Robust Agent Teams via Socially-Attentive Monitoring

Agents in dynamic multi-agent environments must monitor their peers to execute individual and group plans. A key open question is how much monitoring of other agents' states is required to be effective: The Monitoring Selectivity Problem. We investigate this question in the context of detecting failures in teams of cooperating agents, via Socially-Attentive Monitoring, which focuses on monitoring for failures in the social relationships between the agents. We empirically and analytically explore a family of socially-attentive teamwork monitoring algorithms in two dynamic, complex, multi-agent domains, under varying conditions of task distribution and uncertainty. We show that a centralized scheme using a complex algorithm trades correctness for completeness and requires monitoring all teammates. In contrast, a simple distributed teamwork monitoring algorithm results in correct and complete detection of teamwork failures, despite relying on limited, uncertain knowledge, and monitoring only key agents in a team. In addition, we report on the design of a socially-attentive monitoring system and demonstrate its generality in monitoring several coordination relationships, diagnosing detected failures, and both on-line and off-line applications.

[1]  Nicholas R. Jennings,et al.  Commitments and conventions: The foundation of coordination in multi-agent systems , 1993, The Knowledge Engineering Review.

[2]  Milind Tambe,et al.  I'm OK, you're OK, we're OK: experiments in distributed and centralized socially attentive monitoring , 1999, AGENTS '99.

[3]  Austin Tate,et al.  Synthesizing Protection Monitors from Causal Structure , 1994, AIPS.

[4]  Sarit Kraus,et al.  Reaching Agreements Through Argumentation: A Logical Model and Implementation , 1998, Artif. Intell..

[5]  Yaser Al-Onaizan,et al.  On being a teammate: experiences acquired in the design of RoboCup teams , 1999, AGENTS '99.

[6]  R. James Firby,et al.  An Investigation into Reactive Planning in Complex Domains , 1987, AAAI.

[7]  A. Newell.  Unified Theories of Cognition , 1990 .

[8]  Sarit Kraus,et al.  Coordination without Communication: Experimental Validation of Focal Point Techniques , 1997, ICMAS.

[9]  Candace L. Sidner,et al.  Using plan recognition in human-computer collaboration , 1999 .

[10]  Manuela M. Veloso,et al.  Rationale-Based Monitoring for Planning in Dynamic Environments , 1998, AIPS.

[11]  Hiroaki Kitano,et al.  The RoboCup Synthetic Agent Challenge 97 , 1997, IJCAI.

[12]  Milind Tambe,et al.  Intelligent Agents for Interactive Simulation Environments , 1995, AI Mag..

[13]  Sarit Kraus,et al.  Collaborative Plans for Complex Group Action , 1996, Artif. Intell..

[14]  Edmund H. Durfee,et al.  Expecting the Unexpected: Detecting and Reacting to Unplanned-for World States , 1996, AAAI/IAAI, Vol. 2.

[15]  Bryan Horling,et al.  Diagnosis as an integral part of multi-agent adaptability , 2000, Proceedings DARPA Information Survivability Conference and Exposition. DISCEX'00.

[16]  Tomohito Andou,et al.  Refinement of Soccer Agents' Positions Using Reinforcement Learning , 1997, RoboCup.

[17]  Paul R. Cohen,et al.  Early Warnings of Plan Failure, False Positives and Envelopes, Experiments and a Model , 1992 .

[18]  Candace L. Sidner,et al.  COLLAGEN: when agents collaborate with people , 1997, AGENTS '97.

[19]  Paul E. Spector,et al.  The Impact of Cross-Training on Team Functioning: An Empirical Investigation , 1996, Hum. Factors.

[20]  Joseph Y. Halpern,et al.  Knowledge and common knowledge in a distributed environment , 1984, JACM.

[21]  Maja J. Mataric,et al.  Interference as a Tool for Designing and Evaluating Multi-Robot Controllers , 1997, AAAI/IAAI.

[22]  Gerd Wagner,et al.  Distributed diagnosis by vivid agents , 1997, AGENTS '97.

[23]  David Atkinson,et al.  Generating Perception Requests and Expectations to Verify the Execution of Plans , 1986, AAAI.

[24]  W. Lewis Johnson,et al.  Steve: an animated pedagogical agent for procedural training in virtual environments , 1997, SGAR.

[25]  Richard Washington,et al.  Markov tracking for agent coordination , 1998, AGENTS '98.

[26]  Milind Tambe,et al.  Tracking Dynamic Team Activity , 1996, AAAI/IAAI, Vol. 1.

[27]  Edmund H. Durfee,et al.  On Acting Together: Without Communication , 1995 .

[28]  Milind Tambe,et al.  What Is Wrong With Us? Improving Robustness Through Social Diagnosis , 1998, AAAI/IAAI.

[29]  T. A. Harris,et al.  I'm OK, You're OK , 1967 .

[30]  Anand S. Rao,et al.  Means-End Plan Recognition - Towards a Theory of Reactive Recognition , 1994, KR.

[31]  Masaru Ishii,et al.  Cooperation by observation: the framework and basic task patterns , 1994, Proceedings of the 1994 IEEE International Conference on Robotics and Automation.

[32]  Hector J. Levesque,et al.  On Acting Together , 1990, AAAI.

[33]  Nicholas R. Jennings,et al.  Controlling Cooperative Problem Solving in Industrial Multi-Agent Systems Using Joint Intentions , 1995, Artif. Intell..

[34]  Gregory D. Hager,et al.  If at First You Don't Succeed , 2007, AAAI/IAAI.

[35]  Luca Console,et al.  Readings in Model-Based Diagnosis , 1992 .

[36]  Edmund H. Durfee,et al.  Blissful Ignorance: Knowing Just Enough to Coordinate Well , 1995, ICMAS.

[37]  Sarit Kraus,et al.  The Evolution of Sharedplans , 1999 .

[38]  Gil Tidhar,et al.  Planned Team Activity , 1992, MAAMAW.

[39]  Milind Tambe,et al.  Towards Flexible Teamwork , 1997, J. Artif. Intell. Res..

[40]  Lynne E. Parker.  Designing control laws for cooperative agent teams , 1993, [1993] Proceedings IEEE International Conference on Robotics and Automation.

[41]  C. Sidner,et al.  Plans for Discourse , 1988 .

[42]  M. F. Mar,et al.  ModSAF Behavior Simulation and Control , 1993 .

[43]  Sam Steel,et al.  Integrating Planning, Execution and Monitoring , 1988, AAAI.

[44]  R. Arkin,et al.  Behavioral diversity in learning robot teams , 1998 .

[45]  K. Bailey.  Social Entropy Theory , 1990 .