Learning to Explore Indoor Environments using Autonomous Micro Aerial Vehicles

TL;DR

This paper introduces a novel deep learning and reinforcement learning framework enabling autonomous micro aerial vehicles to efficiently explore unknown indoor environments within strict size, weight, and power constraints, significantly improving exploration efficiency.

Contribution

The paper presents a new onboard exploration framework combining DL-based map prediction and DRL-based navigation, optimized for SWaP-constrained aerial robots, with demonstrated superior performance.

Findings

Achieves 50-60% higher exploration efficiency than existing methods.

Successfully runs onboard on SWaP-constrained hardware.

Outperforms state-of-the-art exploration algorithms in experiments.

Abstract

In this paper, we address the challenge of exploring unknown indoor aerial environments using autonomous aerial robots with Size Weight and Power (SWaP) constraints. The SWaP constraints induce limits on mission time requiring efficiency in exploration. We present a novel exploration framework that uses Deep Learning (DL) to predict the most likely indoor map given the previous observations, and Deep Reinforcement Learning (DRL) for exploration, designed to run on modern SWaP constraints neural processors. The DL-based map predictor provides a prediction of the occupancy of the unseen environment while the DRL-based planner determines the best navigation goals that can be safely reached to provide the most information. The two modules are tightly coupled and run onboard allowing the vehicle to safely map an unknown environment. Extensive experimental and simulation results show that our approach surpasses state-of-the-art methods by 50-60% in efficiency, which we measure by the fraction of the explored space as a function of the length of the trajectory traveled.