Human Variability
in Computer Generated Forces (HV-CGF)
Project
Scope
The goal of this HV-CGF project is to design, implement and deliver
a model of human variability in computer generated forces. The
project focuses on the representation of human
and representation
of the effects of external and internal moderating influences on
the Computer Generated Forces (CGF) entity and unit
in an effective and practical manner.
Research
Year 1
The
project began in February 2003. The 1st year's activities are as
follows:
Set up Joint Development Infrastructure
Study Causal Relationships between the moderators
and the Co-JACK
System
Prepare Statement
of Stakeholder Relevance
Create Baseline Scenario with suitable
military doctrine and tactics set
Illustrate how people change
their
when under the influence of certain moderating
factors
Prepare a document on how to map Cognitive Architecture
(e.g. situational awareness) onto the underlying JACK agent technology
Present
at MoD's Corporate Research
events
Research
Year 2
Demonstrate human variability with an initial demonstration
Document
techniques for representing
moderators and their
interactions in a CGF
Deliver finished software and technical
reports to MoD
Close project with a final
demonstration and presentation
Project Outcomes and Novel Features
At a high
level, the project will deliver:
A scientifically justified cognitive
architecture, cognitive models, and
overlays
A
software system built upon the JACK platform, used to validate
the cognitive
techniques, and capable of being readily
integrated with a wide range of Synthetic Environments (SEs), CGF systems and
simulators
A demonstration of the capabilities of the
system, showing the effects of moderators through the resulting
of entities in the OneSAF Testbed Baseline (OTB)
The technological basis for
undertaking this research is intelligent software agents. JACK provide a generic
framework for collaboration
and control of multiple, concurrent systems, which may be either
functionally or geographically distributed.
The novel features
of the project, and their relevance to the MoD customer include:
The project will extend the BDI (Belief, Desire, Intention)
paradigm of intelligent agents with a cognitive architecture
that can
be used to include psychologically based constraints as extensions of the
BDI model
By means of
Moderator Overlays (BMOs), this
work will provide
a mechanism for influencing the
of cognitive
models implemented using the cognitive architecture
This work will allow CGF
entities to be controlled by external agents that incorporate
a cognitive architecture and
moderators
The project allows a unique combination of generic cognitive
responses, doctrinal
and non-doctrinal
to
be
combined into a single coherent
model, which can be directly integrated
with
a range
of
SEs or CGFs
The project extends the current JACK/OTB integration
being conducted by AOS and DSTO (Australia), which is a unique
integration
between intelligent agents and OTB entities
Exploitation
Potential stakeholders for the project include:
Users of CGF systems, e.g. for training, war-gaming, and
mission rehearsal
Acquisition groups which specify the requirements
for new CGF systems, e.g. the UK Defence Procurement Agency (DPA)
Analysts
who research CGF systems and support capability development,
e.g. the UK Defence Science and Technology Laboratory
(Dstl)
International military communities involved with CGF technology
Anticipated benefits arising from the HV-CGF project include:
Increase confidence in the credibility of CGFs so that
they accurately reflect how entities and units change their
over time when exposed to moderating influences
Provide a system that is
applicable to both enemy and friendly forces, so war-gamers can
better investigate force mix
equations and cultural effectiveness
Improve potential for training by representing
non-attritional enemy
(for example surrender, or fleeing)
and
generate more accurate
for non-combatants, asynchronous enemies and
other
non-standard scenarios
Provide capability to model coalition force structures
and interactions through
of cultural idiosyncrasies,
religious factors
and other effects
References
A.
Lucas, Human Variability in Computer Generated Forces,
presented at Behavior Representation in Modeling and Simulation
(BRIMS'03), 2003.
N. Howden, Press Release:
UK Ministry of Defence Research Contract Goes to
Agent Oriented Software, The Age, 2003.
M.L. Fineberg and G.E. McClellan,
Modeling the Effects of Suppression
in Synthetic Dismounted Infantry (SynDI), Draft Report
no. DNA 001-94-C-0024 for US Defense Special Weapons Agency 1997.
P.
Gillis, Realism in Computer Generated
Forces Behavior, Report of the US Army Research
Institute, 1998.
R. Archer, N. Lavine and S. Goldberg, Using Human
Performance Models to Train Tomorrow's Soldiers
for the Objective Force, Report of Micro Analysis & Design
Inc. and the US Army Research Institute, 2001.
J.E. Vanderveen
(chair of the Committee
on Military
Nutrition Research), Caffeine for the Sustainment
of Mental Task Performance: Formulations for Military Operations,
Report of the Institute for Medicine,
published by National Academy Press, 2001.
T.R. Johnson,
Control in ACT-R and Soar, Report
of Ohio State University, and in the proceedings
of the
19th annual conference of the Cognitive Science Society,
1997.
E.
Norling and F.E. Ritter, Embodying
the JACK Agent Architecture, Report of the University
of
Melbourne and Penn State University, and in the proceedings
of the 14th
Australian Joint
conference on
Artificial Intelligence, 2001.
F.E. Ritter and M.N.
Avraamides, Steps Towards Including Behavior Moderators in Human
Performance Models in Synthetic
Environments, Report of Penn State University, 2000.
F.E.
Ritter, M.N. Avraamides and I.G. Councill, Validating Changes to
a Cognitive Architecture to More Accurately
Model the Effects
of Two
Example
Behavior
Moderators, Report
of Penn State University, and in the proceedings
of the 11th Computer Generated Forces and Behavior
Representation
conference,
2002.
F.E. Ritter (editor), Techniques for Modeling
Human Performance in Synthetic Environments: A
Supplementary Review, part
of the Soar State
of the Art
Report, Report of the Human Systems Information
Analysis Center, 2002. http://www.hsiiac.org/hsi/products.do?action=detail&code=HS-2003-2
S.A.
Wallace and J.E. Laird, Towards Automatic Knowledge Validation,
Report of the University
of Michigan,
2002.
R.W. Pew and
A.S. Maver (editors), Modeling Human and Organizational Behavior:
Application
to Military
Simulations,
Report of
the US National Research
Council's panel on Modeling Human Behavior
and Command Decision Making (Representation
for Military Simulations), also published
by National Academies Press, 1998. http://books.nap.edu/books/0309060966/html/index.html
