Decentralized Real-Time Planning for Multi-UAV Cooperative Manipulation via Imitation Learning

TL;DR

This paper introduces a decentralized imitation learning approach for multi-UAV cooperative manipulation that operates effectively without inter-agent communication, achieving real-time, smooth trajectory planning under partial observability.

Contribution

It presents a novel decentralized kinodynamic planning method using imitation learning, trained efficiently on standard hardware, and validated in real-world multi-UAV manipulation tasks.

Findings

Achieves performance comparable to centralized methods

Trains policies in under two hours on a standard laptop

Operates effectively under partial observability without communication

Abstract

Existing approaches for transporting and manipulating cable-suspended loads using multiple UAVs along reference trajectories typically rely on either centralized control architectures or reliable inter-agent communication. In this work, we propose a novel machine learning based method for decentralized kinodynamic planning that operates effectively under partial observability and without inter-agent communication. Our method leverages imitation learning to train a decentralized student policy for each UAV by imitating a centralized kinodynamic motion planner with access to privileged global observations. The student policy generates smooth trajectories using physics-informed neural networks that respect the derivative relationships in motion. During training, the student policies utilize the full trajectory generated by the teacher policy, leading to improved sample efficiency. Moreover, each student policy can be trained in under two hours on a standard laptop. We validate our method in both simulation and real-world environments to follow an agile reference trajectory, demonstrating performance comparable to that of centralized approaches.