An On-Line POMDP Solver for Continuous Observation Spaces

TL;DR

This paper introduces LABECOP, an online POMDP solver for continuous observation spaces that combines Monte-Carlo-Tree-Search and particle filtering, avoiding discretization and observation limiting.

Contribution

The paper presents LABECOP, a novel online POMDP solver that handles continuous observations without discretization or observation limiting, improving planning efficiency and robustness.

Findings

LABECOP performs comparably or better than existing solvers.

It effectively handles continuous observation spaces.

Experiments demonstrate its efficiency across different problems.

Abstract

Planning under partial obervability is essential for autonomous robots. A principled way to address such planning problems is the Partially Observable Markov Decision Process (POMDP). Although solving POMDPs is computationally intractable, substantial advancements have been achieved in developing approximate POMDP solvers in the past two decades. However, computing robust solutions for problems with continuous observation spaces remains challenging. Most on-line solvers rely on discretising the observation space or artificially limiting the number of observations that are considered during planning to compute tractable policies. In this paper we propose a new on-line POMDP solver, called Lazy Belief Extraction for Continuous POMDPs (LABECOP), that combines methods from Monte-Carlo-Tree-Search and particle filtering to construct a policy reprentation which doesn't require discretised observation spaces and avoids limiting the number of observations considered during planning. Experiments on three different problems involving continuous observation spaces indicate that LABECOP performs similar or better than state-of-the-art POMDP solvers.