A Decentralized Multi-Agent Unmanned Aerial System to Search, Pick Up, and Relocate Objects

TL;DR

This paper introduces a fully integrated decentralized multi-agent aerial system capable of autonomously searching, detecting, tracking, and relocating objects in complex environments, demonstrating high reliability in real-world tests and competitions.

Contribution

It presents a novel integrated multi-robot aerial system combining exploration, detection, tracking, and object manipulation using simple components and decentralized coordination.

Findings

Successful deployment in real-world field tests

Achieved second place in MBZIRC 2017 challenge

Demonstrated reliable multi-object search and pickup capabilities

Abstract

We present a fully integrated autonomous multi- robot aerial system for finding and collecting moving and static objects with unknown locations. This task addresses multiple relevant problems in search and rescue (SAR) robotics such as multi-agent aerial exploration, object detection and tracking, and aerial gripping. Usually, the community tackles these problems individually but the integration into a working system generates extra complexity which is rarely addressed. We show that this task can be solved reliably using only simple components. Our decentralized system uses accurate global state estimation, reactive collision avoidance, and sweep planning for multi-agent exploration. Objects are detected, tracked, and picked up using blob detection, inverse 3D-projection, Kalman filtering, visual-servoing, and a magnetic gripper. We evaluate the individual components of our system on the real platform. The full system has been deployed successfully in various public demonstrations, field tests, and the Mohamed Bin Zayed International Robotics Challenge 2017 (MBZIRC). Among the contestants we showed reliable performances and reached second place out of 17 in the individual challenge.