aiqbench analysis
This paper establishes Embodied Scene Rearrangement Planning (ESRP) as a rigorous benchmark for embodied AI, specifically testing agents' ability to handle long-horizon tasks with partial observations and occlusions. By introducing ESRP-Bench with over 5,400 scene pairs, the work highlights current limitations in aligning egocentric vision with global layouts, setting a new standard for evaluating real-world robotic deployment.
Embodied Scene Rearrangement Planning
arXiv:2608.27371v1 Announce Type: new Abstract: This paper introduces Embodied Scene Rearrangement Planning (ESRP), a novel task requiring embodied agents to rearrange furniture in 3D scenes to match a target configuration using only egocentric observations and a top-down target layout. Unlike prior rearrangement tasks, ESRP precludes global state access and introduces mutual object occlusions, reflecting the practical constraints of real-world robotic deployment. These factors make aligning partial egocentric observations with the global target layout particularly challenging for long-horizon planning. To facilitate research, we present ESRP-Bench, a comprehensive benchmark built on OmniGibson featuring over 5,400 scene pairs and 8,200 objects. We define three multi-level metrics to evaluate rearrangement quality and provide four baselines: a hierarchical task-and-motion planning method, a vision-language-model-based method, and two learning-based approaches (IL and RL). Experimental results demonstrate that current methods struggle to complete the task efficiently, highlighting ESRP as a challenging frontier for embodied agents in scene understanding and long-horizon task planning. This work serves as a stepping stone toward deploying intelligent agents in real-world scenarios. Project page: https://pie-lab.cn/ESRP/.
Abstract:This paper introduces Embodied Scene Rearrangement Planning (ESRP), a novel task requiring embodied agents to rearrange furniture in 3D scenes to match a target configuration using only egocentric observations and a top-down target layout. Unlike prior rearrangement tasks, ESRP precludes global state access and introduces mutual object occlusions, reflecting the practical constraints of real-world robotic deployment. These factors make aligning partial egocentric observations with the global target layout particularly challenging for long-horizon planning. To facilitate research, we present ESRP-Bench, a comprehensive benchmark built on OmniGibson featuring over 5,400 scene pairs and 8,200 objects. We define three multi-level metrics to evaluate rearrangement quality and provide four baselines: a hierarchical task-and-motion planning method, a vision-language-model-based method, and two learning-based approaches (IL and RL). Experimental results demonstrate that current methods struggle to complete the task efficiently, highlighting ESRP as a challenging frontier for embodied agents in scene understanding and long-horizon task planning. This work serves as a stepping stone toward deploying intelligent agents in real-world scenarios. Project page: this https URL.
Accepted for publication in IEEE Robotics and Automation Letters (RA-L), 2026. Project page: this https URL Code: this https URL Dataset: this https URL
Robotics (cs.RO)
arXiv:2608.27371 [cs.RO]
(or arXiv:2608.27371v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2608.27371 arXiv-issued DOI via DataCite (pending registration)
https://doi.org/10.1109/LRA.2026.3728329 DOI(s) linking to related resources
Full text under the publisher's open licence · 215 words
Related coverage
ReactHuman: A Physics-Grounded Benchmark for Human-Like Reactive Decision-Making in Embodied Multimodal LLMs
arXiv · Sep 12, 2026
Worlds in One Demo: A Synthetic Data Engine for Learning Open-World Mobile Manipulation
arXiv · Jul 20, 2026
Gemini 4 Argon for Robotics
Google DeepMind · Oct 13, 2026
Running MolmoAct2 Robot Policy on an AMD Ryzen AI Max+ 395 Mini PC
AMD · Sep 30, 2026