Embodying Multi-Hand Manipulation Policies by Searching the Assignment and Null Spaces

2026-07-24

AI Summary

This work addresses the challenges of deploying learned multi-hand manipulation policies on multi-arm robotic systems, which involve trajectory assignment, kinematic constraints, and collision avoidance. The authors propose a unified conflict-based search framework that jointly models the discrete assignment of trajectories to arms and continuous motion planning within the nullspace of redundant-arm Jacobians—an integration not previously achieved. By doing so, the method ensures precise end-effector tracking of policy-specified trajectories while respecting kinematic limits and avoiding collisions, thereby overcoming the safety and feasibility limitations inherent in conventional single-arm inverse kinematics extensions. Experimental results demonstrate that the approach enables efficient and reliable embodiment of multi-hand policies on real multi-arm platforms, exhibiting both theoretical soundness and practical effectiveness.

Abstract

Learned manipulation policies increasingly predict motions for abstract "hands" and are attractive in practice because they rely on easily collected demonstrations and transfer across robot platforms. Executing these trajectories on multi-arm robots, however, is not trivial. Multi-hand policy outputs must be assigned to physical arms, each arm must realize a configuration-space motion that tracks its prescribed end-effector trajectory, and all arms must respect kinematic limits and avoid collisions. In the absence of algorithms that directly address this problem, practitioners typically extend single-arm inverse-kinematics (IK) pipelines in an ad hoc way, with no guarantees of feasibility or safety. In this work, we close this execution gap with a search-based framework that is theoretically complete for grounding policy-generated multi-hand trajectories onto physical multi-arm systems. Building on Conflict-Based Search, our method explicitly searches over both the discrete assignment of trajectories to arms and the continuous Jacobian null spaces of redundant manipulators, using redundancy to avoid inter-arm collisions while tracking the prescribed motions. This unified treatment of assignment and null-space motion yields a practically efficient planner that safely realizes coordinated manipulation-policy outputs on multi-arm robots. See omcbsa.github.io for more.

Problem

• multi-arm manipulation
• trajectory assignment
• inverse kinematics
• collision avoidance
• redundant manipulators

Innovation

• multi-arm manipulation
• trajectory assignment
• null-space optimization
• Conflict-Based Search
• redundant manipulators