DRAL: Deep Reinforcement Adaptive Learning for Multi-UAVs Navigation in Unknown Indoor Environment

TL;DR

This paper presents DRAL, a deep reinforcement learning system enabling autonomous indoor navigation of single and multiple UAVs using only a camera, with adaptive control for complex multi-drone tasks.

Contribution

It introduces a novel deep reinforcement adaptive learning algorithm for UAV navigation and multi-drone coordination in unknown indoor environments.

Findings

Effective real-time navigation in various indoor settings

Successful multi-drone object lifting coordination

Enhanced robustness and adaptability in dynamic conditions

Abstract

Autonomous indoor navigation of UAVs presents numerous challenges, primarily due to the limited precision of GPS in enclosed environments. Additionally, UAVs' limited capacity to carry heavy or power-intensive sensors, such as overheight packages, exacerbates the difficulty of achieving autonomous navigation indoors. This paper introduces an advanced system in which a drone autonomously navigates indoor spaces to locate a specific target, such as an unknown Amazon package, using only a single camera. Employing a deep learning approach, a deep reinforcement adaptive learning algorithm is trained to develop a control strategy that emulates the decision-making process of an expert pilot. We demonstrate the efficacy of our system through real-time simulations conducted in various indoor settings. We apply multiple visualization techniques to gain deeper insights into our trained network. Furthermore, we extend our approach to include an adaptive control algorithm for coordinating multiple drones to lift an object in an indoor environment collaboratively. Integrating our DRAL algorithm enables multiple UAVs to learn optimal control strategies that adapt to dynamic conditions and uncertainties. This innovation enhances the robustness and flexibility of indoor navigation and opens new possibilities for complex multi-drone operations in confined spaces. The proposed framework highlights significant advancements in adaptive control and deep reinforcement learning, offering robust solutions for complex multi-agent systems in real-world applications.