of either materials, products,systems, or processes. Engineering is a complex,information- and communication-intensive activitythat involves invention, problem-solving, andcoordination of many independent efforts."Concurrent engineering," a notion that is currentlypopular in engineering management circles,focuses on the perceived need for better andfaster communication, coordination, and

integration of the work and information contributedby all of the people involved in the development,production, and marketing of a particulartechnology. Many engineering organizations areexploring the ability of computers and electronicnetworks to facilitate concurrent engineedng 1andimprove the performance of engineers and thetechnical quality of their work.2-8 Industrialorganizations hope that by facilitatingcommunication and improving coordination,electronic networks will decrease both the costsand time needed to bring products to market. Dueto proprietary and security concerns, manyengineedng organizations have implemented theirown private, high-speed networks that are usedonly by their own employees and affiliates. Theneed for the completely reliable electronic transfer

of very large amounts of data also makes the use ofmost commercial networks inadequate for someindustriesand applications.

Today, engineers use computers to performcalculations; to produce and evaluate drawings,designs, and prototypes (CAD/CAM); to maintainand archive the "corporate memory," i.e., all thecontracts, designs, schedules, assumptions,constraints, procedures, data, etc., associated witheach particular project; to write and edit documentsand prepare presentations; to run projectmanagement software, and to control equipment.Computer networks are also playing an increasinglyimportant role in engineering work. For example,engineers use networks to receive data collectedby remote instruments. Networks facilitate thetransfer of documents and designs and are used toautomate the manufacturing process. Electronicdata interchange (EDI) is used to exchange ordersand invoices with vendors and suppliers, andcontracts with clients and customers. Networks arealso used for information retrieval in connectionwith both in-house and commercial databases, g-lo

Finally, engineers also use computer networksfor a variety of communication purposes. 11-1sForinstance, they can exploit computer-basedmessage systems to call on the expertise, ideas,and advice of other members of their communityand to locate resources. Electronic mail and variouscomputer conferencing applications are also usedto schedule and coordinate work or even conductmeetings, since they can be used to contactproject team members, managers, people in otherdepartments or divisions, and consultants inoutside organizations. Electronic mail and bulletinboards are sometimes used to facilitatecommunication with customers and funders, aswell.

There is a growing body of empirical researchthat examines the characteristics, use, and effectsof computer-mediated communication.IS-21 Fewstudies attempt to describe these variables in termsof particular kinds of work, except by comparingbroad job categories, for example, managers,professionals, and clerical workers. 22 With therecent proliferation of electronic networks, anumber of empirical efforts dedicated to exploringthe use of electronic networks for communication

2

by scientists and engineers have beenundertaken.23-32Thereseemsto beagreementthat electronic communicationis used foradministrative,technical,and socialpurposes.Muchof thisworkseemscompatiblewithfindingsaboutthe natureof engineeringcommunicationand its relationship to engineering work andproductivity, although virtually no studies havedealt exclusively or extensively with engineers.The capabilities and characteristics of electroniccommunication, in other words, seem to "match,"to some extent, the nature and requirements ofengineering work, knowledge, andcommunication. But new questions and issueshave been raised and a number of conflictingfindings have been presented. All in all, very little isknown about the characteristics, use, and impact ofelectronic communication from the engineer'spoint of view.

The aerospace industry possesses a numberof characteristics that make it a natural environmentfor the implementation of electronic networks. It isa high technology industry, already highlycomputerized. It involves significant R&D, which isan especially communication-intensive activity.Further, its end products are highly complex,calling for a great deal of work task coordinationandthe integration of information created by diversepeople. In describing the business andtechnology strategy in place at British Aerospace,Hall33 emphasized the need for increasedcomputing and communications capabilities inaerospace firms aiming to design, develop, makeand market complex systems while maintaining atechnical competitive edge and reducing unit costs(p. 16-2). He noted that a number of typicalinformation technology opportunities wereparticularly relevant to the aerospace industry,suchas "improved productivity, better competitive edge,reduced timescales, closer collaboration, morestreamlined management, better commonality ofstandards across sites, more operational flexibility,[and] constructive change of workforce skill levels"(p. 16-2).

Rachowitz et al.34describe effortsat Gmmmanto realize a fully distributed computingenvironment. Grumman's goal is to implement asystem of networked workstations in order to "cost-effectively optimize the computing tools available

to the engineers, while promoting the systematicimplementation of concurrent engineering amongproject teams" (p. 38). The network includes PCsand software to be used for communication.Grumman assumes that their computer/informationintegrated environment (CIE) will result in "productoptimization o- quality products manufactured withfewer errors in shorter time and at a lower cost" (p.66).

Black35 presents a brief overview of the usesand advantages of computer conferencingsystems, noting that computer conferencing is a"very powerful tool for the transfer of information inall areas of research and development" and "anatural for the AGARD [Advisory Group forAerospace Research and Development]community .... "(p. 13-4). Moholm3S describes theapplication of the Department of Defense'sComputer-aided Acquisition and Logistics Support(CALS) initiative to the aerospace community.CALS mandates the use of specific standards forthe electronic creation and transmission oftechnical information associated with weaponssystems development. Eventually all Departmentof Defense contractors and subcontractors will berequired to create and distribute in digital form allthe drawings, specifications, technical data,documents, and support information required overthe entire lifecycle of a military project. The CALSsystem may be a significant impetus to networkingfor aerospace firms.

Few empirical studies of computer networkingin the aerospace industry have been conducted,although a number of the surveys conducted aspart of the NASAJDoD Aerospace KnowledgeDiffusion Project have included small componentsassessing the use of computing andcommunications technologies by aeropacestudents, faculty, researchers, and engineers.Beuschel and Kling conducted a case study of CIMin an aerospace firm37 and found that effectivetechnological integration was limited by complexsocial requirements for group coordinationprocesses, such as negotiation and interpretation.

These reports reveal that a number ofengineering organizations, including those inaerospace, are using electronic networks for avariety of communication activities, distributed

3

computing, and shared access to information

resources. Networks are being implemented to

serve organizational goals and business strategies,

i.e., to achieve impacts in terms of better and faster

product development and cost savings. Themotivations for network investments noted in these

reports suggest factors that may encourage

network use in particular engineering organizationsand obviate the need for them in others. These

reports also point to a number of factors that mayhinder network use, such as security and

proprietary concerns, the inability of networks toaccommodate the negotiation and interpretation

aspects of communication, and the substantialfinancial outlays required to implement networked

systems.

III. A User-Based Study of Computer Networking

in the Aerospace Industry: Method

This section describes briefly the method of

the study whose results are reported here. Asnoted above, data to answer the study's research

questions were gathered from a wide variety of

aerospace engineers and the study sought

specifically to collect data that reported network

use (1) from the user's point of view, and (2) fromwithin the context of aerospace engineering work

and communication. The study drew upon

methodological approaches and techniques thathave evolved in the fields of library and information

science, communications, management, computer

science, and sociology. 23,25,28,29,38-42 Because it

is user-based, the study aimed to collect data

directly from individual aerospace engineers on

networking topics and issues that were specifically

related to their personal experiences andconcerns. Understanding relationships among

network use, work, and communication will be

useful to those people and organizations trying to

estimate the potential impact of electronic networks

on aerospace engineers, on their organizations,and on national productivity and competitiveness in

the aerospace industry. Further, the results should

be suggestive of the potential impact of networkson other kinds of work, based on the degree to

which they resemble aerospace engineering work.It is the aim of this research to identify work

characteristics and needs that underlie the use of

networks. This type of user-based research on

information and communication technology is

important because it not only evaluates the status

quo, it points to networking system features,

implementation strategies, and use policies that

could improve the effectiveness of the next

generation of networked systems.

The primary mechanism for gathering data wasa national mail survey, conducted in Spring 1993.

The mail survey was preceded by site visits and in-

depth interviews and a national telephone survey.

These preliminary activities were used to refine the

mail survey instrument, to supply anecdotal and

interpretive data not easily gathered in a mail

survey, and to triangulate study results. This paper

will present results from the mail survey only.

The mail survey°s respondents came from a

stratified, random sample of 2000 U.S. subscribers

to Aerospace Enaineerin_q, a weekly trade

magazine published by the Society of Automotive

Engineers (SAE), whose membership includesboth automotive and aerospace engineers. The

database containing records for the 54,600 journalsubscribers is maintained by SAE, but subscribers

are not required to be SAE members. The

database categorizes individuals according to the

aerospace industry they represent (manufacturing,

government, air transportation, suppliers, and

services -- including consultants, R&D services,

and education) and their self-identified job

classification (corporate management, engineering

management, engineers and designers, R&D,

manufacturing and production, purchasing and

marketing, and other). The database includes

practicing aerospace engineers working on a broadrange of aerospace products, in a wide variety of

organizations and subfields, and with a variety of

professional duties. The SAE sample possesses

characteristics in proportions that are similar to

those reported in NSF employment data on the

aerospace industry as a whole. The final,

unadjusted response rate for the mail survey wasabout 48%, with 950 usable surveys returned.

The mail survey consisted of a ten-page

booklet containing items on network availability and

use, work and communication characteristics and

activities, perceived network impacts, and

demographic and employment characteristics ofrespondents. Most questions required

4

respondents to circle the number of their selectedanswer or to fill in a matrix by placingcheck marks incells corresponding to their answers. Severalquestions called for respondents to supplynumerical answers or open-ended textual replies.

The mail survey's results are presented herewith simple descriptive summaries. Most surveyrespondents were engaged primarily in design orproduct engineering (23%), advanced or applieddevelopment (14%) or research (13%) and wereemployed in industry (54%) or government (30%)settings. Other characteristics of surveyrespondents appear below (figures represent % ofrespondents):

Gender

Male 97Female 3

Age

20-29 yrs. 330-39 2440-49 2450-59 3260+ 17

Size of Parent Organization (if private sectorbusiness)

1-4 employees 1050-99 3100-499 13500-999 61000-4999 215000-9995 109996+ 37

Job Type (self-identified)

Engineer 46Manager 39Scientist 5Other 10

Branch of Aerospace (self-identified)

Aerodynamics 6Structures 12Propulsion 9Flight Dynamics & Control 5Avionics 12Materials & Processes 14Other 42

Primary Job Function (self-identified)

Administration 10Research 13Advanced or Applied Dev. 14Design or Product Engr. 23Industrial Engr. 6Quality Control 6Production 1Sales or Service 7Information Processing 3Teaching 5Other 12

In general, survey results paint a picture ofwidespread use of electronic networks. Themajority of respondents (74%) reported that theypersonally used networks, while 11% usednetworks through some kind of intermediary, suchas a secretary or a librarian. Only 15% declared thatthey never used any kind of computer network(from linked workstations within an organization, toa personal computer connected to a printer downthe hall or a supercomputer across the country, to adial-up link to the Internet) in their work. Indescribing the extent of computer networking attheir workplace, 40% of respondents reported that"Networksare used by most people; many tools areavailable on networks; most computer systems arelinked together by a network; and network use isrequired or strongly encouraged." A slightly higherproportion (48%) characterized the extent ofnetworking at their workplace as use by "some"people, and only 7% reported use by "few" peoplewith "little"organizational encouragement or evendiscouragement of network use.

Respondents also reported on availability anduse of different types of networks (see Table 1).Computers connected to commercial networks thatlink users to people, tools, or information outside of

TYPE OF NETWORK

A computer orterminal

connecled

to such a network

isAVAILABLE

for my use

85

Available

networkisUSED

Local 91

Organizational 7 4 8 9

External/Research 5 0 8 8

External/Corn mercial 3 0 8 5

Table 1. Network Availability and Use(% of Respondents Representing Each

Response)

their own organization --such as Compuserve --were available to about 30% of respondents; 50%had access to an exlernal research network such asthe Internet; 74% reported that they wereconnected to an organizational network that linkedthem to resources beyond one workplace building;and 85% reported access to a local area network.Respondents were about equally likely to use anytype of network available to them. Between 85%and 91% of respondents reportedly used eachtype of available network. As Table 2 indicates, theoverwhelming majority of respondents usedcomputer networks at work as opposed to at homeor at some other location; of the various types ofnetworks, external/commercial networks were, notsurprisingly, most likely to be used at home.

Work Home Other

TYPE OF NETWORK

!Local 8 4 1 0 4

OrganizalJonal 7 6 1 1 3

Extemal/Research 5 2 S 2

External/Commercial 2 8 1 9 2

Table 2. Location of Network Use(% of Respondents Selecting Each Location)

The mail questionnaire also askedrespondents to describe the availability, use, andperceived value of various types of computernetwork applications (see Table 3). File transferwas the computer network application reportedlyavailable to the greatest percent of respondents(85%), followed by electronic mail (82%),accessing remote data files (82%), remote Iogin torun a computer program (80%), and electronicbulletin boards or conferencing systems (77%).

Y. W'_ conside% Who say %

VALUE olthat

AVAILABLE (H avalll_le) "_F.Id'orAPPUCAllONS ",ore*"

iE-._ 82 84 83

iBBs, mail lists, conleren¢ing 77 70 67

_eal-tirne inte,actrve messa_lir_ 70 51 54

V_leoconferencin_l 66 44 S B

Voice mail 77 7B 76

Fax 94 96 g4

Eiectronic journals 61 41 50

E:_' 61 23 42

Run p_o_r_ on remote computer 80 71 73

_ccess data on remote,computer 82 72 75

Search l_ovl,commercial dalabatm 66 49 69

Card catalo_ search 62 57 62

Operate remote devices S2 27 43

C_ 63 24 41

Transfer data between computers 85 81 81

Access images 74 56 S9

Other 69 50 52

Table 3. Network Applications

These applications were also the most likely to beused. Less available were applications thatsupported access to published literature, such aselectronic journals or newsletters (61%) or onlinelibrary catalog searching (62%). It should be notedthat these responses indicate a lack of 0erceivedavailability;some aerospace engineers may simplynot be aware that certain applications are available.As a point of general comparison, 94% ofrespondentsindicated that fax was available in theirworkplace, and 77% reported the availability oftelephone voice mail. The percent of respondentsconsidering the value of each application to be"great" or "some" varied from 83% for electronicmail to a low of 41% for computer-integratedmanufacturing (CIM). Throughout the survey, value

judgmentsweremade by all respondents, whetheror not they currently had access to or used thenetwork item in question. Overall value judgments,in this particular instance, may be colored bywhether or not a specific application is used by alarge number of respondents, even thoughrespondents were also given the answer option of"Application is NOT APPLICABLE to My Work."For example, CIM may be assessed by a smallerpercent of respondents as valuable to their work,because it is directly applicable to the work of arelatively smaller number of aerospace engineers.

Tables 4-5 report the availability, use, andvalue of network access to various work resources

in aerospace engineering. In describing networkaccess to human resources (Table 4), morerespondents (about 85%) were able tocommunicate electronically with people within theirown organization than with people in otherorganizations. Private sector colleagues orassociates were least likely to be accessible overthe network, with between 61% and 66% ofrespondents reporting such access. Networkaccess to people in other departments of one'sorganization was judged valuable by the greatestnumber of respondents (81%), while access toexternal colleagues, customers, vendors, etc., wasapparently considered slightly less important. Thismay reflect the feeling -- accepted as common

knowledge by observers of the engineeringenterprise -- that internal communication of anykind is generally more critical in engineering workthan is external communication. On the other hand,the number of aerospace engineers who do usenetworks to communicate with various kinds ofpeople outside their own organizations (between52% and 72%) may surprise those who thoughtthat such links, at least in the private sector, werestill largely prohibited due to proprietary andsecurity concerns.

% Who conslde%WHh %

VALUENet

o( acctss IsAO_ Nel

WORK RESOURCESUSED IoResource Acces| "great'or"ioffte •

Document ¢ilations, abstracll 69 76 74

Journal, trade ma_tzina articles 55 50 63

Equipment / procedures manuals 59 57 $ 2

Intema.I technic al reports 66 71 72

Compan ), newsleners, bulletins 70 7S 61

Suppliers' catalO_l 52 34 $ I

Coclesof stand&rds&pra=ce= S6 s 7 e3Oireclorms of people 73 79 72

Trainir_ material, tools, pro1_raml, 67 87 69

Irlernal financial dall 71 73 70

Prodt_ction cOntrol data 70 69 $4

Experimental data 66 73 76

Proclucl, maleriaJ characteristcs 60 81 71

Technical specifmalions 62 69 79

Design change forms 61 58 61

Lab nolebOOkS 50 33 47

Drawin_ls and designs 71 74 79

3omputer code / pro_tart_ 77 62 79

_her 61 56 78

WORK RESOURCESUSED

People in your workgroup or ¢k_.

,Other peo_ole in your organization

Colelgues in academia, governmenl

Colleagues in privale m¢lustry

External clle_s, customers.

sponsors

E,xten'ml vendors, _ppli4_s

Other

%tVno¢on=l<_"

%IM1h Net%U_NO NIt VALUE

ACCESS ID_e of acc_4 as

RN,ource"grul"

•iom, e°

65 86 78

86 09 81

70 72 66

S6 62 62

82 58 66

61 52 83

46 22 42

Table 4. Work Resources and NetworkUse: People

Table 5. Work Resources and NetworkUse: Information

Network access to information resources(Table 5) ranged from a low of 50% for labnotebooks to a high of 77% for computer codeand programs. Other information resources towhich at least 70% of respondents reportedly hadelectronic access were company newsletters orbulletins, directories of people, internal financialdata, production control data, and drawings ordesigns. Those resources actually accessed vianetworks by at least 70% of respondents weredocument citations and abstracts, internal technicalreports, company newsletters and bulletins,directories of people, internal financial data,experimental data, drawings and designs, andcomputer code and programs. The range of

resourcesheresuggeststhat networkaccesstoinformationsupportsa broad arrayof specificengineeringtasks. Network access to thoseresourcesmostcrucialto the actualdesignandproductionof technologies-- such as technicalspecifications and designs -- was considered of"great" or "some" value by the greatest number ofrespondents.

Respondents were also asked to report thetwo most significant communication channels theyused to perform an important work task. They couldeither choose one of the twenty-one work taskspresented in a list, or supply a task not listed. Thetasks selected by the greatest number ofrespondents were:

• Identify requirements• Conduct experiment or run test• Interpret results of experiments, tests• Produce drawings, designs• Assure conformance with requirements• Plan tasks, projects, programs, etc.• Coordinate work

• Negotiate with co-workers, clients, vendors,students, etc..

• Solve technical problem• Write proposal, report, paper, etc.

Figure 1 portrays the extent to which differentcommunication channels were used in taskperformance, regardless of which task wasreported. Face to face communication was used bya clear majority of respondents (69%), followed bythe examination of printed material (37%) and useof the telephone (36%). Use of a computernetwork link to people, information, or a computerwas greater than that of either voice mail or U.S. orcompany mail service. In examining the use ofnetwork channels for specific tasks, "Learning howto do something" was the one task that accountedfor substantial use of all three kinds of networkchannels. Network links to information were alsoused most heavily for producing drawings ordesigns and identifying problems. Network links topeople were also used most extensively to supportwork coordination and writing proposals andreports. Finally, network links to computers werealso used to develop theories and concepts orproduce drawings or designs.

Survey results discussed so far address extentof network use in the aerospace industry and theuse of networks to support aerospace engineeringwork and communication tasks. Another aim of thestudy was to explore factors that might beassociated with network use. One questionnaire

100

90

SO69

7O

.oiiiiiiiiii!iii5O

,o ;iiiiii!i!ii!i3,30

14 17

zo _!_i!i_ii 11 lz,o ,, Ji i iiiiiiiilliiiii!iiiii!0

36

19

FACETO PRINT DIRECT NET: NET: NET: COMPUTERPHONE VOICE MAIL FAXFACE EXAM PEOPLE INFO COMPUTER MAJL

CHANNELUSED(tl of Respondents)

Figure1. Useof Networks,Comparedto OtherChannels,For PerformingWorkTasks

matrixaskedrespondentsto reporttheextenttowhichtheyagreedor disagreedwitha numberofstatementsdescribingtheirworkandnetworkingenvironments.Comparingthe responses ofnetwork users to nonusersreveals possiblerelationships between network use and variousfactors (see Table 6).

For example, a greater percent of networkusers, compared to non-users, agreed that theirwork is integrated with the work of others, that thepeople they need to communicatewith are all in

their building, that they require a diverse range ofinformationfrom a wide variety of sources, and thattime pressures in their work are tremendous. Agreater percent of network nonusers agreed thatthey spent their day working independently. Theaccessibility of a networked computer is stronglyassociated with network use, as is work output thatis stored in computerized form; these arefrequently cited in the literature as factors thatencourage network use, but they may also reflectresults of extensive network use. Organizationalreward and external demand seem to be significant

FACTORSThe'results of my work are integrated with the work of others

spend my day working independently

All the people I need to communicate with are in my building

require a range of information from a variety of sources

Time pressures ius tTamendous in my work

My work is routine, predictable

Nork discussions require having documents, devices & drawings

examine physical devices, instruments, materials, processes

The products I design, develop, produce are highly complex

work in a field that is extremely compelJtive

M.y.'organization is hierarchically structured (not project-based)

My organizational culture is rigid and authoritative

My work is classified

Results of my work are proprietary

Results of my work are stored in computerized form

started my professional career without networks

% of USERS % of NON-USERS

agreeing with agreeing withstatement statement

89 77

42 63

75 26

84 65

76 59

7 13

67 66

59 62

69 59

69

48

34

22

49

67

59

41

24

21

55

4O

88 84

like to learn new computer things just for the fun of it 65 56

Networking requires too much effort to learn and keep up with 23 16

know about networked information services relevant to my work 19 7

Networking help comas from training or support programs

Network transmission is unreliable

Existing network applications are wall-suited to my work

NI the people, tools, resources I need are on Iha network

Networks are not perfact-many incompatible systems

Networking costs outweigh its benefits

Network usa is encouraged, rewarded by my organization

Lack of experience makes it hard to predict costs/benefits

J, networked computer is easily accessible to me

Customers, clients, sponsors are demanding that I use networks

25 16

15 5

44 16

16 4

61 21

11 12

35 11

45 36

77 15

20 9

Table 6. Factors Affecting Network Use

factorsinencouragingnetworkuse.Interestingly,morenetworkusersagreedthatnetworkingis notseamlessand that manyincompatiblesystemsexist; nonusers, perhaps, are simply moreoptimisticaboutnetworkcapabilities.

Cross tabulating various respondentcharacteristicswith networkuse (seeTable7)revealed, for the most part, only small differencesuse due to respondent characteristics. Networkuse did not vary greatly by age, except for thoseover sixty, who were much less likely to be networkusers. Network use appears to increase witheducational level. Network use is more extensive inin academia, as opposed to other sectors and ismore widespread in very large organizations.Table8 reports variations in network use according todifferent work characteristics. Scientists appear touse networks more than engineers do. In terms ofprimary job function, network use is most extensiveamong those engaged in teaching, research,advanced or applied development, and industrialengineering. Aerospace engineers working inaerodynamics or flight dynamics are slightlymore

s o.oe rOtARAC/EFISTIC$

__._.._:Male

Femade

&me:20.29

30-3Q40-49

50-59SO+

E_,_,__-*.._n Lewd:

Hi_'t Schoo4TechniCal DeQree

BacheloJ'| Degres

Master'= D*Qreel_.O

Poll Doctorate

Type of Organlzatkm:

Indult_/Mim ufnct u rinaOov_nme_

.4kr.Ademe

Non-_)t

Retired

OIhM

<5O

5O-99

100-490

500-9961000-49Q9

5OOO-9995

NEVI_

USE i_:NetworKs Netwodcl.

85 15

81 19

e9 i_93 792 8 __

86 1461 39

80 20

69 3183 17

ee 12

93 7

tO0 0

85 15

92 8 __96 2

63 17

12 8860 40

58 42

53 47

61 19

87 1387 13

95 5

94 6

Table 7. Respondent Characteristics andNetwork Use

CHARACllERISTICS

Job Title:

Enoinem

Mar_w_11

OtherJob Funcllon:

Aclrnin_raI_n

Reesaroh

Advanced/Applied Developme_De_arVPtodu¢l Er_neerIndust hal Eng*n*er

QuaJJly ControlProduction

Sales or MarketingSerwce O( Maintenance

Intorma2ion ProcessingTeaching

/_ospoc. l_mch:

., AerodynamicsStructurss

Propulsion

Fl_hl DynamosAv_

Male_aJe

U_E UWNetworKe N*two_l

84 16

87 1391 9

84 15

80 20

94 691 9

81 1991 9

65 15

80 2073 2775 25

88 12

98 2

94

85 1585 15

90 10

55 1583 17

Table 8. Work Characteristics and NetworkUse

likely to use networks than are those in otherbranches of aerospace.

The final aspect of networking considered inthis study was its impact on aerospace engineeringwork and communication. The percent ofrespondents selecting various replies to thequestion "Overall, how would you describe yourcurrent reaction to computer networks/" ispresented below:

• They have revolutionized aerospace work(21%)

• They are very useful in many respects (55%)• They have certain worthwhile uses (19%)• I am neutral or indifferentto them (4%)• I have reservations about their value (1%)• They have limited value and can cause serious

problems (.4%)• They are worthless and should not be

implemented (0%)

Thus, the overwhelming majority of aerospaceengineers surveyed perceived a very positiveimpact from networks.

The survey also solicited aerospace engineers'

10

assessmentsofspecificnetworkingimpacts.Inonequestionnairematrix,respondentsfirst indicatedwhethertheythoughtnetworksdecreasedgreatly,decreasedsomewhat,hadnoeffecton, increasedsomewhat,orincreasedgreatlyeachoftheaspectsof work and communicationlisted.Theythenindicatedwhethertheyconsideredthe perceivednetworkingeffectto bea majorproblem,a majorbenefit,or neither.Table 9 presentsselectedresultsfromthissectionofthesurvey.Responsesrelatedto degreeof increaseor decreaseweregrouped,andonlythelargerofthetwo resulting

values is reported in the table. Results appear indescending order, with the effects perceived bythe greatest percent of respondents listed first.The table also shows the percent of respondentswho felt that each network effect represented amajor problem or benefit in aerospace work. Overhalf of the respondents felt that major benefits ofnetworks were that they increased:

ASPECTS OF WORKAND COMMUNICATION

% Reporting Network Effect is to:

OECREASE I_CI:EASE

,n_ Effect is:

MAJOR BENERT

Amounl of information available 87 76

74 72

70 64

65 64

65 65

64

64

Exchange of information, ideas across organizational boundaries

Efficiency of contacting people

Ability to complete proiacts, on schedule

Responsiveness to customers, clients, etc.

Ability to stay on the cutting edge of new knowledge

Documentalion, evaluation of work processes

Ability to communicate with otherwise inaccessible people

Use of expensive cqmputers & devices

Ability to express ideas at point of need

Need for face-to-face interaction

61

60

63 62

62 28

60 57

55 34

Performance of work at home. on the road, off-site 53 51

Management control 53 49

Feasibility. size of collaborative efforts 53 51

53 48Flexibility in work structures, patterns

Coherence with one's work community

Duplication of effort

JJt_ility to complete projects within budget

Turnaround time on solving problems

Major system security problems

_mount of time spent fooling around

Leaks of proprietary or sensitive information

:Number or changes required in final products

OeQree of status among one's peers

Sense of ownership, committment to work product

Rate of career advancement

52

47

47

43 45

29

41

52

32

43

38

45

48

Communication with people NOT on the network

Number of etaff employed

46

70

42

30 21

29 27

24 22

22 22

22 19

Table 9. Network Impacts

11

• Theamountof informationavailable• The exchange of information and ideas across

organizational boundaries• The efficiency of contacting people• The ability to complete projects on schedule• Responsiveness to customers, clients, etc.• The ability to stay on the cuffing edge of new

knowledge• The documentation, evaluation of work

processes• The ability to communicate with otherwise

inaccessible people• The ability to express ideas at pointof need• The performance of work at home, on the road,

off-site

• The feasibility and size of collaborative efforts• The turn-around time on solving problems

Citing the JD£,f.93.S._ turnaround time in solvingproblems as a major benefit seemscounterintuitive, if one assumes that it is alwaysadvantageous to solve problems as quickly aspossible. It may be that that some respondentshad difficulty with the "decrease/increase" scaleused in that question, applying it rather as thedegree of "bad" to "good" infl_Jenceof networks.Another possible explanation is that somerespondents felt that networks allowed for moreinput into the problem-solving process, whichincreased the time required to arrive at a solution,but also improved the quality of the solution.

Of the major problems cited, the risks ofsystem security and leaks of proprietary informationwere perceived by over 40% of respondents.Almost a third of aerospace engineers surveyedfelt it was a major problem that networks increasedthe time that people spent "fooling around," whileabout a fifth cited the problem that communicationwith nonusers of networks was reduced.

A number of these impacts, such as "increasesthe amount of information available" are generic inthe sense that they may be felt as well by othertypes of users beyond those in the engineeringcommunity. Some of the reported impacts relatedirectly to efficiency or effectiveness gains.Others, such as the increased "coherence withone's work community" describe second ordereffects, which are also important within the generalwork context.

Few studies have appeared that examinenetworking in engineering, as opposed to scientificor scholarly work, or that relate electroniccommunication determinants and effects to thesituations and environments of particularcommunities of users. The current study hopes toextend existing knowledge by employing a user-based approach to explore the role of electronicnetworks in engineering work and communication.

This paper has reported selected data on theuse of electronic networks in engineeringenvironments. Networks appear to be used widelyfor both communication and computation purposesby engineers in the aerospace industry, withinterorganizational links available to half of thosesurveyed. Nonetheless, respondents perceivedinternal electronic links as being more valuable thanexternal communication capabilities. A significantnumber of respondents reported their networkaccess to a variety of tools and resources andjudged network access highly valuable for a varietyof resource types, from analytical tools likecomputer programs to experimental data toliterature citations and abstracts. While computernetworksare apparently not as important as face-to-face, telephone, and print channels in the conductof aerospace engineering work, they were usedmore often than voice mail or regular mail services,and almost as often as fax. Electronic mail and filetransfer are the applications that are most available,most used, and judged most valuable.

While organizational sector and size andprimary job function appear to influence networkuse, other demographraphic characteristics ofrespondents do not, generally, seem todifferentiate network users from nonusers as wellas specific job and organizational environmentcharacteristics, such as the accessibility ofnetworked computers, whether network use isrewarded by one's organization, or whether onerequires a wide range of information to performone's job. Lack of network training and awarenesswere noted by both network users and nonusers;this may be one area that organizations could targetif they wish to increase network use by theiremployees.

12

The impactof computer networks on theaerospace industry has apparently beenoverwhelmingly positive, with respondentsgenerally reporting gains in certain areas of workefficiency, effectiveness, and satisfaction. Anumber of significant problems were alsoperceived, including lack of ubiquitousconnections and inadequate security controls.

In addition to the questionnaire findings,comments made by study respondenls in in-depthinterviews suggest some of the limitations andadvantages of electronic communication inengineering work. Although electroniccommunication is perceived to contribute toengineering efficiency and effectiveness, its use islimited (at least in terms of today's technology) byengineers' need for immediate, highly interactivediscussion of complex problems of both a technicaland non-technical nature. Networks do not provideadequate means to convey the multi-faceted,multimedia information that is typically exchangedin those situations where, for example, engineersdiscuss issues and negotiate while simultaneouslyconsulting drawings, contracts, financial data, testresults, and physical devices. Use may also belimited by organizations' lack of experience withelectronic communication: while dangers are easyto imagine and costs easy to tallyl benefits areharder to predict and quantify.

Research conducted from a user perspectivecan be utilized by network policy makers, systemdesigners, and service providers in a number ofways. It can help them:

Anticipate and avoid conflicts by discoveringwhere attitudes and expectations vary amongdifferent groups;Understand and estimate networking impactsand benefits by revealing both direct andsecond order effects;Develop products and services well-suited tocustomer/client needs;Choose appropriate network designs andfeatures to meet users' real needs;Devise strategies to promote network use;Develop appropriate management and usepolicies;Implement effective mechanisms for usertraining and support by finding out who is

having what kind of problem;Prepare appropriate evaluations of networksystems and services by identifying a variety ofgoals and objectives and assessing thedegree to which they have been met.

Thus, user-based research offers an importantcomplement to networking investigations thatconcentrate on technical and financial analyses andcan help assure that networking goals will beoptimally met.

References

(I) J.W. Lewis, et al., The CE Testbed:

Methodoloqy and Tools for DevelopinqConcurrent Enqineering Systems. GECorporate Research and DevelopmentReport 90CRD025. (Schenectady, NY:General Electric, 1990).

(2) Timothy L. Dirt and R. Gordon Stockdale,"Distributed Computing at 3M," Computer-Aided Enqineerinq, 8 (no. 6, 1989) 46-54.

(3) Sandra Heiler and Arnon Rosenthal,"Engineering Databases, Tools, andManagement: An Integration Framework." InDiqest of Papers, COMPCON 89: IntellectualLevera,qe, 34th IEEE Computer SocietyInternational Conference (Washington, DC:IEEE Computer Society Press, 1989) 431-437.

(4) Peter G. W. Keen, Competin.q in Time:Usinq Telecommunications for CompetitiveAdvantaqe (Cambridge, MA: Ballinger,1986).

(5) Gerardo L. Lastra, Jose L. Encarnacao, andAristides A. G. Requicha, Applications ofComputers to En.qineerin.q Desiqn, Manu-facturin,q, and Manaqement: Proceedinqs ofthe IFIP TC 5 Conference on CAD/CAM

(6)

(7)

Technology Transfer, Mexico City Mexico,22-26 August, 1988. (New York, NY: North-Holland, 1989).

Henry C. Mishkoff, "The Network NationEmerges," Mana,qement Review, 75 (no. 8,1986) 29-31.

Bernard I. Rachowitz, et al., "UsingWorkstations Efficiently: DistributedComputing Power with Workstations Paves

13

(8)

(9)

the Way for Concurrent Engineering with High

Productivity," IEEE Spectrum, 28 (no. 4,1991) 38-39, 48, 68.

Willie Schatz, "Making CIM Work," Datamation

(Dec. 1, 1988) 18-21.

Constance C. Gould and Karla Pearce,Information Needs in the Sciences: An

Assessment (Mountain View, CA: Research

Libraries Group, 1991 ).

(10) Elizabeth N. Mailloux, "Engineering

Information Systems," In Martha E. Williams

(Ed.), Annual Review of Information Science

and technoloqy, 24 (Amsterdam: Elsevier

Science Publishers, 1989) 239-266.

(11) Beverly A. Beckert, "Technical Office Tools:

Communication and Management ToolsStreamline the Design Office," Computer-

Aided Engi_need.ng, 9 (no. 12, 1990) 68-70.

(12) John K. Borchardt, "Improve In-House

Communications," Chemical Engineering, 97(no. 2, 1990) 135-138.

(13) Arnoud C. L. De Meyer, "Tech Talk: How

Managers are Stimulating Global R&D

Communication," Sloan Management

Review, 2 (no. 3, 1991) 49-58.

(14) Chandler Harrison Stevens, "Electronic

Organization & Expert Networks: Beyond

Electronic Mail and Computer Conferencing."

In Proceedings, 1987 IEEE Conference on

Manaqement and Technology: Managementof Evolving Systems (New York, NY: IEEE,

1987) 360-370.

(15) Tekla S. Perry and John A. Adam, "E-Mail,

Pervasive and Persuasive: Special Report on

Electronic Mail," IEEE Spectrum, 29 (Oct.1992) 22-33.

(16) Tora K. Bikson and J. D. Eveland, "The

Interplay of Work Group Structures and

Computer Support," In Jolene Galegher,

Robert E. Kraut, and Carmen Egido (Eds.),Intellectual Teamwork: Social and Technical

(17)

Foundations of Cooperative Work (Hillsdale,

NJ: Lawrence Erlbaum, 1990) 245-290.

Starr Roxanne Hiltz, "Productivity

Enhancement from Computer-Mediated

Communication: A Systems ContingencyApproach," Communications of the ACM, 31

(18)

(no. 12, 1988) 1438-1454.

Ronald E. Rice, "The Impacts of Computer-

Mediated Organization and Interpersonal

Communication," In Marlha E. Williams (Ed.),Annual Review of Information Science and

Technoloclv, 15 (New York, NY: KnowledgeIndustry, 1980) 221-249.

(19) Charles W. Steinfield, "Computer-Mediated

Communication in an Organizational Setting:Explaining Task-Related and Socioemotional

Uses. In Communication Yearbook., 9 (NewBrunswick, NJ: International Communication

Association, 1986) 777-804.

(20) Charles W. Steinfield, "Computer-Mediated

Communication Systems." In Martha E.

Williams (Ed.), Annual Review of Information

Science and Technoloqv, 21 (New York, NY:

Knowledge Industry, 1986 ) 167-202.

[21] Lee Sproull and Sara Kiesler, Connections:

New Ways of Working in the Networked

Organization (Cambridge, MA: MIT Press,

1991).

(22) Ronald E. Rice and Douglas E. Shook,

"Relationships of Job Categories andOrganizational Levels to Use of

Communication Channels, IncludingElectronic Mail: A Meta-Analysis and

Extension," Journal of Manaqement Studies,27 (March 1990) 195-229.

(23) Elizabeth B. Bizot, Nancy D. Smith, andThomas Hill, "Use of Electronic Mail in a

Research and Development Organization,"

In Advances in the Implementation and

Impact of Computer Systems, 1 (Greenwich,CT: JAI Press, 1991) 65-92.

(24) J.D. Eveland and Tora K. Bikson,"EvolvingElectronic Communication Networks: An

Empirical Assessment," Office: Technology

and People, 3 (1987) 103-128.

(25) Martha S. Feldman, "Electronic Mail and

Weak Ties in Organizations," Office:

Technology and People, 3 (1987)83-101.

(26) Humberlo Gerola and Ralph E. Gomory,

"Computers in Science and Technology:

Early Indications," Science, 225 (no. 4657,

June 6, 1984) 11-18.

14

(27) Bradford Hesse, Lee Sproull, Sara Kiesler,and John P. Walsh. "Returns to Science:

Computer Networks in Oceanography,"

Communications of the ACM 36 (no. 8, 1993)90-101.

(28) Starr Roxanno Hiltz, Online Communities: A

Case Study of the Office of the Future

(Norwood, NJ: Ablex, 1984).

(29) Charles R. McClure, Ann P. Bishop, Philip

Doty, and Howard Rosenbaum, The NationalResearch and Education Network (NREN):

Research and Policy Perspectives (Norwood,

NJ: Ablex, 1991).

(30) Bruce R. Schatz, "Building an ElectronicCommunity System," Journal of Mana.qement

Information S_, 8 (no. 3, 1992) 87-107.

(31) Lee Sproull and Sara Kiesler, "ReducingSocial Context Cues: Electronic Mail in

Organizational Communication,"

Manaqement Science_, 32 (1986) 1492-1512.

(32) Davis A. Foulger, Medium as Process: TheStructure_.._U_s_e_and Practice.of Computer

Conferencing on IBMPC Computer

Conferencinq Facility. Draft of unpublished

doctoral dissertation. (Philadelphia, PA:

Temple University, 1990).

(33) K. Hall. "Information Technology

Applications: A British Aerospace Military

Aircraft Ltd View," In North Atlantic Treaty

Organization. Advisory Group for Aerospace

Research and Development. ElectronicTransfer of Information and Ils Impact on

(34)

(35)

Aerospace and Defence Research and

Development. AGARD Conference

Proceedings No. 466 (16-1--16-16) ([S.I.]:

AGARD, 1990).

Bernard I. Rachowitz, et al. "Using

Workstations Efficiently: Distributed

Computing Power with Workstations Paves

the Way for Concurrent Engineering with High

Productivity," IEEE Spectrum, 28 (no. 4,1991), 38-39, 48, 66.

John B. Black. "Computer Conferencing:

Minds Meeting Anywhere/Anytime." In North

AtlanticTreaty Organization. Advisory Group

for Aerospace Research and Development.

(36)

Electronic Transfer of Information and Its

Impact on Aerospace and Defence Research

and Development. AGARD Conference

Proceedings No. 466 ([S.I.]: AGARD, 1990)13-1--13-4

Kurt N. Moholm. "Applications to the

Aerospace and Defense R&D Community

[of] the DoD Computer-Aided Acquisition

and Logistics Support (CALS) Initiative," In

North Atlantic Treaty Organization. AdvisoryGroup for Aerospace Research and

Development. Electronic Transfer of

Information and IIs Impact on Aerospace and

Defence Research and Development.

AGARD Conference Proceedings No. 466

([S.I.]: AGARD, 1990) 14-1--14-8.

(37) Werner Beuschel and Rob Kling.

"Coordination Processes and the Integration

of Manufaturing Domains Through

Computerization." Manuscript, 1993.

[38] Brenda Dervin and Michael S. Nilan,"Information Needs and Uses," In Martha E.

Williams (Ed.), Annual Review of Information

Science and Technoloqy, 21 (White Plains,

NY: Knowledge Industry Publications,1986)3-33.

[39] Use of the Critical Incident Technique to

[40)

[41)

(42)

Evaluate the Impact of MEDLINE: Final

Re______.AIR-64600-9/89-FR (Pale Alto, CA:

American Institutes for Research, 1989).

D. Wixon et al. "Contextual Design: An

Emergent View of System Design," In CH..__[I'90 Conference Proceediqns: Human

Factors in Computin.q Systems, Empowerinq

People, (Seattle, WA: ACM, 1990) 329-336.

J. D. Gould et at., "Making Usable, Useful,

Productivity-Enhancing Computer

Applications," Communications of the ACM

34 (no. 1, 1991) 75-85.

David K. Murotake. A Double-Edqed Sword.

Relationships Between the Engineerinq Useof Computer Tools and Project Performance.

Doctoral dissertation, Sloan School of

Management. (Cambridge, MA: MIT, 1990).

15