arXiv 2509.05397

RoboBallet: Planning for Multi-Robot Reaching with Graph Neural Networks and Reinforcement Learning

By Matthew Lai, Keegan Go, et al.

Published 2025-09-05

Mindmap

Browse the paper's core ideas, clusters, and relationships in a structured outline.

Modern robotic manufacturing requires collision-free coordination of multiple robots to complete numerous tasks in shared, obstacle-rich workspaces. Although individual tasks may be simple in isolation, automated joint task allocation, scheduling, and motion planning under spatio-temporal constraints remain computationally intractable for classical methods at real-world scales. Existing multi-arm systems deployed in…

View the original paper on arXiv