Regrasp Maps for Sequential Manipulation Planning

2025-07-16

AI Summary

Manipulation in cluttered environments under complex constraints requires repeated regrasping from unknown object poses, posing significant challenges for robust task and motion planning (TAMP). Method: We propose the “Regrasp Graph”—a state-space abstraction that explicitly models feasible regrasp regions and their associated grasp combinations. This graph enables prior-guided mode-switching reasoning within TAMP and is dynamically updated via online failure feedback to enhance planning robustness. Integrated with regrasp-sequence inference and iterative planning repair, our approach accelerates search convergence. Results: Experiments on challenging multi-stage regrasp tasks demonstrate that our method achieves higher success rates and greater computational efficiency compared to baseline approaches. It provides a scalable, adaptive planning framework for dexterous manipulation in cluttered scenes, advancing the capability of autonomous robotic systems to handle uncertainty and partial observability during physical interaction.

Technology Category

• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Intelligent Robots: Motion and Path Planning
• Multiagent Systems: Multiagent Planning

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We consider manipulation problems in constrained and cluttered settings, which require several regrasps at unknown locations. We propose to inform an optimization-based task and motion planning (TAMP) solver with possible regrasp areas and grasp sequences to speed up the search. Our main idea is to use a state space abstraction, a regrasp map, capturing the combinations of available grasps in different parts of the configuration space, and allowing us to provide the solver with guesses for the mode switches and additional constraints for the object placements. By interleaving the creation of regrasp maps, their adaptation based on failed refinements, and solving TAMP (sub)problems, we are able to provide a robust search method for challenging regrasp manipulation problems.

Problem

• Solving manipulation in cluttered spaces with unknown regrasp locations
• Speeding up TAMP via regrasp maps and grasp sequence guidance
• Enhancing robustness in regrasp planning through iterative map adaptation

Innovation

• Optimization-based TAMP solver with regrasp areas
• State space abstraction using regrasp maps
• Interleaving map creation and TAMP problem solving