Accurate Pose Estimation Using Contact Manifold Sampling for Safe Peg-in-Hole Insertion of Complex Geometries
🤖 AI Summary
For robotic peg-in-hole assembly of non-convex parts under micro-gaps (0.1–1.0 mm), inaccurate state estimation frequently causes jamming and excessive contact forces. To address this, we propose a contact-only SE(3) pose estimation framework operating online. Our method innovatively constructs a short-horizon, motion-driven online contact submanifold and achieves high-precision pose estimation via efficient matching against an offline precomputed contact manifold. It integrates active exploration, force–pose coupled modeling, and lightweight learning—training completed within 6 seconds. Evaluated on five industrial-grade complex geometries, the approach achieves a 96.7% insertion success rate, a sixfold improvement over the no-state-estimation baseline. It significantly reduces both mean insertion force and time, demonstrating the feasibility of safe, robust, and efficient micro-gap assembly.
Application Category
📝 Abstract
Robotic assembly of complex, non-convex geometries with tight clearances remains a challenging problem, demanding precise state estimation for successful insertion. In this work, we propose a novel framework that relies solely on contact states to estimate the full SE(3) pose of a peg relative to a hole. Our method constructs an online submanifold of contact states through primitive motions with just 6 seconds of online execution, subsequently mapping it to an offline contact manifold for precise pose estimation. We demonstrate that without such state estimation, robots risk jamming and excessive force application, potentially causing damage. We evaluate our approach on five industrially relevant, complex geometries with 0.1 to 1.0 mm clearances, achieving a 96.7% success rate - a 6x improvement over primitive-based insertion without state estimation. Additionally, we analyze insertion forces, and overall insertion times, showing our method significantly reduces the average wrench, enabling safer and more efficient assembly.
Problem
Research questions and friction points this paper is trying to address.
Estimating full SE(3) pose for tight-clearance peg-in-hole insertion
Preventing jamming and damage during robotic assembly
Improving success rate for complex non-convex geometries
Innovation
Methods, ideas, or system contributions that make the work stand out.
Uses contact states for SE(3) pose estimation
Constructs online submanifold via primitive motions
Maps to offline manifold for precise estimation
🔎 Similar Papers
No similar papers found.
