Estimating Value of Assistance for Online POMDP Robotic Agents

Yuval Goshen, Sarah Keren
Proceedings of The 9th Conference on Robot Learning, PMLR 305:1079-1101, 2025.

Abstract

Robotic agents operating in dynamic, partially observable environments often benefit from teammate assistance. We address the challenge of determining when and how to assist in multi-robot systems where agents can modify the physical environment, such as moving obstacles that block perception or manipulation. For robots using online POMDP planning, evaluating assistance impacts requires computationally intensive policy evaluation, making real-time decisions difficult. We formulate Value of Assistance (VOA) for POMDP agents and develop efficient heuristics that approximate VOA without requiring complete policy evaluation. Our empirical evaluation on both a standard POMDP benchmark and a collaborative manipulation task demonstrates that our Full Information heuristic enables real-time assistance decisions while maintaining sufficient accuracy for effective helping action selection.

Cite this Paper

BibTeX
@InProceedings{pmlr-v305-goshen25a, title = {Estimating Value of Assistance for Online POMDP Robotic Agents}, author = {Goshen, Yuval and Keren, Sarah}, booktitle = {Proceedings of The 9th Conference on Robot Learning}, pages = {1079--1101}, year = {2025}, editor = {Lim, Joseph and Song, Shuran and Park, Hae-Won}, volume = {305}, series = {Proceedings of Machine Learning Research}, month = {27--30 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v305/main/assets/goshen25a/goshen25a.pdf}, url = {https://proceedings.mlr.press/v305/goshen25a.html}, abstract = {Robotic agents operating in dynamic, partially observable environments often benefit from teammate assistance. We address the challenge of determining when and how to assist in multi-robot systems where agents can modify the physical environment, such as moving obstacles that block perception or manipulation. For robots using online POMDP planning, evaluating assistance impacts requires computationally intensive policy evaluation, making real-time decisions difficult. We formulate Value of Assistance (VOA) for POMDP agents and develop efficient heuristics that approximate VOA without requiring complete policy evaluation. Our empirical evaluation on both a standard POMDP benchmark and a collaborative manipulation task demonstrates that our Full Information heuristic enables real-time assistance decisions while maintaining sufficient accuracy for effective helping action selection.} }
Endnote
%0 Conference Paper %T Estimating Value of Assistance for Online POMDP Robotic Agents %A Yuval Goshen %A Sarah Keren %B Proceedings of The 9th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2025 %E Joseph Lim %E Shuran Song %E Hae-Won Park %F pmlr-v305-goshen25a %I PMLR %P 1079--1101 %U https://proceedings.mlr.press/v305/goshen25a.html %V 305 %X Robotic agents operating in dynamic, partially observable environments often benefit from teammate assistance. We address the challenge of determining when and how to assist in multi-robot systems where agents can modify the physical environment, such as moving obstacles that block perception or manipulation. For robots using online POMDP planning, evaluating assistance impacts requires computationally intensive policy evaluation, making real-time decisions difficult. We formulate Value of Assistance (VOA) for POMDP agents and develop efficient heuristics that approximate VOA without requiring complete policy evaluation. Our empirical evaluation on both a standard POMDP benchmark and a collaborative manipulation task demonstrates that our Full Information heuristic enables real-time assistance decisions while maintaining sufficient accuracy for effective helping action selection.
APA
Goshen, Y. & Keren, S.. (2025). Estimating Value of Assistance for Online POMDP Robotic Agents. Proceedings of The 9th Conference on Robot Learning, in Proceedings of Machine Learning Research 305:1079-1101 Available from https://proceedings.mlr.press/v305/goshen25a.html.

Related Material