Research

Paper

AI LLM March 04, 2026

Large-Language-Model-Guided State Estimation for Partially Observable Task and Motion Planning

Authors

Yoonwoo Kim, Raghav Arora, Roberto Martín-Martín, Peter Stone, Ben Abbatematteo, Yoonchang Sung

Abstract

Robot planning in partially observable environments, where not all objects are known or visible, is a challenging problem, as it requires reasoning under uncertainty through partially observable Markov decision processes. During the execution of a computed plan, a robot may unexpectedly observe task-irrelevant objects, which are typically ignored by naive planners. In this work, we propose incorporating two types of common-sense knowledge: (1) certain objects are more likely to be found in specific locations; and (2) similar objects are likely to be co-located, while dissimilar objects are less likely to be found together. Manually engineering such knowledge is complex, so we explore leveraging the powerful common-sense reasoning capabilities of large language models (LLMs). Our planning and execution framework, CoCo-TAMP, introduces a hierarchical state estimation that uses LLM-guided information to shape the belief over task-relevant objects, enabling efficient solutions to long-horizon task and motion planning problems. In experiments, CoCo-TAMP achieves an average reduction of 62.7 in planning and execution time in simulation, and 72.6 in real-world demonstrations, compared to a baseline that does not incorporate either type of common-sense knowledge.

Metadata

arXiv ID: 2603.03704
Provider: ARXIV
Primary Category: cs.RO
Published: 2026-03-04
Fetched: 2026-03-05 06:06

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.03704v1</id>\n    <title>Large-Language-Model-Guided State Estimation for Partially Observable Task and Motion Planning</title>\n    <updated>2026-03-04T04:07:22Z</updated>\n    <link href='https://arxiv.org/abs/2603.03704v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.03704v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Robot planning in partially observable environments, where not all objects are known or visible, is a challenging problem, as it requires reasoning under uncertainty through partially observable Markov decision processes. During the execution of a computed plan, a robot may unexpectedly observe task-irrelevant objects, which are typically ignored by naive planners. In this work, we propose incorporating two types of common-sense knowledge: (1) certain objects are more likely to be found in specific locations; and (2) similar objects are likely to be co-located, while dissimilar objects are less likely to be found together. Manually engineering such knowledge is complex, so we explore leveraging the powerful common-sense reasoning capabilities of large language models (LLMs). Our planning and execution framework, CoCo-TAMP, introduces a hierarchical state estimation that uses LLM-guided information to shape the belief over task-relevant objects, enabling efficient solutions to long-horizon task and motion planning problems. In experiments, CoCo-TAMP achieves an average reduction of 62.7 in planning and execution time in simulation, and 72.6 in real-world demonstrations, compared to a baseline that does not incorporate either type of common-sense knowledge.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.RO'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-03-04T04:07:22Z</published>\n    <arxiv:primary_category term='cs.RO'/>\n    <author>\n      <name>Yoonwoo Kim</name>\n    </author>\n    <author>\n      <name>Raghav Arora</name>\n    </author>\n    <author>\n      <name>Roberto Martín-Martín</name>\n    </author>\n    <author>\n      <name>Peter Stone</name>\n    </author>\n    <author>\n      <name>Ben Abbatematteo</name>\n    </author>\n    <author>\n      <name>Yoonchang Sung</name>\n    </author>\n  </entry>"
}