DAGSIAFRL Project AgentBased MixedInitiative Collaboration WSU Role JerSen Chen, Michael T' Cox,

1
DAGSI/AFRL ProjectAgent-Based Mixed-Initiative
CollaborationWSU RoleJer-Sen Chen, Michael T.
Cox, S. NarayananWright State University
2
Outline

• Overall Project Goals
• Interrelated Research Areas
• WSU Participants
• Background, Interests, Expertise
• Summary of Prior Research
• Research Components Role of the Participants

3
Project Goals

• To investigate agent-based architectures for
  distributed collaborative planning
• Framing collaboration as a distributed
  mixed-initiative planning process in a
  goal-centered framework
• Modeling users for human-centered information
  systems development
• Designing, implementing, and evaluating agent
  architectures that support collaborative planning
  in prototypical problems in the Air Force and the
  commercial sector

4
Characteristics of Intelligent Collaboration

• Dynamic Environments
• Distributed Cognitive Activity
• Complex Human-Machine Interaction
• Complex Human-Information Interaction

5
Interrelated Research Areas

• Distributed Mixed-Initiative Planning
• Tailored Information Presentation
• Agent Architectures
• User Modeling

6
WSU Participants

• Jer-Sen Chen, Computer Science
• Michael R. Cox, Computer Science
• S. Narayanan, Industrial Human Factors
  Engineering
• Several Students with Interdisciplinary
  Backgrounds

7
Jer-Sen Chen, Ph.D.

• Research Assistant Professor in Computer Science
• Areas Interactive Visualization, HCI, Human
  Visual Perception
• Projects
• ONR Assembly Planning in Manufacturing
• OBR NSF HCI Workbench
• AFOSR Human Visual Perception

8
Michael T. Cox, Ph.D.

• Assistant Professor in Computer Science
• Theme How Goals Interact with and Influence
  Cognitive Processes in Planning, Understanding,
  Learning
• Areas Case-Based Reasoning, Mixed-Initiative
  Planning, HCI, Automated Understanding, Machine
  Learning

9
Mike Cox (Projects)

• DARPA JADE (Joint Assistant for Deployment
  Execution) - Developed Mixed-Initiative Component
• AFOSR Large-Scale, Multi-Agent, Distributed
  Mission Planning and Execution in Complex Dynamic
  Environments (Constraint-Based Reasoning,
  Autonomous Agent Emphasis)
• GT Meta-AQUA Meta-TS Instrospective
  Multi-Strategy Learning

10
S. Narayanan, Ph.D., P.E.

• Associate Professor in Industrial Human Factors
  Engineering
• Theme Modeling Analysis of Complex Systems
  Understanding Aiding Human Decision Making
  through Computational Tools
• Areas Cognitive Modeling, Interactive
  Simulation, Applied AI

11
Narayanan ( Related Projects)

• AFOSR/AFRL JADIS JADIS-WEB, Interactive
  Modeling Simulation of AF Logistic Systems
• LEXIS-NEXIS Modeling Information Seeking
• AFRL Adaptive Agent Architecture in Supervisory
  Control Systems
• AFRL UMAST
• GT Meta-TS, OOSIM
• DAGSI/AFRL Integrating Optimization,
  Simulations, Human Problem Solving in AF
  Logistics to Support EAF

12
WSU Role in Research Components

• Jer-SenChen
• Information Manager, Agent Manager, Interface
  Usability
• Interactive Visualizations
• GUI Development

13
WSU Participants Role

• Mike Cox
• Plan Manager, Information Manager, Apply to Air
  Campaign Planning, Apply to ATM Scheduling,
  Computational Studies on Planning, Interface
  Usability
• Overall Project Coordination
• Theoretical Applied Issues in Goal-Directed
  Dynamic Control of Distributed Planning
  Information Presentation

14
WSU Role in Research Components

• S. Narayanan
• Plan Manager, Information Manager, Apply to Air
  Campaign Planning, Empirical Evaluating Using
  Humans, Computational Studies on Planning,
  Interface Usability
• Studies on Effective Collaboration in Natural
  Systems How to Translate it to Human-Agent
  Collaboration
• Simulation Cognitive User Modeling for Tailored
  Information Presentation

15
Finally

• DAGSI/AFRL grant is for fundamental research
• Keys are flexibility, open-mindedness, room to
  improvise under the overall theme as we make the
  research progress