Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences...
-
Upload
reynold-banks -
Category
Documents
-
view
217 -
download
3
Transcript of Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences...
![Page 1: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/1.jpg)
Soar: Soar: An Architecture forAn Architecture for
Human Behavior RepresentationHuman Behavior Representation
Randall W. Hill, Jr.
Information Sciences Institute
University of Southern Californiahttp://www.isi.edu/soar/hill
![Page 2: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/2.jpg)
What is Soar?What is Soar?
Artificial Intelligence Architecture– System for building intelligent agents
– Learning system
Cognitive Architecture– A candidate Unified Theory of Cognition
(Allen Newell, 1990)
![Page 3: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/3.jpg)
HistoryHistory
Inventors– Allen Newell, John Laird, Paul Rosenbloom
Officially created in 1983– Roots in 1950’s and onwards
Currently on version 8 of Soar architecture– Written in ANSI C for portability and speed
In the public domain
![Page 4: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/4.jpg)
User CommunityUser Community
Academia– USC, U. of Michigan, CMU, BYU, others
International– Britain, Europe, Japan
Commercial– Soar Technology, Inc.– ExpLore Reasoning Systems, Inc.
![Page 5: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/5.jpg)
Objectives of ArchitectureObjectives of Architecture
Support multi-method problem solving– Apply to a wide variety of tasks and methods – Combine reactive and goal directed symbolic processing
Represent and use multiple knowledge forms– Procedural, declarative, episodic, iconic– Support very large bodies of knowledge (>100,000 rules)
Interact with the outside world Learn about all aspects of tasks
![Page 6: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/6.jpg)
Cognitive Behavior:Cognitive Behavior:Underlying AssumptionsUnderlying Assumptions
Goal-oriented Reactive Requires use of symbols Problem space hypothesis Requires learning
![Page 7: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/7.jpg)
Soar ArchitectureSoar Architecture
Working Memorysituational assessment, intermediate results, actions, goals, …
Long Term Knowledgee.g., Doctrine, Tactics, Flying Techniques,
Missions, Coordination,Properties of Planes, Weapons, Sensors, …
[ ][ ][ ]
[ ][ ][ ]
Match Changes
Perception / Motor Interface
![Page 8: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/8.jpg)
Soar Decision CycleSoar Decision CyclePerception Cognition Motor
Input Phase
Elaboration Phase
Output Phase
Decision Phase
• Fire rules
• Generate preferences
• Update working memory
• Evaluate operator preferences
• Select new operator OR
• Create new state
• Sense world
• Perceptual pre-processing
• Assert to WM
• Command effectors
• Adjust perception
![Page 9: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/9.jpg)
Which Rule(s) Should Fire?Which Rule(s) Should Fire? Fire all matched rules in parallel until quiescence Sequential operators generate behavior
– e.g., Turn, adjust-radar, select-missile, climb
– Provides trace of behavior comparable to human actions
Rules select, apply, terminate operators.– Select: create preferences to propose and compare operators
– Apply: modify the current situation, send motor commands
– Terminate: determine that operator is finished
Inp
ut
Elaboration(propose operators)
Decide(select operator)
Elaboration(apply operator)
Ou
tpu
t
Inp
ut
Dec
ide
Ou
tpu
t
Inp
ut
Dec
ide
Elaboration(terminate operator & propose)
![Page 10: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/10.jpg)
Example RulesExample Rules
PROPOSE: If I encounter the enemy, propose an operator to break contact with the enemy.
SELECT: If I am enroute to my holding area and I come into contact with an enemy unit, prefer breaking contact over engaging targets.
APPLY: If the enemy is to my left, break to the right.
APPLY: If the enemy is to my right, break to the left.
TERMINATE: If break contact is the current operator, and contact is broken, then terminate break operator.
![Page 11: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/11.jpg)
Goal Driven BehaviorGoal Driven Behavior
Complex operators are decomposed to simpler ones– Occurs whenever rules are insufficient to apply operator
– Decomposition is dynamic and situation dependent
– Over 90 operators in RWA-Soar
Execute-Mission
Fly-Flight-Plan Engage Prepare-to-return-to-base
Fly-control-route Select-point
Select-route
High-level
Low-level
Contour NOE
Mask Unmask Employ-weapons
Initialize-hover
Return-to-control-point
![Page 12: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/12.jpg)
Coordination of Coordination of Behavior & ActionBehavior & Action
Combines goal-driven and reactive behaviors– Suggest new operators anywhere in goal hierarchy
– Generate preferences for operators
– Terminate operators
Provides limited multi-task capability– Constrained by single goal hierarchy in Soar
Other possible approaches– Multiple goal hierarchies
– Flush and re-build goal hierarchies when needed
![Page 13: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/13.jpg)
ModelingModelingPerceptual Perceptual
AttentionAttention
Problem
• Naïve vision model— Entity-level resolution
— Unrealistic field of view (360o, 7 km)
• No focus of attention— Perceptual overload often occurs
— Pilot crashes helicopter
Approach
• Zoom lens model of attention— Gestalt grouping in pre-attentive stage
— Multi-resolution focus
• Control of attention — Goal-driven: task-based, group-oriented
— Stimulus-driven: abrupt onset, contrast
![Page 14: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/14.jpg)
Model of Attention• Gestalt grouping
• Zoom lens effect
• Goal and stimulus driven
Naïve Vision Model• Entity-oriented
• Stimulus-driven
• No perceptual control
![Page 15: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/15.jpg)
Underlying Underlying Technologies/AlgorithmsTechnologies/Algorithms
Optimized RETE algorithm– Enables efficient matching in large rule sets
Universal subgoaling– Operator-based architecture– Truth Maintenance System (TMS)
Learning algorithm– Chunking mechanism
![Page 16: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/16.jpg)
Soar ApplicationsSoar Applications
Agents for Synthetic Battlespaces– Commanders and Helicopter Pilots (USC)
– Fixed Wing Aircraft Pilots (UM, Soar Technology)
Animated, Pedagogical Agents– Steve (Rickel and Johnson, USC)
Game Agents– Quake (Laird and van Lent, UM)
![Page 17: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/17.jpg)
Intelligent Synthetic ForcesIntelligent Synthetic Forces
Helicopter pilots Teamwork C3I Modeling
– Planning– Execution– Re-planning– Collaboration
![Page 18: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/18.jpg)
Steve: An Embodied Intelligent Steve: An Embodied Intelligent Agent for Virtual EnvironmentsAgent for Virtual Environments
3D agent that interacts with students in virtual environments
Can take different roles: teammate, teacher, guide, demonstrator
Multiple trainees and agents work together in virtual teams
Intelligent tutoring in the context of a shared team environment
![Page 19: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/19.jpg)
Soar/Games ProjectSoar/Games Project Build an AI Engine around the Soar AI architecture
– Soar/Quake II project– Soar/Descent 3 project
U. of Michigan, Laird and van Lent
InterfaceDLL
Soar/QuakeAI
AI Engine(Soar)
KnowledgeFiles
Actions
Sensor Data
Socket
![Page 20: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/20.jpg)
Validation EffortsValidation Efforts
Intelligent Synthetic Forces– Synthetic Theater of War ‘97 experience– Subject Matter Experts
Human Factors / HCI studies– e.g., B. John (CMU) & R. Young (U.K.)
Better models for validating integrated models of human behavior needed
![Page 21: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/21.jpg)
Summary of Summary of Capabilities/LimitationsCapabilities/Limitations
Capabilities– Mixes goal-oriented and reactive behavior– Supports interaction with external world– Architecture lends itself to creating integrated
models of human behavior Limitations
– Learning mechanism not easily used
![Page 22: Soar: An Architecture for Human Behavior Representation Randall W. Hill, Jr. Information Sciences Institute University of Southern California .](https://reader035.fdocuments.in/reader035/viewer/2022062517/56649f395503460f94c55b20/html5/thumbnails/22.jpg)
Future Development /Future Development /Application PlansApplication Plans
Integrate emotion with cognition Learn from experience
– Incorporate inductive models of learning Continue work on models of collaboration
in complex decision-making– Extend the current C3I models