aiqbench analysis
This research addresses the critical challenge of online planning in high-dimensional partially observable environments by reducing Monte Carlo variance through hybrid belief representations. By analytically propagating uncertainty within the Rao-Blackwellized framework, the proposed method enables more efficient decision-making for complex robotic tasks like search-and-rescue without sacrificing computational feasibility. The work represents a significant advancement in making rigorous probabilistic planning tractable for real-world AI agents operating under uncertainty.
Scalable Rao-Blackwellized Online Planning for High-Dimensional POMDPs
arXiv:2609.01351v1 Announce Type: cross Abstract: Online planning under uncertainty remains a fundamental challenge for robotic systems operating in partially observable environments with high-dimensional state spaces. While sampling-based POMDP solvers enable approximate decision-making in large or continuous domains, their performance degrades as belief dimensionality increases due to the high variance inherent in Monte Carlo-based estimation. In this work, we extend the Rao-Blackwellized online POMDP (RB-POMDP) framework to improve its generalizability in high-dimensional settings through hybrid continuous-discrete belief representations. By analytically propagating uncertainty associated with marginalized state components during tree-based planning, the proposed approach reduces sampling-induced variance in value estimation. We demonstrate the effectiveness of this framework in a robotic search-and-rescue task by integrating it with FastSLAM 2.0. Experimental results show that the proposed planner achieves higher cumulative rewards using significantly fewer particles and planning simulations than purely sampling-based methods under equivalent computational budgets. These results suggest that structured high-dimensional robotic problems admitting tractable sufficient statistics can be effectively leveraged within the RB-POMDP framework for computationally feasible online decision-making.
Abstract:Online planning under uncertainty remains a fundamental challenge for robotic systems operating in partially observable environments with high-dimensional state spaces. While sampling-based POMDP solvers enable approximate decision-making in large or continuous domains, their performance degrades as belief dimensionality increases due to the high variance inherent in Monte Carlo-based estimation. In this work, we extend the Rao-Blackwellized online POMDP (RB-POMDP) framework to improve its generalizability in high-dimensional settings through hybrid continuous-discrete belief representations. By analytically propagating uncertainty associated with marginalized state components during tree-based planning, the proposed approach reduces sampling-induced variance in value estimation. We demonstrate the effectiveness of this framework in a robotic search-and-rescue task by integrating it with FastSLAM 2.0. Experimental results show that the proposed planner achieves higher cumulative rewards using significantly fewer particles and planning simulations than purely sampling-based methods under equivalent computational budgets. These results suggest that structured high-dimensional robotic problems admitting tractable sufficient statistics can be effectively leveraged within the RB-POMDP framework for computationally feasible online decision-making.
Robotics (cs.RO); Artificial Intelligence (cs.AI)
arXiv:2609.01351 [cs.RO]
(or arXiv:2609.01351v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.01351 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 258 words
Related coverage
Think Fast and Far: Long-Horizon Online POMDP Planning via Rapid State Sampling
arXiv · Jun 4, 2026
VIP: Variation-based Iterative-learning Planning for Robotic Navigation
arXiv · Aug 26, 2026
Denoising Multi-Robot Trajectories
arXiv · Sep 29, 2026
Robust Belief-State Policy Learning for Quantum Network Routing Under Decoherence and Time-Varying Conditions
arXiv · Jul 23, 2026