Modeling Teamwork in MultiAgent Systems: The CAST Architecture

1
Modeling Teamwork in Multi-Agent Systems The
CAST Architecture

• Dr. Thomas Ioerger, Jianwen Yin, and Michael
  Miller
• Computer Science, Texas AM University
• March 21, 2001

2
Agenda

• Teamwork
• CAST
• MALLET
• DIARG
• Dynamic Role Selection
• Conclusions

3
What is a Team?

• Definition of a Team
• A group of entities that are working together to
  achieve a shared (or common) goal
• Team needs
• Coordination
• Sharing of information
• Distributed decision-making
• Focus of (human) teamwork optimizations
• Situational awareness
• Communications efficiency
• Effectiveness of group decision-making

4
Previous work on Teamwork

• Multi-agent teams
• Cohen Levesques joint intention theory
• Planning and plan specification/recognition
• Team-oriented programming
• STEAM based on SOAR by Tambe based on
  joint-intentions
• Groszs Shared Plans
• Psy. Research on Human Team-Training
• Shared mental models

5
Issues in Multi-Agent Team Training System Design

• How to represent a team? (Team ontology and
  MALLET)
• How to effectively reason about other team
  members roles and beliefs to collaborate?
• How to initiate communication? (DIARG)

6
Shared Mental Model

• Team structure roles
• responsibilities for individual steps in plans
• mutual belief assumption
• Team process team plan
• use Petri Nets as an approximate finite and
  computable model of team process
• use the token flow in Petri Nets to monitor and
  track the plan execution)

B
A
start
D
end
C
7
From teamwork to agent model

• A compositional multi-agent architecture

8
CAST (Collaborative Agents for Simulating
Teamwork)

• Model effective teamwork by capturing team
  structures and teamwork process
• Enable agents in a team to have flexibility for
  adapting the team to changes in the environment

9
CAST Architecture
10
MALLET (Multi-Agent Logic Language for Encoding
Teamwork)

• Basic Object Type Predicates
• Actions and Plans
• Responsibilities
• Capabilities
• Belief
• Communication

11
MALLET Specification

• Team plans
• role specifications (with constraints)
• process specification

(team-plan make-dinner (?entrée) (role shopper
?X (has-money ?X)) (role cook ?Y
) (process (if (need-ingredients
?entree) (sequence (do ?X (go-to store))
(do ?X (buy-ingredients ?entrée))) (parall
el (do ?Y (prepare ?entrée)) (do ?X
set-table)) (forall ?Z (team-member ?Z) (do ?Z
eat)) (do-all (clean-up dishes)))) // an AND
team-operator
12
Team Operators

• Joint actions coop-mode
• and lifting a couch together
• xor/or hitting a volleyball
• Team operator
• team, co-mode, arguments
• precond, postcond

13
Responsibilities

• Responsibilities are relationships between a set
  of roles, a goal, and a stage, represented as
  Responsibility(ltrole-setgt, ltgoalgt, stage)
• Redundant responsibilities Or-
  responsibility(ltrole-setgt, ltgoalgt, stage)
• Shared competitive responsibilities
  Xor-responsibility(ltrole-setgt, ltgoalgt, stage)
• Shared complementary responsibilities And-
  responsibility(ltrole-setgt, ltgoalgt, stage)

14
Capabilities

• Capabilities are relationships between roles and
  actions, which is represented as
  Capability(ltrole-setgt, ltactiongt)
• Backup capability Or- capability(ltrole-setgt,
  ltactiongt)
• Shared competitive capability Xor-
  capability(ltrole-setgt, ltactiongt)
• Shared conflicting capability And-
  capability(ltrole-setgt, ltactiongt)

15
Three Major Algorithms

• Petri Net generation algorithm
• DIARG
• Dynamic role selection

16
Petri Nets

• place/transition Petri nets vs. PrT nets
• propositional ? first-order predicate
• PrT nets vs. logic programs
• happy(x) ? cat(x), purrs(x).
• Concurrence
• Firing rule vs. meta-predicates

17
Petri Net Generation

• Sequential process
• Parallel process

18
Petri Net Generation (Cont.)

• If process

sub-net
Then cond
else cond
sub-net
start
end
19
DIARG (Dynamic Inter-Agent Rule Generator)

• Proactive information exchange
• inferring what might be needed for others to
  complete their (current) tasks
• Proactively provide timely information
• Reduce communication overhead
• Examines needers and providers of information in
  order to allow dynamic generation of information
  flow

20
Dynamic Role Selection

• Whenever there is a choice on who performs an
  action (i.e.role redundancy)
• Postpone role selection until reach the
  individual plan or individual operator level
• Might need to communicate to
• resolve ambiguities of responsibility
• synchronize for joint actions
• Advantages
• Provide flexibility in team formation
• Others can play backup role, in case of failure
• Improve teamwork efficiency, Balance workload

21
Future Work

• Simulating more dynamic teams
• handling failures, load-balancing, delegation,
  unreliable agents
• Handling more types of communication
• resolving conflicting information
• situation assessment (cognitive models)
• Reasoning about agent capabilities
• Goal regression and planning
• User-modeling inferring state/beliefs/intent
  from monitoring actions

22
Conclusions

• First version (CAST 1.0) running now
• see demo
• Target domains Team Training
• AWACS
• Fire Fighting
• Army Brigade Operations Staff
• NASA Flight Controllers

23
Acknowledgements

• Dr. Richard Volz, Dr. John Yen, Dr. Dianxiang Xu
• Heejin Lim, Sen Cao, Yue Zhou, Xueqi Cheng, Colby
  Johnson