From: Proceedings of the Eleventh International FLAIRS Conference. Copyright © 1998, AAAI (www.aaai.org). All rights reserved.
E1-Tech: A Performance Support System with Embedded
Training for Electronics Technicians
Alberto Cafias, John Coffey, ThomasReichherzer, Greg
Hill, Niranjan Suri, Roger Carff, Tim Mitrovich and
Derek Eberle
Institute
for Humanand Machine Cognition
University of West Horida
Pensacola FL 32514
Abstract
The integration of modern computer-based inferencing,
networking, computer-mediated training and multimedia
technologyholdthe promiseof improved
training and performance
support for personnelin industry and government.This paper
reports an effort to developa network-enabled
knowledge
editing
and modelingenvironment
that offers just-in-time training and
performancesupport, anytime, anywhere.The instructional
componentof the system is based upon a knowledgemodel
elicited froma domainexpertand representedin a pedagngicallymotivatedmultimediabrowser.Theexpert’s diagnostic processes
are mappedinto an integrated rule-based advisory systemthat
assists the user in task performance.The advisory systemis
tightly coupled with the knowledgemodelin a fashion that
provides context-dependentexplanations of the diagnostic
process. Thedistributed, network-based
designallowsthe system
to be accessed any time, on demand,running seamlessly over
the Intemetor similar networks.EI-Tech(Electronic Technician)
is a first prototypeof this approach
to next-generation,
just-in-time
performancesupport systems with embedded
training. EI-Tech
is being developedas a joint project with the Navy’sChief of
NavalEducationand Training (CNET).
1. Introduction
Thetechnologies and methodsthat underlie current approaches
to computer-mediated
learning, distancelearning, expertsystems,
performance
supportsystems,world-wide-web
browsers,and justin-time training systemsare rapidly converging.Training and
performancesupport that were once presented in hardcopyare
increasinglybeing delivered in computerized
formats. However,
simplyautomatingor computerizing
current training materials
Copyright1998,American
Associationfor Artificial Intelligence
(www.aaai.org).
All rights reserved.
and processes is not enough- qualitative changesto enhance
instruction and job performance
are both neededand feasible.
Systemsthat embody
the knowledge
and reasoningcapabilities
of experts in the performance
of skilled tasks hold the promise
of providingmuchgreater utility than current training programs
provide. The goal, as Wehrenberg
(1989, p 38) aptly expressed
it is "to put the right personin the right placeat the right time
withtherightskills ..."
Bradshaw(1993) discusses integrated performancesupport
systems(IPSS)that are designedto improvethe performance
field-deployedtechniciansby makingtraining and other critical
informationavailable immediatelyas needed.Bradshawstates
that pressures for efficiency will require moretraining to be
deliveredonly as neededon the job, in somedegreesupplanting
the training that is currently offeredin dedicatedprograms.As
an example, it is possible that computer-basedperformance
supportsystemsmightprovidevideo demonstrations
of equipment
checkoutor maintenance
proceduresin a fashion that can provide
just-in-timetrainingfor technicians.
Thispaperreportsan effort to createtools to supportdistributed
knowledgemodelingand inferencing capability. This tool set
allows domainexperts and knowledgeengineers to collaborate
in the development
of performance
supportsystemsto assist the
end user of the systemwithperformance
of a task. Thedeployed
systemwill havethe ability to explainthe conceptual
logic behind
its diagnosticreasoningstrategy. Thisresearch effort extends
workthat started with NUCES,
NuclearCardiologyExpert System
(Ford et al, 1996) by extending the knowledgemodelingand
inferencing componentsto operate seamlesslyover networks,
and by addingadditional supportcapabilities as appropriateto
the applicationdomain.
Thesystemholds the promiseof supportingselected ancillary
services in a morecost-effective wayvia networktechnology.
Sinceit is anticipatedthat suchsystemscanfacilitate just-in-time
training in additionto performance
support, the deployedsystem
will havethe capabilityto trackusers’useof instructionalmaterials,
and to provide other services. The systemhas beenimplemented
ExpertSystems 79
in Java to achieve platform-independence. The first working
prototype of this system, FA-Tech,demonstratesthe feasibility of
this type of cost-effective just-in-time training.
EI-Tech is being constructed in a joint research effort with
the Chief of Naval Education and Training, that is aimed at
advancing and integrating the aforementioned technologies into
a remotely accessible, just-in-time performancesupport system
for electronics technicians. The Navy, at any given time, has
approximately45,000students in training, mainly through courses
taught by approximately6,500 instructors. It is typically the case
that Navyelectronics technicians find themselves in-fleet and
responsible for mission-critical equipmentmonthsafter receiving
rapid-fire, short duration training on the equipment,and after a
long-period of time without any contact with the equipment.
Likewise, regular equipment upgrades and evolving safety
procedures and warnings regarding maintenance and repair of
equipment require continuous refresher courses for the
technicians. With budgets regularly being reduced, the Navyis
looking to distance learning and newtechnologies to provide a
morecost effective delivery of continuouseducation and training.
WWW
Servers
MediaSen~rs
The target audience for EI-Tech is the enlisted electronics
technician. Typically, this technician has had 16 weeksof basic
electronics training followed by as muchas five months of
advanced training on specialized equipment. The advanced
training is especially fast-paced and covers a lot of equipment.
As stated above, very often it is monthsafter the courses end
before technicians finds themselves responsible for the
maintenance of the equipment for which they were trained. The
role of EI-Techwill be to assist these electronics technicians in
their job performance.
2. Computer-based
Systems
Training
Model 8e~’er=
KB Sewers
¯..,.....
EndUsers
Figure1. SystemArchitecture
Cafias
Support
Collis and Verwijs (1995) present a survey of software that
originally fell outside the educational process, but that are now
being applied increasingly to educational situations, and that are
characterized as hybrid electronic performancesupport systems.
These products are grouped into three categories: support for
groupfunctioning, information retrievai and handling assistance,
Pedagogicaily.motivated
browse#editor
80
and Performance
Expert
and guidanceand informationsupportthat are integratedinto the
regular workplace.Our workencompasses
elementsof the latter
twocategories.
Computerized
u’aining and performancesupport has evolved
siguificanfly fromthe early computer-based
training programs.
Early workin computerizedtraining and performancesupport
featured lockstep, programmed
instruction that grewout of the
workof B. F. Skinnerand the BehavioralPsychologists(Shrock,
1995). Theseearly systemspresenteda standardcurriculumand
sequence
of instructionfor all, withtesting andfeedbackto assess
the attainmentof pre-specifiedgoals. Suchsystemsmirroredthe
approach (in computer-mediatedform) of the traditional
classroom,and lacked explicit performancesupportcapability.
Theapproachof such early systemshavebeencriticized both for
beingdecontextualized,and for not beinggearedto the needsof
the individual student, and hence,ineffectual in bringingabout
learning.
The current workseeks to avoid deficiencies attributed to
suchsystemsby presentingthe studenta constructivist modelof
a knowledge
domainbaseduponknowledge
that has beenelicited
from an expert in the domain. The knowledgemodelis well
situated in the contextof a specific problemdomain,rather than
in generic first principles. Theperformancesupportcomponent
of the systemassists withthe solutionof contextualized,
real-world
problems.
domainknowledge.The knowledgemodelis stored on network
servers that contain concept maps and other mediating
representations (see Ford, et al, 1991)such as textual media,
audio,video,still and animatedgraphics.Thearchitectureof the
knowledgemodelpermits its components
to reside in a central
repositoryor to be distributedovera network.
The knowledgemodelis accessible from anywhereon the
network. The pedagogically motivated browser can switch
seemlessiyfroma knowledge
modelof one domainto a knowledge
modelof a related domain.Navigationlinks, based uponthe
occurrenceof a given conceptin various conceptmapswithin a
givenknowledge
domain,can be establishedto other suchmodels.
Theknowledge
modelitself adoptsa specifically Constructivist
approachby affording the user a freely browsableknowledge
space which the user mayperuse until satisfied with the
explanation.
3. An Architecture for Performance Support Systems
with Embedded Training
The increasing maturity and integration of network
communications,
multimedia,performancesupport and training
systemsmakespossible a newarchitectural approachto networkbased performancesupport and training systems. Figure #I
presentsa graphicof our systemarchitecturewhichincludes:
(1) a pedagogically-motivated
multimediabrowser
(2) a consultationcomponent
for performancesupport
An integrated developmentenvironmentfor the creation of
browserand consultationcomponent
is part of the system.
3.2. The Consultation Component
Knowledge-based
systems (KBS)in general, and expert systems
in particular, havetraditionally beenusedfor classification or
diagnostic tasks. Ourapproachextends traditional knowledgebasedsystemscapabilities in order to providetraining as needed,
and reference information to support job performance.The
consultationcomponent
acts bothas a consultantprovidingdomain
knowledge
andas a diagnosticfront-endin supportof just-in-time
training.
After the initial knowledge
modelhas beenelicited fromthe
domainexpert, the consultationcomponent
is createdas a system
of rules that mapsto the expert’s diagnostic process. The
consultationpart can present a full complement
of mediain the
formof examplesto assist in the diagnostic interaction. The
consultationcomponent
offers great promiseto mitigatethe need
for access to human
domainexperts. Byintegratingpedagogicallymotivatedbrowserswith knowledge-based
systems, just-in-time
training is providedfor the learner/practitionerwhois not in a
traditional setting - somebody,
for example,whoneedsto learn
howto fix a piece of equipmentwhile on the job actually
conducting
the repair.
3.1. The Pedagogically-Motivated Multimedia Browser
Conceptmaps, developedby Novak(1984), have been designed
to elicit andrepresent part of a person’scognitivestructure by
externalizing concepts and propositions. A concept mapis a
planar representation of a set of concepts that makes
interrelationships amongthemevident. Conceptmapshaveproven
to be useful for knowledge
elicitation (Fordet al, 1991),and
an interface to a modelof a knowledgedomain(Ford, Cafias,
and Coffey, 1993).
The pedagogically motivated browser is based upon a
hierarchically structured representationof a knowledge
domain,
organized by concept maps. This pedagogically-motivated
browserprovides a unique wayof organizinga modelof expert
3.3. The Editing Component
Theediting environment
containstwoparts. Theknowledge
model
editor fosters collaborationbetweenthe expertand the knowledge
engineer to create the multimediaknowledgemodelthat forms
the basis of the system. The knowledgemodelis basedupona
hierarchyof conceptmaps.Theknowledgeengineer has the use
of a modeleditor that allowsfor creationand linking of concept
maps,and for the populationof the mapswithaudio, video, text
and graphics. The modeleditor is an outgrowthof work on
ICONKAT
(Integrated Constructionist Knowledge
Acquisition
Tool) (Ford et al, 1991) and an initial version has been
implemented.
Theconsultationeditor providesbasic editing capability to
ExpertSystems 81
create the rule bases. Anobject oriented approachhas been adopted
in whichthe role object includes attributes such the various media
that mightbe accessed by the user during an interaction involving
that rule, as well as an entry point into the knowledgemodel,
should the user seek greater elaboration of the question at hand
in the consultation. The consultation editor is currently under
development.
Knowledgeelicitation started with an informal discussion
during whichthe three areas were identified, and then proceeded
with the creation of concept mapsto elicit the importantconcerns
of the expert in each area. Aniterative process of refining the
concept mapsand creating and locating support materials such
as schematics, block diagrams, video of the expert, etc, then
ensued.
4. EI-Tech
4.2. The EI-Tech System Interface
The first prototype developedwith this system, E1-Tech,models
the knowledgeof an expert electronics technician on the RD379A(V)/UNH,a fault-tolerant
magnetic recorder/reproducer
manufactured by Magnasync/Moviolacorporation.
Figure 2 presents a graphic of the systeminterface. The leftmost
windowin Figure 2 depicts an interaction with the advisory
componentof the system, in whichthe system is asking the user
about the machine’s performance in "direct mode." The system
Figure 2. The EI-Tech Interface
4.1. Knowledge Elicitation
for EI-Tech
Knowledge elicitation
proceeded along several dimensions
simultaneously. First, it wasdeemeddesirable to create a basic
description of the RD-379A(V)/UNH,that documents its
components, their locations, and how they individually and
collectively operate. Second,in order to support the performance
of deployedelectronics technicians, the routine maintenanceand
checkout procedures had to be identified and described in the
system. The third area of knowledge elicitation
involved
identifying the symptomsof problemsthat might be noted by the
technician, and the diagnostic strategy that might be employed
to repair the problem.
82
Cafias
explanation of the question has been invoked and the "direct
mode", as it pertains to a malfunction has been located in the
"Malfunctions" concept map. The user has opened a window
that gives a textual description of the importanceof direct mode
in the diagnosis of the fault, and the user is currently viewinga
digital video of the expert elaborating on the point.
The malfunction model is closely coupled with a model that
describes the nominal operation of the system so that the user
can easily navigate through a description of a component’snormal
behaviors and failure modes. The knowledge model contains
two hierarchies of concept maps,the first presenting the detailed
description of the equipment, and the other the detailed account
of howto diagnoseand repair faults.
The knowledgemodel includes block diagramsof the equipment,
electronics schematics, photos that showcomponentsand their
locations, textual passages transcribed from interviews with the
expert, digital video that illustrates checkoutand fault isolation
procedures, and references to supplementary information
contained in external sources.
The inference componentof EI-Tech is a rule-based expert
system, with rules culled from the knowledge model elicited
from the expert. The inference componentis implementedusing
JESS (Java Expert System Shell) (Friedman-Hill, 1997) which
allows creation and maintenance of centralized knowledgebases
on a network. The Fault Isolation knowledgebases were first
developedin CJ.,IPS (Riley, 1997). Oursystemincludes a graphical
interface for the consultation session that runs a Java Applet, to
replace the commandline interface of CLIPS. A distributed
collection of servers house these components (the knowledge
model and the inference component)together with all the other
media, makingthe entire system available anywherein-fleet via
a network connection.
Users of the system can seek a simple advisory consultation
in whichthe systemasks the user a series of questions concerning
symptomsto be found in the equipment. The inference component
converges on probable failure points at the componentlevel individual transistors, relays, switches, motors, amplifiers, or even
a wiringfault, typically after the user has answered3 to 6 questions.
At any time the user can ask for an explanation of the line of
reasoning the system is pursuing. Uponsuch a request, the
inference componentwill transport the user into the knowledge
model where the user can view graphics of the schematics and
literal pictures that showtest points of the componentsunder
consideration, oscilloscope settings, waveforms, and values for
whichthe user should look.
5, Summary and Future Work
This project links navigational and knowledge-basedtools into
an integrated, network-enabledsystem suitable for distance, justin-time learning and performance support. This research has led
to the production of a working prototype system (El-Tech) that
assists electronics technicians in their job performance. This
approach to computer-mediatedlearning decouples training from
the schedulingand location constraints of traditional instruction.
The development and deployment of systems like EI-Tech will
makeit possible for electronics technicians to have access to
knowledge when and where it is needed. This approach is
generalizable and is currently being used in another project with
the National Imagery and Mapping Agency (NIMA).
was discovered, length of time to repair, etc. The Navyhas an
extremely comprehensive record of these reports. A retrieval
system for previous failures could prove a useful adjunct to the
current system.It is anticipated that application of this technology
to other knowledge domains, might well reveal other domainspecific ancillary services that can be integrated into the system.
6. References
Bradshaw, J. M., D. Madigan, D., Richards, T. and G. Boy.
(1993). Emerging technology and Concepts for ComputerBased Training, Proceedings of the 6th Florida Artificial
Intelligence Research Symposium,Ft. Lauderdale, FL.
Collis, B. A., and Carla Verwijs. (1995). A Humanapproach
Electronic Performance and Learning Support Systems:
Hybrid EPSSs. Educational Technology. January.
Ford, K. M., Stahl, H., Adams-Webber,
J. R., Caflas, AJ., Novak,
J.D., & Jones, J. C. (1991). ICONKAT:
An integrated constructivist
knowledge acquisition tool. Knowledge
Acquisition Journal, eel 3, pp. 215-236.
Ford, K. M., Caflas, A. J., Coffey, J. (1993). Participatory
Explanation. Proceedings of 6th Annual Florida AI Research
Symposium(pp. 85-90). Ft. Landerdale, FL.
Ford, K. M., Coffey, J., Caflas, AJ., Andrews,E.J., Turner, C.
W. (1996). Diagnosis and Explanation by a Nuclear
Cardiology Expert System, International Journal of Expert
Systems, vol 9, number4, pp. 499-506.
Friedman-Hill, E. (1997). Jess, the Java Expert System Shell.
[online reference] http://herzberg.ca.sandia.gov/jess/
Novak, J. D. & Gowin, D .B. (1984). Learning Howto Learn.
Ithaca, NY:Cot’nell University Press.
Riley, G. 0997). CLIPS, A Tool for Building Expert Systems.
[online reference] http:l/www.ghg.netlclips/CLIPS.html.
Shrock, S. A. 0995). A Brief History of Instructional
Development.Instructional Technology, Past, Present and
Future. 2nd edition. Gary Anglin ed. Libraries Unlimited,
Englewood CO.
Wehrenberg,S. B. (1989). The Future Just-In-Time WorkForce,
PersonnelJournal,. pp. 36 - 44. February.
Future workwill explore the feasibility of incorporating other
formsof reasoning into the system, such as case-based reasoning.
System problems in the Navy trigger an OpNav47902K
trouble
report that details the nature of the problem, howand whenit
Expert Systems
83