Scale-Invariant Sampling in Multi-Arm Bandit Motion Planning for Object Extraction

2026-04-15

AI Summary

This work addresses the inefficiency of conventional sampling-based motion planners in narrow-space disassembly tasks—such as bolt or pin extraction—where high collision rates severely degrade performance. To overcome this limitation, the authors propose a scale-invariant sampling strategy that dynamically identifies high-entropy, effective sampling scales through an expand-and-contract search mechanism. This approach is integrated with principal component analysis (PCA) to determine favorable extraction directions and embedded within a multi-armed bandit RRT (MAB-RRT) framework to enable efficient motion planning. The proposed method introduces, for the first time, a scale-invariant sampling mechanism that substantially enhances planning efficiency in complex disassembly scenarios. Evaluated across eight 3D extraction tasks, it achieves up to an order-of-magnitude higher success rate than existing baselines in seven of the cases.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Search and Optimization: Sampling/Simulation-based Search
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Security and Privacy: Large-scale security measurements

Abstract

Object extraction tasks often occur in disassembly problems, where bolts, screws, or pins have to be removed from tight, narrow spaces. In such problems, the distance to the environment is often on the millimeter scale. Sampling-based planners can solve such problems and provide completeness guarantees. However, sampling becomes a bottleneck, since almost all motions will result in collisions with the environment. To overcome this problem, we propose a novel scale-invariant sampling strategy which explores the configuration space using a grow-shrink search to find useful, high-entropy sampling scales. Once a useful sampling scale has been found, our framework exploits this scale by using a principal components analysis (PCA) to find useful directions for object extraction. We embed this sampler into a multi-arm bandit rapidly-exploring random tree (MAB-RRT) planner and test it on eight challenging 3D object extraction scenarios, involving bolts, gears, rods, pins, and sockets. To evaluate our framework, we compare it with classical sampling strategies like uniform sampling, obstacle-based sampling, and narrow-passage sampling, and with modern strategies like mate vectors, physics-based planning, and disassembly breadth first search. Our experiments show that scale-invariant sampling improves success rate by one order of magnitude on 7 out of 8 scenarios. This demonstrates that scale-invariant sampling is an important concept for general purpose object extraction in disassembly tasks.

Problem

• object extraction
• narrow spaces
• sampling-based planning
• motion planning
• disassembly tasks

Innovation

• scale-invariant sampling
• multi-arm bandit
• PCA-based direction extraction
• narrow-passage planning
• object extraction