Task Decomposition and Dynamic Role Assignment for Real-Time Strategic Teamwork

Multi-agent domains consisting of teams of agents that need to collaborate in an adversarial environment offer challenging research opportunities. In this paper, we introduce periodic team synchronization domains, as timecritical environments in which agents act autonomouslywith limited communication, but they can periodically synchronize in a full-communication setting. We present a team agent structure that allows for an agent to capture and reason about team agreements. We achieve collaboration between agents through the introduction of formations. A formation decomposes the task space defining a set of roles. Homogeneous agents can flexibly switch roles within formations, and agents can change formations dynamically, according to pre-defined triggers to be evaluated at run-time. This flexibility increases the performance of the overall team. Our team structure further includes pre-planning for frequent situations. We fully implemented this approach in the domain of robotic soccer. Our simulator team made it to the semi-finals of the RoboCup-97 competition, in which 29 teams participated. It achieved a total score of 67-9 over six different games, and successfully demonstrated its flexible team structure. Using the same team structure, our small robot team won the RoboCup-97 small-robot competition, in which 4 teams participated. It achieved a total score of 13-1 over 4 games and also demonstrated its flexible team structure.

[1]  Manuela M. Veloso,et al.  Layered Approach to Learning Client Behaviors in the Robocup Soccer Server , 1998, Appl. Artif. Intell..

[2]  Manuela M. Veloso,et al.  Communication in Domains with Unreliable, Single-Channel, Low-Bandwidth Communication , 1998, CRW.

[3]  Luke Hunsberger.  Making SharedPlans more concise and easier to reason about , 1998, Proceedings International Conference on Multi Agent Systems (Cat. No.98EX160).

[4]  Peter Stone,et al.  CMUnited: a team of robotics soccer agents collaborating in an adversarial environment , 1998, CROS.

[5]  Michael J. Prietula,et al.  Simulating organizations: computational models of institutions and groups , 1998 .

[6]  Hiroaki Kitano,et al.  RoboCup: A Challenge Problem for AI , 1997, AI Mag..

[7]  Maja J. Mataric,et al.  Interaction and intelligent behavior , 1994 .

[8]  Manuela M. Veloso,et al.  The CMUnited-97 Simulator Team , 1997, RoboCup.

[9]  Michael L. LaBlanc,et al.  The World Encyclopedia of Soccer , 1993 .

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

[11]  Anup Basu,et al.  Computer Vision: Systems, Theory and Applications , 1993, World Scientific Series in Computer Science.

[12]  Rafael H. Bordini,et al.  Moral Sentiments in Multi-Agent Systems , 1998, ATAL.

[13]  Jörg P. Müller,et al.  The Right Agent (Architecture) to do the Right Thing , 1998, ATAL.

[14]  Hitoshi Matsubara,et al.  Soccer Server and Researches on Multi-Agent Systems , 1996 .

[15]  Alan K. Mackworth,et al.  Real-time control of soccer-playing robots using off-board vision: the dynamite testbed , 1995, 1995 IEEE International Conference on Systems, Man and Cybernetics. Intelligent Systems for the 21st Century.

[16]  Manuela M. Veloso,et al.  Towards collaborative and adversarial learning: a case study in robotic soccer , 1998, Int. J. Hum. Comput. Stud..

[17]  B. Grosz.  Collaborative Systems , 1996 .

[18]  M. Veloso,et al.  The Cmunited-97 Simulator Team in Robocup-97: Robot Soccer World Cup I , 1998 .

[19]  Manuela M. Veloso,et al.  Using Decision Tree Confidence Factors for Multiagent Control , 1997, RoboCup.

[20]  Manuela M. Veloso,et al.  The CMUnited-97 Small Robot Team , 1997, RoboCup.

[21]  Sascha Ossowski,et al.  Social Structure in Artificial Agent Societies: Implications for Autonomous Problem-Solving Agents , 1998, ATAL.

[22]  Minoru Asada,et al.  Coordination of multiple behaviors acquired by a vision-based reinforcement learning , 1994, Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS'94).

[23]  Frank Dignum,et al.  Autonomous Norm Acceptance , 1998, ATAL.

[24]  Hiroaki Kitano,et al.  RoboCup-97: The First Robot World Cup Soccer Games and Conferences , 1998, AI Mag..

[25]  Cristiano Castelfranchi,et al.  Commitments: From Individual Intentions to Groups and Organizations , 1995, ICMAS.

[26]  Barbara J. Grosz,et al.  Collaborative Systems (AAAI-94 Presidential Address) , 1996 .

[27]  Manuela M. Veloso,et al.  Using decision tree confidence factors for multi-agent control , 1998, AGENTS '98.

[28]  Luke Hunsberger.  Making Shared Plans More Concise and Easier to Reason About , 1998, ATAL.

[29]  Keith S. Decker,et al.  Task environment centered simulation , 1998 .

[30]  Randy Sargent,et al.  Dynamic Object Capture Using Fast Vision Tracking , 1997, AI Mag..

[31]  Alan K. Mackworth.  On Seeing Robots , 1993, Computer Vision: Systems, Theory and Applications.