 |
|
|
|
| | Site Map | Search | Home | Meeting Information | Online Presentations (for this meeting) | |
Monday January 30, 2006 CMI Subcommittee: Bill McDonald, Alteon
Training, Chairman Bill gave an overview of the CMI Activities including
CMI001 Version 4.0+, Simple Deployment Architecture,
SCORM Convergence, CMI Testing, PENS Specification, and
CMI Offline Specification. Bill has been
investigating methods for developing a more robust test
suite. Tom King has written a test suite for
PENS in PHP. He will talk about that later in the
morning. Simple Deployment Architecture (SDA) Working
Group – Draft Document Review: Bill McDonald,
Alteon Training SDA can fix the “cross-domain” problem.
Allows more content delivery scenarios for content
providers. The SDA group has been formed and there
have been 4 meetings. Bill reviewed the concepts behind the Simple
Deployment Architecture. He gave recap of a
previous presentation “XML and Service Enhancement
to HACP API and more.” He restated the goal of
the effort and highlighted what is important: do not lose
backward compatibility, solve the cross domain issue,
increase configurability, reduce compatibility issues
between content LMS/CMI Bill described the problem
showing several scenarios. The draft specification is in work. Kevin
Schlipper reviewed the draft document. It specifies
how a relay service should be created to work with the
API. He reviewed definitions that are part of the
document: Content Server Listener (CSL), Content
Server Relay (CSR). The document standardizes how
the CSL communicates with the CSR. Jack asked
about the data model. Kevin explained it is the
model for data communication to be XML. Jack
recommended that the XML work he is presenting next may
be usable Kevin the reviewed the Use Cases document
which describes all the possible communication scenarios
which shows were the content, API, and firewalls are.
The working group is Plateau (Ed), EEDO (Kevin
Schlipper), Recombo (Steven Forth), SkillSoft (Mike
McCauley), VTN(George Urich). PENS: Tom King, EEDO Tom discussed changes to the PENS forum detailing
documentation revisions. Tom reviewed the
documentation revisions. Tom updated the test
harness (written by Paul Roberts, QuestionMark) and
described the changes. AI: Tom King, Bill McDonald, Kevin
Schlipper: Tom to update PENS specification (CMI010) with
corrections/clarifications posted on AICC forum (and send
to Bill) Tom to send updated test harness to Bill (one with
corrections, but NOT testing-specific extensions added) Tom to send PHP validation code for PENS server to
Bill (and cc: Kevin) Bill to get PHP support added to "aicc.org" Tom to draft some test scenarios for testing a) PENS
clients and b) PENS servers Bill and Tom to
draft a test procedure (scenarios, data, and document
creation). Make a recommendation to determine
whether the test lab should offer testing. Kevin
Schlipper will assist with this. Tom recommends a
Beta certification stage before the real full cost
testing phase. Review of AICC Document: XML Communication
for CMI: Jack Hyde AICC Jack presented an overview of the agenda. He
discussed the overview of the standard including a
discussion of the communication (file, HACP, API),
Offline Communication, SDA (Simple Deployment
Architecure). The XML will encapsulate the Student
data into a separate document for an XML binding for the
student data. Kevin asked about changing student
preferences and demographics to random order, it means
that all items in preferences and demographics are
optional. Jack asserted that these are optional.
Jack reviewed the draft document version (1.1) and
discussed the changes. Jack showed an obligation
table that shows: data elements, flow
(communication direction), and obligation (AU and CMI) Bill thinks we might define XSD and caveats and just
include in CMI001. Jack said CMI spec is very large
and it may be better to have multiple small
specifications and refer to by CMI. Tom Kings
thinks that separate document makes easier to reference
because old CMI document is easier to use (antidotal
comment to Tom) than the new. Even Bill refers to
old document. Jack reviewed several questions: Details-Number Question: we left of
number in XSD but Jack thinks we need to add the number
which would be used if not all tries reported. Bill
said all tries must be reported. Bill said that
this does not meet the spec. Jack outlined the try
options. Kevin said we might want to clarify the
try order so however your parser works the try
specification shows the correct order. Andrew
Lucking said that beauty of XML doesn’t require
order but that requires that the try number should be an
attribute. After the discussion, it was decided to
leave in the try number. Bill an Kevin vote for
“n”. Bill thinks the rationale for
keeping the try number should be in the spec. Will
use “n” for the try tag and there will be
rationale why the “n” is to be included. Do
not need try number in other items: score, status, etc.
AI: Jack to use “n” for try number and
provide the rational for inclusion. Glossary Question: Jack requested
audience to recommend terms for the glossary besides CMI
& LMS. Jack to make sure any new terms appear
in Glossary. No action taken Order Question: Jack doesn’t think
the data within the groups should be in any order. Bill
doesn’t think there should be unless there were some
reasons for parsing. Tom things that since some are
required and some are optional, there may be a reason to
make some appear before others. Kevin said that
under demographics: city, class, company cannot occur
twice. If would have companies would need method to
show it. When there is a multiplicity: you
need a wrapper to allow provide a means for the
multiplicity. Kevin will change Objectives, Student
Data, Try, Sessions, (within Sessions Journal), Path
(within Path. AI: Kevin made changes to document as the
presentation and discussions occurred. Jack to
review Kevin’s changes and validate before
incorporation into final document.. Do we need a conformance clause in the document.
Bill says we need a few sentences on conformance. AI: Jack to add a conformance clause to
document. Next steps: make changes, create AGR, proof, and
vote. Offline CMI Specification Update: Shirkant
Pattahil, Harbinger Systems Shirkant said the only missing piece to the Offline
spec is the XML. He reviewed the agenda. updates:
DTDs to XSDs, deleted elements. Shirkant an example
of the XSD and how the corresponding XML is represented.
Shirkant noted he added version to session ID. Bill
related problems when adding version number to the CMI
spec and he problems encountered by how vendors handle.
Bill suggested that Shirkant add information on who to
process when encountering a version. AI: Jack and Shirkant will add language
on the intended use of version to assist vendors. XSD Representation – Reponse: Shirkant
discussed Progress: course element ID, dateTime,
data. One progress type node with multiple
progresses what would contain multiple AU sessons. Kevin
thinks we may need to implement name spacing to minimize
conflict. Discussion resulted (Bill, Kevin,
Shirkant) on the progress node. Will there be 2
different types AU input and AU output? Originally
there was just one type not there will be 2 elements data
from LMS and data to LMS. AI: Shirkant: After much discuss it
was that data will be separated into 2 data elements
(child node): from LMS to AU data type and from AU
to LMS data type. Shirkant discussed deleted data element. Shirkant
explained the rational behind deleting the elements.
Tom thinks another use case would be to write to USD
drives (or other disks) and move data from one notebook
to another. The group discussed work done by IEEE
that may have applicability to use. Roadmap to release: Formal review of document at
the I/ITSEC Recap & Highlights: Bill
Shook, Boeing Bill described the purpose IITSEC. He showed the
size of the conference (16500 registrants). Bill
described the themes of the show. Bill will talk
primarily on the training, education, and human
performance information. Bill’s intent to go
to the show is to find out where training and performance
are going over the next few years. Bill talked
about Key themes of courseware: gaming, EPSS, ITS,
Navy ILE. Gaming is getting more important and
prevalent. Bill could not get anyone to give the
instructional value of gaming. Bill described his
interpretation of what gaming is. Kris said that
the best example is AI: Tom King recommends that we have an
S1000D session and speakers for the fall meeting. Anne
work with Tom K and Bill S on an S1000D agenda. AICC-SCORM Differences Document: Jack Hyde,
AICC Jack gave a convergence summary. This is a recap
of a previous presentation. Jack reviewed the AICC,
IEEE, and SCORM documents and what they cover. Jack
showed overlap between the AICC and ADL SCORM in terms of
Conceptual and Technical overlap. Jack says the greater we make the alikeness of the
specifications the better it is for the industry. Jack
says we are adopting the IEEE/SCORM metadata. AICC
has been work on this for a while and is getting close to
a specification. AICC is adopting the SCORM?IMS
content packaging/course description. AICC has
begun an initiative with Giunti. Tom recommends a
hybrid package that contains all the SCORM 2004 and AICC
.ini files. Minimally we should formalize the AICC
zip format and say the references in the AICC ini file
are relative. Ed’s vision is a folder solution
that contains more than the content. Inside the zip
that root folders: on-line content, documentation. No
restriction in SCORM on this. AICC Executive Committee Meeting Scott passed around the agenda: Finances: Scott reviewed the financial summary. Then we
discussed the ways that AICC can economize to reduce
costs. We discussed meeting fee charges, raising
the membership fees, charging for test suites. It
was decided to meet at cheaper hotels and charge a
nominal meeting fee of $50. There was a discussion
on whether to charge for non-AGR publications. Several
ideas were presented. For example, revenue from
downloading documents may be a good idea. Tom King
said we need to quantify some of these ideas. Is
the potential revenue stream worth the cost to provide
the capability? We will change a $50 member internet access/meeting
fee and $100 non-member internet
access/meeting fee per person (not per company) starting
for the Tom King will serve as the new Communications
subcommittee chairman. Elections: Neil Cramer as Chair. Mike Sharp, Treasurer Air Frame: Dave Jacobson, Boeing; Rolen Weeks,
Rockwell Collins Vendor: GianLuca Rolandelli, Giunti; Kevin
Schlipper, EEDO Airline: Kari Itkonen, Finnair, Will Schart, UAL, Ray
Butler, UPS Future Meetings: Other Business: For the TUESDAY January 31, 2006 AICC Chairman’s Report: Bernard
Bouyt, AICC Chairman Bernard discussed the financial situation of the
organization and the steps AICC has discussed to improve
the financial situation. AICC would like to set a
meeting fee for attendees: speakers-no fee,
members-$50, non-members-$100. AICC may charge for
access to documents on the AICC website but that is still
under study and evaluation. Bernard then discussed the elections to be conducted
next. He then discussed future meetins AICC Elections: Scott Bergstrom, AICC Scott conducted the elections. Because there
were not enough nominations to require a ballot, the
election will be just a show of hands to agree to the
current slate. The following are the new officers: Neil Cramer: Chairman Mike Sharp: Secretary/Treasurer Air Frame: Dave Jacobson, Boeing; Rolen Weeks,
Rockwell Collins Vendor-at-Large: GianLuca Rolandelli, Giunti;
Kevin Schlipper, EEDO Airline: Kari Itkonen, Finnair, Will Schart, UAL, Ray
Butler, UPS Training Technology Subcommittee: Jean Louis Bravo,
Chairman Airbus CBT Desktop Simulation Communication:
Jean Louis Bravo, Airbus Jean-Louis discussed by Airbus has chosen desk top
simulation: best practice, time saving in terms of
reuse of the simulation, gives a real world feel,
and the simulation is certified by the manufacturer.
Airbus uses simulations from CAE. The simulation is
the A320 simulation. The simulation links with
other applications: CBT, A/C walk-around, and
on-line documentation. The desk-top simulation runs
in several configurations: laptop, desktop with 2/3
monitors, etc. Jean-Louis discussed device configurations –
multiple laptops can be linked to the training device but
only one lap top can use the TD at a time; the
connection is via USB as stated earlier. This is
used at Airbus for procedure study or maintenance. The
synchronization/initiation of CBT and training device
session is from an AU or directly from CMI/LMS. Communication
is via the LAN or a direct link (USB easy to implement
and use). The USB link is used for communication
and is fast enough for the requirements. Jean-Louis
showed the architecture: TCP/IP and the training
device have the same address and the laptop has an IP
address based on the laptop equipment number. This
ensures that there is no conflict between laptops and the
training device. There is a software proxy on the
laptop side. Jean Louis described the training
device API using a Get URL encoding command to control
the Training Device and the response is an HTML string.
There are a minimal set of functions on the API: query,
start a simulation, preset a scenario, stop. Jean
Louis discussed how it was implemented: the CMI
uses an EXE launcher and the LMS uses an ActiveX control
inside an HTML page. Client (CMI/LMS AU <>
USB proxy client) Server (USB stub server <>
http/get IIS Web Server <>dll<> Training
device simulation Jean-Louis then gave a demonstration of the Airbus
Desktop Simulation. Airbus simplified the commands to start, stop, and
running commands to make sure the simulation-AU sequence
is running correctly before the training actual starts.
CAE delivered the simulation with the ability to record
sequences that Airbus calls templates. Jack asked if you could call up the thrust levers or
any panels. JLB said that he is running a true
simulation of the A320. The CAE templates allow the
instructor to setup the attributes of the training device
to meet the training objectives of the CBT AU. Courseware Management and Processes Subcommittee:
Yvonne Johnson, Chairwoman How to Identify Differences Between Successive
Updatings: Bernard Bouyt, Airbus Introduction When an airline gets a training course from an air
frame manufacturer, the airline may want to
change/customize the content. Then the manufacturer
delivers an update and the airline needs information
about the changes to update the current (airline) content
most effectively. This could be the result of a
language translation or just changes to the original
content. Bernard review slides on translation. International
airlines use English. Domestic airlines use local
language or languages) Bernard discussed the levels of complexity in
translation: external audio, call-outs, etc are
simpler to translate, audio synchronization and rename
files to the original the course maintains the same
structure and modification is relatively easy. Any
audio translation that changes synchronization of audio
to schematic presentation, will be more difficult. Bernard
has not worked with any multi-language courses. Bernard
then recapped the problems when a manufacturer needs to
translate a course. Bernard told of a project that
used synthetic voice (text to speech) where the voice was
very realistic (undistinguishable from human voice) so
the technology can be very useful. Bernard looked
into voice technology and things like CDU was a problem. AI: Bill recommends a presentation on
comparison of various synthetic voice technologies (Anne
ask Kevin S if he would make a presentation; compare and
contrast; what needs to be used on the client and server.
Tom King recommended we ask Carol Wideman. How do
you get the speech into a Flash application?). Looking
for an integration where the text is being read by the
speech engine; this makes the band width for audio nil.
Issue is load speech engine on client). Bernard discussed the updating of an existing course
and the course structure differences. He discussed
the identification of changed, added, deleted AUs. Next
Bernard identified content differences and how Airbus
determine what, how, and when to make changes. Bernard
discussed how some metadata elements will help identify
content differences; this could be done by a tool to find
content differences using the metadata. Bernard discussed how Airbus uses metatdata, what is
delivered to airlines, tools used by Airbus. Bernard
is not sure that the airlines know about/use the Airbus
metadata for identification of changes. AI: Bernard will demonstrate an Airbus Course
Structure Differences Tool in Integration of Training and S1000D: Bill
Shook, Boeing S100D was started by ASD (Europe Space and Defense).
International Standard Specification for technical
documentation. Bill showed the S100D Maintenance
organization. ASD signed an agreement with AIA to
bring S1000D into the Bill discussed why S1000D is being accepted. The
spec covers Type I Data (manuals) and Type II Data (data
driven from a source data, no manual). Boeing is
delivering all A/C as Type II data. Tom asked if a
style could be applied if you could get a print (the
original set was designed not be printed – no page
type style sheets). Now moving to style sheets.
Kevin raised a concern about trying to print
non-printable data (e.g., Flash files). Tom King
cited an example where a procedure was a Flash animation
but also provide a checklist that could be printed.
Kevin thinks focusing on paper driven solutions is a step
backwards; Tom says try to live a day without some kind
of paper. Bill says there will be customers that
will need paper. S1000D enforces view commonality
– Bill says this is an interesting debate. Bill discussed the benefits of Type II data: reduction
of sustainment costs, maintenance turn around time, and
footprint (distribution and warehousing costs). Bill
sowed the S1000D timeline and roadmap. Release in
2006 will have some training integration (ADL)
requirements. Plan is that all releases of S1000D
will be backward compatible. OSD (Office of
Secretary of Defense) is close to mandating S1000D
(analyzing services impact) and is mandating SCORM. Bill gave a description of S1000D. A large data
base with 9 types of data modules (e.g. functional
description, operational information, parts information,
information diagnostics, etc) with a specific DTD (how
the data must be structured to accommodate the DTDs).
Bill gave an overview of S1000D and SCORM integration
issues. Bill has been working to bring S1000D
and ADL groups together. Bill looked for CMS/LCMS
Vendors that supported S1000D – SCOMR: Ecosystems,
Siberlogic, Immedius, Bill discussed reasons why to integrate training with
tech data. He gave an example where there were
incompatibilities and integration in systems released
with separate databases instead of an integrated tech
data/training data database. Bill’s Type 45
project is taking some initial tasks and move them into
the EPSS training environment – this means learning
on the job in stead of in the classroom. Jack
re-interated that this does not reducing the training
time; you are moving a portion of the training from the
classroom to on-the-job. Bill demonstrated a project that used an integrated
data base for S1000D and reuse of S100D elements for tech
pubs and training materials. This was done as a
collaboration between, Boeing, HarvestRoad, and
Australian Dept. of Education. He discussed the
idea of author once and use many time – this uses a
reuse repository. Data comes from LSAR (for the
landing gear system) and was the basis of data for the
training and pubs demonstrations. For the purposes
of the demo Bill used bit-map graphics not vector
graphics. Metadata was automatically populated by
the HarvestRoad product. Bill demonstrated reuse of
information from the tech pubs in the training material.
Bill McDonald said that context can change the meaning of
the reusable information. Bill Shook agreed and
said this is just the first step to show it is possible. The S1000D spec will change in the future to allow
tagging of data at a more granular level and using XML
for tagging. Robby’s concept is S1000D is a
data repository that can be disaggregrated for training
reuse. Robby said that this does not take advantage
of the context (or source) of the S1000D data. Bill
countered that the project he demonstrated was just a
beginning and could be extended to include everything
Robby had discussed. Bill says these concepts in
conjunction with metadata can support multiple versions
of the tech pubs and training modules. Bill discussed integration issues: philosophical
issues, business process and rules. Philosophical
issues include: different purposes, different
design needs. Training is designed to tech concepts
(no need to verify the data) and tech data focused on
providing procedures. Business process impacted by
integration: business rules must be established,
data integration and support must be considered
(additional metadata- cost, data management environment
–LCMS/CMS, data delivery - LMS, S1000D viewer).
There are Technical Issues such as data management
(granularity, data structure, metadata) and data delivery
(viewers, LMS interfaces, performance tracking). Bill
said S1000D is great but there are proprietary viewers
but the government wants non-proprietary viewers. Bill
was asked to develop a non-proprietary viewer, went back
to his management; the management said “What’s
in it for me? Bill did not have a project for a
free S1000D viewer. Jack asked if S1000D had any
restrictions on data formats. Bill S. said
the multi-media will have guidelines for any multi-media
type. Bill McDonald said that it is highly unlikely
anyone is going to develop a single one-sized fits all
viewer. Bill S. said you need a style sheet. Robby
said there are 2 different concepts: a viewer and a
renderer. The viewer needs to go beyond rendering
to handle a behavioral component. The training
committee is working on bring up S1000D to a training
standard. Tom King asked if the taxonomy was
suitable for training. Bill Shook thinks there
needs to be a common taxonomy that supports both. Bill discussed granularity of SD1000A data. For
Bill’s project he used SCORM data model but thinks
the AICC model might work better. Bill discussed
possible changes to the LOM to support S1000D including:
objective, knowledge/skill, audience,
effectivity/applicability, and relationships. Bill
discussed impact to SCORM: terminology, run-time to
support EPSS, adding structure to SCO black box,
competency model, business rules. Bill McDonald asked if there were S1000D DTD’s in
work for training. Bill S said not yet. There
are efforts working on this. Bill Shook’s
hypothesis is that if there were a set of training DTDs
that would be associated with the S1000D database this
would provide an environment for training and tech pubs. Metadata Flexible Hierarchy: Jack Hyde,
AICC This presentation was not given because of the
discussions involved in the Metadata document review. Metadata Document Review: Jack Hyde, AICC
and Bill Shook, Boeing Bill S reviewed the goals of the session. Jack
Hyde reviewed the metadata document current status,
identify changes, and collected comments from the
audience. Jack’s focus is for the audience to
understand the document. If the audience
doesn’t like what is there the audience should bring
it up with Jack. Jack talked about document
organization, what has changed, what’s left to do,
and two new sections: Obligations and Collections.
AI: Bill Shook/Jack Hyde: Tom King
recommended that a document be created containing all the
examples instead of embedding examples within the main
document. Jack reviewed the areas where there are questions of
what the document should do Obligations: Jack explained obligations and gave
several alternatives for the document. The group
will vote on preference after Jack’s presentation. Collections: Jack described the concept of a
Collection and reviewed his recommended metadata. Jack
summarized the Collection concept. The LOM can
describe a Collection but has a weakness: you can
not talk about members of the collection. Jack
discussed the LOM relation category and showed why it is
only designed for a one-to-one or one-to-few relationship
and cannot be used to correctly describe a collection.
Jack described why a separate metadata instance for each
member of the collection. This would store a lot of
redundant information and loose the collectedness of the
group (couldn’t see the entire collection). Jack
discussed the data element called Collection with type:
uniform or mixed and member information Jack gave
an example how the collection could represent an exam.
Jack presented a set of Collection Questions when
reviewing the document. AI: Tom King noted that the abstract
needs to be written and the contact information needs to
be updated, delete the registration, use admin@aicc.org for
contact, update the abstract and keywords. Robby recommended a scope and purpose. What is
the purpose of metadata, who is going to use, and how
will it be used. Robby gave an example: Purpose
of LOM is to exchange metadata outside your community.
Another use of metadata is for managing your own
repository. What metadata that systems will see.
Systems doesn’t know that an obligation is embedded
in the metadata. Is this for managing metadata in
repositories or exchanging data among systems, or
communicating metadata between training systems. Robby
mentioned METS had a complex Collection set of metadata. AI: Jack. Add specificity to the
new sections on Scope and Purpose. Bill recommended
that we should explicitly state that we are extending LOM
not the AICC LOM. Recommend change the second sentence in
the Introduction. Bill had an overall question:
The PENS specification had a new outline on the document
format. Do we want to use it? The document is
IEEE sections, formatting, fonts, and conventions with
AICC front matter. This would add more discipline.
This will match ISO which has specific requirements.
Change the NS Obligation to Optional – this only
applies to one element: Semantic Density, Coverage,
Robby recommended to consider what is mandatory for a
repository (all elements in a repository), an instance, a
training system. Does anyone building a repository
for AICC required to support all elements. In
authoring tools you may want to consider what you need.
People will not generate metadata unless it is automated.
Metadata is most valuable for non-text based objected and
aggregated objects. Metadata helps with maintenance
and aggregation of non-text based data and helps with
search and discovery. Bernard does not see why anything is mandatory. Robby
recommends that anything that was mandatory in SCORM
would be mandatory in AICC LOM. Manage, locate,
evaluate, maintain, and exchange learning objects. Add
specificity and properties of identifiers. Robby
does not think objectives should be mandatory. Neil
suggested that some data should be mandatory working with
organization and should be optional in a broader context.
Once you define your scope and purpose what principles
apply (e.g., minimal obligations, exchange data in and
out of AICC – use the LOM elements and don’t
overload them). Kevin recommends looking at
breaking of structure from data model and taking data
model elements and putting them into a broader structure. Metadata Working Group Meeting: Bill
Shook, Group Leader Based on comments from the meeting attendees, a small
group worked late implementing the suggestions from the
floor to provide a scope and purpose. WEDNESDAY February 1, 2006 Overview of NATO Advanced Authoring Panel
Activities, Dr Dexter Fletcher, IDA Dexter discussed the formation and goals of an
exploratory group on advanced authoring technologies.
They formed task group in Dexter discussed an ADL (biased perspective) on
authoring tools. Combine education training and
performance aids – anytime/anywhere. Dexter
discussed an ADL vision of objects floating in the global
ether and persons using the objects in various ways and
places. The problems is: How do you build the
server to reach into the objects and build the content in
real time for delivery anywhere/anytime? Direction
they are going: fewer lessons/more learning; fewer
tests/more assessment; personal learning associates/PDAs
(classrooms, out of classrooms, anytime/anywhere); less
authoring/endless material. Neil asked for elaboration on meta-authoring. Meta-authoring
is authoring of authoring – how do you assemble the
tools for authoring. Robby disagrees with the
concept of a central server and wonders how you
assemble content in a distributed system. Dexter is
less on concerned with the architecture rather than the
concept. Avron Barr noted that there may be
different architectures for different needs. Allen is thinking of releasing his tool for general
use. If anyone is interested contact him at
munro@usc.edu. iRides and Simulation Interoperability: Dr.
Allen Munro, USC Allen will discuss iRides and how simulations can work
more closely with instructional systems. Allen will
discuss how practical the SCORM notion is. Allen
discussed simulation behavior authoring and cited several
examples. Allen demonstrated a simulation with the
ability to insert several malfunctions at various
points in the simulation. He provides tools which
can change the behavior of the simulation. The
simulation has simple/complex rules which govern
behavior. Allan demonstrated how the change
simulation behavior by changing the rules on the fly.
He clicked on the object, view the rules that govern that
object and change the rule. Allen showed a
simulation that is a test to plan the defense of a
carrier group. Allen has the capability to change
instructional assessment. Because there is no right
answer for the task, there are methods to assess goodness
of tasks, done with conditional rules; you get credit for
components of the task – separate scores – core
scores and extra credit scores. The iRides tool has a simulation engine and an
instruction interpreter. This is a Java app and can
be delivered as an application or can be delivered as an
applet. There are wrappers that allow the applet to
report SCORM data back. Can package it to talk to a
records broker which can post all kinds of stuff.
Now the records broker is not packaged to be 2004 SCORM.
There was a discussion on object size. Neil said
SCORM did their best to strip context out of learning
objects. Robby gave a counter response. Allen demonstrated a fairly complex simulation; the
result of the exercise was to show that the simulation
can augment the student’s understanding of the
problem/solution. Allen then showed how iRides can
allow the instructor to create and edit the instructional
component providing a training lesson. Finally there was a review of iRides. IRides
provides authored interactive graphical simulations with
an extensible behavior language, authored
lessons/learning objects that work with the simulations
(these are on-the-fly, modifiable). The instruction
specifications are XML-based. Allen would like a
level of instructional control that would be general.
Jack and Bill said that is Jack’s view of
Simulation-Instruction Interoperability. Jack
talked briefly on the AICC opaque approach. Allan
discussed his experience with that type of approach and
some problems he encountered. Allan suggested that
there are a basic set of instructional interactions that
can be added to or combined/built upon to create more
complex interactions. Jack asked if Allen would
look at Jack’s list of commands and/or if Jack could
look at Allen’s list of commands and exchange ideas.
Allen agreed. AI: Jack to contact Allen Munro to
compare/collaborate on the instructional commands that
govern simulation behavior. Developing and Exploiting Reusable Simulations
in Flash: Dr. Doug Towne, USC Doug will discuss Flash and its capabilities, ReAct
– a simulation and delivery system, examples. The
introduction showed a diagram of the authoring software
resources that make up authoring tools running from
general purpose to special purpose. The release of
Flash Action scripts was the point in Flash when it
became the foundation for building a simulation
development system. Doug talked about the richness
of the Flash Object Library and the availability of them
on the internet. The ReAct system also has a large
library of Flash objects. Almost every PC on the
internet has a Flash player. The ReAct system is funded by ONR and is the heart of
IETAMS (Interactive Tech Manual System). ReAct adds
and extends the capabilities of Flash. ReAct has an
explicit specification for building objects. If you
build objects to this spec, then the ReAct simulation
engine will use them. ReAct has libraries of Object
Prototypes for model-based behavior. The prototype
object has the behavior built in, all you need to do is
build the graphic. The key is model: ReAct
models are collections of objects that can be used for
various purposes. Doug showed an example of a
simple react model. Doug demonstrated a basic
objects controlled by a mouse but is also effected by
states. Doug showed how to move from tech manual to
operable model. Doug showed a list of ReAct
applications. The application developers new Flash
and had an understanding of Action Scripts. They
learned how to use ReAct and then developed the
applications Doug described the Command Information Center (CIC)
application. This application was built and then a
training model was added after the fact. ReAct is a set of Flash objects, a simulation engine,
functions. This is all written in Flash. Flash
can communicate externally. The object
specification allows the objects to be controlled
externally and can be queried externally done thru the
Flash environment (API to get and set states) but this
could be done thru web services (Tom King: this
will be greatly easier in Flash 8). Doug demonstrated several ReAct training/simulations
from small simulations to very large multi-component
simulations. In conclusion, Doug gave an overview of reasons to use
simulations in general and ReAct in particular. The
powerful reasons for ReAct are: usability, reusability (a
very powerful and varied set of reusable tools and
simulations) and scalability, deployability,
interoperability, supportability (because is built on
Flash ReAct gains from improvements in the tool set). Questions: ReAct is a home grown tool built from
the ground up – no software system could do what
needed to be done. When Flash came about, ReAct was
developed at the university with government funding. Then
the government funded it to be brought up to commercial.
It is not off the shelf but are available commercially.
Overview of Simulation Interoperability
Standards Organization (SISO): Dr. Katherine L.
Morse, SAIC Katherine presented the SISO Vision and Mission
Statements. She is interested in discussions
between standards activities and technical activities.
The role of SISO is to identify standards for
interoperability, evangelize and promote buy in for
standards, provide a stable open interface for COTS
products. Katherine described the SISO
organizational structure. Katherine described SISO,
Inc, the non-profit corporation to enter into contracts
and other financial agreements. SISO has 3
simulation interoperability workshops a year. The
workshop has a conference type format but is a workshop.
The workshop has user community forums. Katherine
described the balloted products development process, a 6
step process. Katherine described the 11 active
product development groups (base object model, COTS
simulation package interoperability, distributed
interactive simulation, etc), the activity study groups,
standing study groups. For more information on or to join SISO, go online to
www.sisostds.org. AICC Simulation and Smart Graphics: 2 Approaches Smart Objects: Smart Graphics and Smart
Model: Kris Rockwell, Hybrid Learning &
Sebastien Fraysse, I-COMPONENT Sebastien discussed Smart Graphics for Instructional
Designers. He provided a definition of the scope of
the task to design interactive scenarios. The
objective is to provide an understandable language for
Designers to Designers and Designers and Developers when
creating interactive presentations. The methodology
has 4 steps: task hierarchy, navigation paths, global
(permanent) resources, leaf (step) resources. Sebastien discussed applying templates as the next
step: Learning Template (non interactive), Aiding
Template (interactive), Testing Template. After the
analysis then the information needs to be formalized
using 3 specifications: IMS Content Packaging
Model, IMS Simple Sequencing Model, AICC Smart Graphics
Model Kris Rockwell discussed a Model for Smart Objects.
He has a 3D Model imported into the Unity game engine.
He can build Smart Object for several platforms. The
object communicates over TCP/IP. The Broker engine
is Java based TCP/IP server that process XML messages.
The messages are very simple. Kris then ran his
demo. The function of a smart graphic is to
communicate state information. Kris’s demo
shows that simulation and smart graphics blur. He
wants guidance on the XML spec and further recommendation
from other groups interested in help to complete the spec
development. Simulation/CBT Interoperability: An AICC
Approach: Jack Hyde, AICC Jack defined the terms that AICC uses like content,
courseware, and CBT. They are used interchangeably.
Jack defined his term for Simulation. Next he
talked about strategies using simulation in training.
Jack then described with examples each strategy: Demo,
Path Simulation, Allowed Deviations, Intelligent
Tutoring, Free Play. Jack then provided a problem
statement: How do we enable simulations with our
favorite courseware building tool(s). Build a
specification that will allow simulation and content
building tools to work seamlessly together, allow any
simulation (that conforms to our spec) with any content
building tool, enable multiple strategies. Jack gave an overview of how a simulation will connect
with courseware. CBT may not or may launch simulation. Simulation
may or may not run on the CBT system. Content launches simulation, content sets conditions,
content may obtain final state when student finished Continuous communication between content and
simulation to maximize the number of strategies we can
use Next Jack talked about an architecture with 4
components: CBT logic (authoring/deliver) , CBT
Display(s), Simulation Logic (not in one box often
multiple simulations to run alone or with other
components), Simulation Display(multiple displays).
There needs to be communication between CBT deliver and
CBT display or Simulation logic and simulation display.
We are concerned with communication between Simulation
pile and CBT pile. Jack sees the Data Block Driver
as the point that handles the communication. Jack then talked about what does a simulation have to
do to work in a training environment. Jack talked
about what can/should a specification for
interoperability include. Jack talked about the
flow control model. AI: Jack consider the suggestion from
Robby to look at how to launch and how to terminate
should be added to the standard. Jack talked about what to communicate. Do not
want to know all internal states of the simulation.
Only want to know what the student did not everything the
simulation did. Standard designed with single
student in mind. Would be desirable to have
multiple students run simultaneously. Primary
scenario is one-on-one self-paced. Sean said with
the new airplanes, maintenance training will be more
challenging than pilot training. When working with
multiple students you are working with multiple model
views. Allan said there are 2 lists of services
that needs to be addresses. Jack discussed categories of communication: Display
Management, Simulation Management, (has set mode and get
mode which Jack defined – sequential, concurrent,
step, free-play, command). Robby recommended a
smaller set of commands that could be combined to provide
the modes that Jack defined. Robby said that
middleware can handle the commands. Allan said he
uses a mode where the simulation tells back the student
actions: all the interactions going into the CBT and CBT
interface processes them, turning off or filtering where
needed. Robby says to build the logic into the user
experience. With that model the modes are use
cases. This would then be the modes that the
authoring system would support simplifying the developer
tasking. Jack discussed the problem with standard names –
this is too big a job. The next idea was name
nothing – just store and retrieve blocks of
information, the opaque approach. Jack explained
this approach. The next approach lets the simulation maker name the
controls and functions: booklet or query. What
is required is a standard format for all the names and
the functions in a simulation. Then there is a
standard set of commands – the API. Documents on the AICC web site: MPD10 –
Simulation Interoperability and SSG10v7 – Content
and Simulation Communication will be in the online
presentations for the Jack talked about next steps. Allan will look at
Jack’s commands. THURSDAY February 2, 2006 Macromedia – Captivate Presentation: Silke
Fleischer, Adobe Silke gave an overview on Captiva. It creates
simulations and demonstrations (was RoboDemo). She
explained the areas in a corporation you can use it.
Captivate is similar to Powerpoint. Publish to
Flash files. Create Flash content without learning
Flash. Silke gave a quick demo of the tool. You can
record with a custom size. Bring up the web page
(or application you want to record.) and the recording
mode. Captivate does not generate an flv file but
you can export to an flv file. There are timelines
so you can add audio, etc. There are many different
publishing options. You can search for content in
the Captivate files. Captive integrates with many other products. Question
about integration with QuestionMark. You can mix
QuestionMark assessment with Captivate content. More
information on the QuestionMark website. You can
put captivate content inside a .pdf. You can use
Breeze playback for Captivate slides. All
participants in the meeting can interact (hit the
unsychronize button) with the Captivate portion. Silke reviewed a Case Study about SafetyKleen and
training to understand new outlook setup/access. She
asked for anyone to contribute to her Case Study file.
The next version of Captivate will use Flash video Silke gave is brief description of what will be in the
next release of Captivate. She showed the project
wizard and how to build scenarios. The new version
provides a branching view. There are many new
features to show paths, interactive objects, etc to
view/change path, tracking data, editing, libraries,
hide/show objects. This version adds flexibility;
no firm release date but they will start beta soon. Silke likes to drop simulations into Powerpoint
slides. The SwiffPoint Player (www.globfx.com) allow
you to put Swiff files in your Powerpoint. Many
examples are available on the web site. You can
drop “easter eggs” into a presentation (www.eeggs.com). Silke
talked about pseudo-video – images with slight
changes and a fading effect – big plus is small file
size. In captivate, select pictures and put on time
line. Silke showed a list of resources for captivate. These
are shown in the presentation on the aicc web site. Authoring It is important how we communicate. The words
you say is 38% of how we communicate – intonation,
way say things, body language adds the other factors how
we communicate. Why don’t we add 3-D virtual
mentors in our training. Vcom3D has authoring and
other tools to provide virtual mentors. Carol gave
a brief overview of the Vcom3D applications. Carol demonstrated a flash application with virtual
mentors. The interactive character is used to focus
the learner and draw attention to the information. Carol
showed the set of authoring tools. She will show
how she created the virtual mentor. Mentors are 3
dimensional. She demonstrated how easy it is to
build a new gesture in 3-D. After you author the
gesture, you export it into a library. The mentor
does whatever the script tell him to. The tool is
very quick to learn. Carol showed how to change focus of attention. One
you create the animation then you can export to the
Vcom3D simulation engine, Flash/Captivate, Video,
Animated FIG, or Microsoft Agent. Carol show a Captivate demo. She showed a slide
on the content development process with Vcommunicator ,
Flash, & Captivate. Using VcommunicatorStudio
captured uncompressed AVI, the export to Flash6 and
remove with background color. To export to Flash 8
there is significant compression in the flv size. She
explained how she used Flash 6 & Captivate and Flash
8. Question: Is there a way to get a vector image?
Answer: No, not a feasible approach. Carol showed the content architecture. There is
a game engine capability in addition to their simulation
engine. They also have a project to incorporate a
Cognitive/Affective Agent (give the character a mind).
They are adding libraries of languages and cultural
gestures for work with the military. Carol Showed Vstudio. She showed an automatic
translator, VReader. She highlighted some text and
the virtual mentor read the text and change the
inflection. Converting Legacy Courseware In Various Formats
Into XML-Based SCORM Content With A Styleable UI: Tyde
Richards, EduWorks SCORE is a tool that converts content in various
formats into a SCORM- compliant structure. Tyde
showed Authorware content converted into SCORM SCO’s
in HTML. There are sets of XML files that
correspond to the screen states that are shown. The
converted content was a large learning experience. Tyde
discussed the project results. The key to taking
content: source content in different formats, SCORM
conversion tool, SCORE exchange format, SCORE tool, SCORM
package. You need to build a converter to take
source content into an exchange format. Then you
need to do editing. There needs to be a data model
that underlies everything. The SCORE tool provides
the consistent methodology. The success of the data
model depends on a card deck (linear presentation)
metaphor. Tyde talked about XML SCORM Studio. The project
decided to use the OASIS Open Document Format (ODF).
There is an open source library for this. Smart Graphics: Overview of the
i-Component 2.0 Project: Sebastien Fraysse,
i-COMPONENT Sebastien presented a definition of smart graphics and
smart objects. He described the concept of smart
graphics and smart models. He showed the components
of a smart model. Then he gave a demonstration of
smart graphics and smart models. Enhancing Virtual Training with Networked
Simulation Environments: Bill Andrews, DiSTI Bill asked the question “What is a Simulation”.
Simulator folks usually have no instructional designers
usually involved. Courseware people usually build
small instructionally significant simulations with
specific educational targets. It may hard to get
these groups to agree on what a simulation is. Bill
discussed why we need network accessed simulation: execution
footprint (much too much model for student station),
download time (model too big), mobile code (not allowed
to install executable code), operating environments
(model doesn’t run on your browser), licensing
(model is keyed software – HLA), Bill
discussed current DoD simulation technologies
(mainstream: DIS, ALSP, HLA; other: CORBA, DCOM,
SOAP, custom). Simulation folks are spoiled because
they can pick their hardware, networks. Simulation
programs have long timeline, higher costs, higher
security, good (high end or custom) computers, lack of
instructional designers but many software engineers.:
object oriented, two related but functionally independent
standards (object model and runtime
infrastructure). Bill discussed DIS/HLA design
principles. Bill discussed middleware (translator)
software. Bill discussed Rumtime Infrastructure
(RTI). Bill discussed HLA. RTI is a fundamental
software component of HLA. Single vesion must be
used by all of the interfacing applicataions/simulations.
Bill discussed the RTI services. Bill said that
there were only 2 RTI vendors. Bill discussed how
are you going to deploy. SCORM is specific to
web-based environments, There are many aspects to SCORM.
There is no guarantee what you are going to deploy into.
Bill discussed typical security problems. He
discussed mobile code; it is very difficult to produce
good multimedia without mobile code. DOD
policies do allow mobile code within restrictions. Most
installations do not allow mobile code that accesses the
network. If you try to create a Flash app that
tries to talk over the network there will be a problem.
The next generation of courseware will have mobile code.
There are several implementations that have done SCORM
– HLA interfaces. Bill wants everyone to think
about Gateways. A gateway device with structured
queries are a good solution. This is free-standing
middleware.. SCORM needs to account for
communications outside the LMS. ADL folks need to
take on the security and infrastructure folds. Prepare
a few templates for inclusion of simulation in
courseware. Decide how data comes/goes from network
asset. Levels of Interactivity: David Castillo David discussed using levels of interactivity
framework to guide your interaction development
decisions. David shows one piece of content to show
what a lelvel 4 piece of activity. The frame work
should help make decisions about what level of activity
you need. Once this decision is made you can
allocate resources better. Present and guide,
explore and discover, ass & remediate. He wants
to discuss how a design too can map to levels of
activity. Factors influencing levels of interactivity:
aim, domain flow, prompt, feedback, logic (input device,
media type). David doesn’t believe that input
device and media type are not factors. David will
address tools and design methodology. David describe a framework showing factors, levels,
and proficiency/exchange of information, amount of
instructional feedback, logic, tools. David
described an example from a C-21 Air Medical Evacuation.
He compared the example simulation to the levels of
activity for the different components. David showed
an underlying model that would take the simulation from a
level 2 to level 3 simulation. David demonstrated
that a choice tree can be advantageous for an
instructional designer and will provide a limit (based on
choices) on the interactivity (and cost). Captivate
lends itself to this type of development. David
showed a state chart that models the process. The
state chart helps provide a picture to communicate with
the developer on what I want done. This provides a
conceptual visualization that can be used to communicate
between designer and developer. It is difficult to model
intelligence so David is looking for alternatives. Don’t
model the instructor, create a learning lab. He
displayed a concept that has middle SISO-SAC/IEEE LTSC Simulation Interface Standards Project LTSC and the SISO Joint Project to Standardize
Interfaces between SCORM and Simulations:
Welcome, Goals, Scope, Logistics: Robby Robson,
IEEE & Dr. Katherine L. Morse, SISO All position papers will be available via a link from
the Meeting Presentations on the AICC web site. This
is a joint effort to discuss integration of SCORM
content and simulations. Robbie says he does not
have answers and doesn’t even know the questions.
He hopes he will have the questions and some of the
answers by tomorrow. There were several
presentations which can be found via a link on the AICC
website under the information for this meeting. A
short synopsis of the following papers were presented: Simulation/CBT Interoperability: An AICC
Approach: Jack Hyde, AICC Integrating Simulations in SCORM using the Open
Document Format Framework: Robby Robson, Eduworks Position Paper on IEEE Simulation Interface Standards
- A 3D Simulation Component Framework: Carol
Wideman, Vcom3D Representing Simulations within Shared Content Object
Reference Model (SCORM): Shane Gallagher, SAIC SCORM-Simulation Interface Standards Position Paper:
Brandt Dargue, Boeing Extending SCORM to Support Simulations: Jim Ong,
Stottler-Henke Advanced Distributed Learning Technology: Feasibility
Study for ESA Programmes (Remote Presentation):
Luis Arguello, European Space Agency Experiences and Lessons Learned Integrating HLA and
SCORM: Katherine L. Morse, SAIC/DMSO IAI Position Paper on SCORM and Simulation Interface:
Bob Pokorny, Intelligent Automation Stand-Alone Simulations for Individual Training:
Geoffrey Frank, RTI Integrating Simulations into SCORM Learning
Environments: Brent Smith, Engineering &
Computer Simulations Integrating HLA with SCORM for Effective E-learning:
Brian Spaulding, MAK Technologies A View From a SCORM Immigrant: Bill Ferguson,
BBN
|
| If you have questions, suggestions, or problems with this website, please send e-mail to webmaster@aicc.org |
| Last Updated: 03-Feb-2006 08:19:51 AM Pacific Standard Time |
| Copyright (c) AICC 1988-2005, All Rights Reserved |