Search-Based Path Planning in Interactive Environments among Movable Obstacles

TL;DR

This work tackles Path planning Among Movable Obstacles (PAMO), aimed at finding a minimum-cost, collision-free path for a robot that may push movable objects in a known grid with static obstacles. It introduces PAMO* for bi-objective and resource-constrained formulations, guaranteeing completeness and optimality by employing a multi-objective A*-like search with non-dominated labels, implicit state generation, and global/local pruning. To bridge to continuous spaces and higher fidelity interactions, it extends to Hybrid-state PAMO* (H-PAMO*) by discretizing space, using motion primitives, and simulating robot–object interactions via a forward model and Box2D, albeit at the expense of optimality guarantees. Empirical results demonstrate rapid discovery of Pareto-optimal paths or optimal RC solutions on grids with up to 400 objects and show practical feasibility of H-PAMO* in realistic office/warehouse environments and Unity3D simulations, highlighting the method’s scalability and applicability to interactive environments.

Abstract

This paper investigates Path planning Among Movable Obstacles (PAMO), which seeks a minimum cost collision-free path among static obstacles from start to goal while allowing the robot to push away movable obstacles (i.e., objects) along its path when needed. To develop planners that are complete and optimal for PAMO, the planner has to search a giant state space involving both the location of the robot as well as the locations of the objects, which grows exponentially with respect to the number of objects. This paper leverages a simple yet under-explored idea that, only a small fraction of this giant state space needs to be searched during planning as guided by a heuristic, and most of the objects far away from the robot are intact, which thus leads to runtime efficient algorithms. Based on this idea, this paper introduces two PAMO formulations, i.e., bi-objective and resource constrained problems in an occupancy grid, and develops PAMO*, a planning method with completeness and solution optimality guarantees, to solve the two problems. We then further extend PAMO* to hybrid-state PAMO* to plan in continuous spaces with high-fidelity interaction between the robot and the objects. Our results show that, PAMO* can often find optimal solutions within a second in cluttered maps with up to 400 objects.

Figures (8)

• Figure 1: PAMO example. (d) and (e) show two alternative solutions, trading off arrival time for number of push.
• Figure 2: Runtime of MO-PAMO* and RC-PAMO* in Empty 8x8 grid map with varying percentage of objects.
• Figure 3: Numbers of expansion of MO-PAMO* and RC-PAMO* in the Empty map with varying object percentage.
• Figure 4: Runtime of MO-PAMO* and RC-PAMO* in three grid maps Random 32x32, Room 32x32 and Random 64x64, with 10% objects.
• Figure 5: Numbers of expansion of MO-PAMO* and RC-PAMO* in three grid maps Random 32x32, Room 32x32 and Random 64x64, with 10% objects.

Theorems & Definitions (2)

• Theorem 1
• Theorem 2