TY - GEN

T1 - Policy synthesis for factored MDPs with graph temporal logic specifications

AU - Cubuktepe, Murat

AU - Xu, Zhe

AU - Topcu, Ufuk

N1 - Funding Information:
Partially funded by the grants AFRL FA9550-19-1-0169, and ONR N00014-18-1-2829. Proc. of the 19th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2020), B. An, N. Yorke-Smith, A. El Fallah Seghrouchni, G. Sukthankar (eds.), May 9–13, 2020, Auckland, New Zealand. © 2020 International Foundation for Autonomous Agents and Multiagent Systems (www.ifaamas.org). All rights reserved.
Publisher Copyright:
© 2020 International Foundation for Autonomous.

PY - 2020

Y1 - 2020

N2 - We study the synthesis of policies for multi-agent systems to implement spatial-temporal tasks. We formalize the problem as a factored Markov decision process subject to so-called graph temporal logic specifications. The transition function and the spatial-temporal task of each agent depend on the agent itself and its neighboring agents. The structure in the model and the specifications enable to develop a distributed algorithm that, given a factored Markov decision process and a graph temporal logic formula, decomposes the synthesis problem into a set of smaller synthesis problems, one for each agent. We prove that the algorithm runs in time linear in the total number of agents. The size of the synthesis problem for each agent is exponential only in the number of neighboring agents, which is typically much smaller than the number of agents. We demonstrate the algorithm in case studies on disease control and urban security. The numerical examples show that the algorithm can scale to hundreds of agents.

AB - We study the synthesis of policies for multi-agent systems to implement spatial-temporal tasks. We formalize the problem as a factored Markov decision process subject to so-called graph temporal logic specifications. The transition function and the spatial-temporal task of each agent depend on the agent itself and its neighboring agents. The structure in the model and the specifications enable to develop a distributed algorithm that, given a factored Markov decision process and a graph temporal logic formula, decomposes the synthesis problem into a set of smaller synthesis problems, one for each agent. We prove that the algorithm runs in time linear in the total number of agents. The size of the synthesis problem for each agent is exponential only in the number of neighboring agents, which is typically much smaller than the number of agents. We demonstrate the algorithm in case studies on disease control and urban security. The numerical examples show that the algorithm can scale to hundreds of agents.

UR - http://www.scopus.com/inward/record.url?scp=85089573987&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85089573987&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85089573987

T3 - Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, AAMAS

SP - 267

EP - 275

BT - Proceedings of the 19th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2020

A2 - An, Bo

A2 - El Fallah Seghrouchni, Amal

A2 - Sukthankar, Gita

PB - International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS)

T2 - 19th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2020

Y2 - 19 May 2020

ER -