Cooperative behaviour of agents within highly dynamic and nondeterministic domains is an active field of research. In particular establishing responsive teamwork, where agents are able to react to dynamic changes in the environment while facing unreliable communication and sensory noise, is an open problem. Unexpectedly changing situations force agents to react and adapt under tight time-constraints. Hence they often cannot communicate or agree upon their decisions before acting upon them. We present a novel model for cooperative behaviour geared towards such domains. In our approach, the agents estimate each other’s decision and correct these estimations once they receive contradictory information. We aim at a comprehensive approach for agent teamwork featuring intuitive modelling capabilities for multi-agent activities, abstractions over activities and agents, and clear operational semantics for the new model. We show experimentally that the resulting behaviour stabilises towards teamwork and can achieve a cooperative goal.
[1]
Philip R. Cohen,et al.
Plans for Discourse
,
2003
.
[2]
Koen V. Hindriks,et al.
Agent Programming in 3APL
,
1999,
Autonomous Agents and Multi-Agent Systems.
[3]
Michael Wagner,et al.
A Language for Interactive Cooperative Agents
,
2009
.
[4]
Sarit Kraus,et al.
Collaborative Plans for Complex Group Action
,
1996,
Artif. Intell..
[5]
Milind Tambe,et al.
Toward Team-Oriented Programming
,
1999,
ATAL.
[6]
Milind Tambe,et al.
Towards Flexible Teamwork
,
1997,
J. Artif. Intell. Res..
[7]
Hector J. Levesque,et al.
On Acting Together
,
1990,
AAAI.