Deep Learning-based Robust Autonomous Navigation of Aerial Robots in Dense Forests

TL;DR

This paper introduces an advanced deep learning framework for autonomous aerial navigation in dense forests, combining semantic depth encoding, motion primitives, and real-time safety measures to enhance robustness and success rates.

Contribution

It presents novel modules and improvements over existing algorithms, including lateral control, temporal consistency, stereo odometry, and safety filtering, tailored for real-world dense forest navigation.

Findings

Higher success rates in dense forest environments

More stable and collision-free trajectories

Successful autonomous flights in complex natural settings

Abstract

Autonomous aerial navigation in dense natural environments remains challenging due to limited visibility, thin and irregular obstacles, GNSS-denied operation, and frequent perceptual degradation. This work presents an improved deep learning-based navigation framework that integrates semantically enhanced depth encoding with neural motion-primitive evaluation for robust flight in cluttered forests. Several modules are incorporated on top of the original sevae-ORACLE algorithm to address limitations observed during real-world deployment, including lateral control for sharper maneuvering, a temporal consistency mechanism to suppress oscillatory planning decisions, a stereo-based visual-inertial odometry solution for drift-resilient state estimation, and a supervisory safety layer that filters unsafe actions in real time. A depth refinement stage is included to improve the representation of thin branches and reduce stereo noise, while GPU optimization increases onboard inference throughput from 4 Hz to 10 Hz. The proposed approach is evaluated against several existing learning-based navigation methods under identical environmental conditions and hardware constraints. It demonstrates higher success rates, more stable trajectories, and improved collision avoidance, particularly in highly cluttered forest settings. The system is deployed on a custom quadrotor in three boreal forest environments, achieving fully autonomous completion in all flights in moderate and dense clutter, and 12 out of 15 flights in highly dense underbrush. These results demonstrate improved reliability and safety over existing navigation methods in complex natural environments.