cuArms: GPU-Accelerated Conflict-Based Planning for Multi-Arm Manipulation

cuArms is a GPU-accelerated conflict-based planner designed for multi-arm manipulation tasks. It decomposes the search space using reusable multi-modal roadmaps for each robot and resolves conflicts and deadlocks via a lazy search, while a high-level scheduler proposes candidate skeletons for joint optimization of task assignment, schedules, motions, and manipulation skills. Experimental results demonstrate real-time initial solutions and anytime improvements in motion quality and scheduling, outperforming previous methods that required extensive domain-specific training.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

• robotics
• efficiency
• multimodal