TY - GEN
T1 - A factored approach to deterministic contingent multi-agent planning
AU - Shekhar, Shashank
AU - Brafman, Ronen I.
AU - Shani, Guy
N1 - Publisher Copyright:
© 2019 Association for the Advancement of Artificial Intelligence. All rights reserved.
PY - 2019/1/1
Y1 - 2019/1/1
N2 - Collaborative Multi-Agent Planning (MAP) under uncertainty with partial observability is a notoriously difficult problem. Such MAP problems are often modeled as Dec- POMDPs, or its qualitative variant, QDec-POMDP, which is essentially a MAP version of contingent planning. The QDec- POMDP model was introduced with the hope that its simpler, non-probabilistic structure will allow for better scalability. Indeed, at least with deterministic actions, the recent IMAP algorithm scales much better than comparable Dec- POMDP algorithms (Bazinin and Shani 2018). In this work we suggest a new approach to solving Deterministic QDec- POMDPs based on problem factoring. First, we find a solution to a MAP problem where the results of any observation is available to all agents. This is essentially a single-agent planning problem for the entire team. Then, we project the solution tree into sub-trees, one per agent, and let each agent transform its projected tree into a legal local tree. If all agents succeed, we combine the trees into a valid joint-plan. Otherwise, we continue to explore the space of team solutions. This approach is sound, complete, and as our empirical evaluation demonstrates, scales much better than the IMAP algorithm.
AB - Collaborative Multi-Agent Planning (MAP) under uncertainty with partial observability is a notoriously difficult problem. Such MAP problems are often modeled as Dec- POMDPs, or its qualitative variant, QDec-POMDP, which is essentially a MAP version of contingent planning. The QDec- POMDP model was introduced with the hope that its simpler, non-probabilistic structure will allow for better scalability. Indeed, at least with deterministic actions, the recent IMAP algorithm scales much better than comparable Dec- POMDP algorithms (Bazinin and Shani 2018). In this work we suggest a new approach to solving Deterministic QDec- POMDPs based on problem factoring. First, we find a solution to a MAP problem where the results of any observation is available to all agents. This is essentially a single-agent planning problem for the entire team. Then, we project the solution tree into sub-trees, one per agent, and let each agent transform its projected tree into a legal local tree. If all agents succeed, we combine the trees into a valid joint-plan. Otherwise, we continue to explore the space of team solutions. This approach is sound, complete, and as our empirical evaluation demonstrates, scales much better than the IMAP algorithm.
UR - http://www.scopus.com/inward/record.url?scp=85085592729&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85085592729
T3 - Proceedings International Conference on Automated Planning and Scheduling, ICAPS
SP - 419
EP - 427
BT - Proceedings of the 29th International Conference on Automated Planning and Scheduling, ICAPS 2019
A2 - Benton, J.
A2 - Lipovetzky, Nir
A2 - Onaindia, Eva
A2 - Smith, David E.
A2 - Srivastava, Siddharth
PB - AAAI press
T2 - 29th International Conference on Automated Planning and Scheduling, ICAPS 2019
Y2 - 11 July 2019 through 15 July 2019
ER -