Quadrotor navigation adapts to complex wind and obstacle challenges
WAND: Learning Robust Navigation under Complex Wind Disturbances and Dense Obstacles for Quadrotors
Robotics
Summary
Flying small drones like quadrotors can be tricky when strong, changing winds and lots of obstacles are around, as these winds can push the drone off course and cause crashes. The authors created a new system called WAND that helps drones understand and react to wind disturbances by learning from past flight data. Their method uses a special neural network to estimate wind effects and adjusts the drone's controls both to avoid obstacles and to fight the wind. Tests in simulations and with real indoor fans showed WAND helps drones fly more reliably and safely in windy, cluttered spaces.
What this means in practice
- •For drone flight engineers: Improve drone flight controllers to better handle unpredictable wind effects in cluttered spaces using learned disturbance estimation integrated with navigation policies.
- •For warehouse automation teams: Enable drones to reliably navigate indoors near fans or ventilation systems by estimating wind disturbances and adjusting flight paths in real time.
Authors
Zhonghan Tang, Chenhui Li, Shuai Liang, Zhongrui You, Jianan Li, Bin Zhao, Zhigang Wang, Xuelong Li
Abstract
Robust navigation in cluttered environments remains a fundamental challenge for quadrotors, particularly when strong wind disturbances arise, which perturb vehicle dynamics, limit control authority, and substantially increase collision risk. Existing learning-based navigation policies typically rely on obstacle perception and proprioceptive observations, requiring the policy to infer time-varying disturbance effects implicitly and thereby limiting robustness under partial observability. This paper proposes WAND (Wind-Aware Navigation with Disturbance Estimation), a reinforcement learning framework for navigation under time-varying wind disturbances in dense obstacle fields. Specifically, WAND estimates wind-induced disturbance acceleration from historical proprioceptive states using a Temporal Convolutional Network (TCN). This estimation is integrated into the policy via a zero-initialized residual module, \emph{WindAdapter}, while simultaneously providing feedforward compensation for low-level control. The dual use of the estimate couples disturbance-conditioned navigation with feedforward disturbance rejection. Across 12 wind-disturbed simulation settings, WAND improved the observed success rate by 8.3 percentage points on average relative to feedforward compensation alone. Controlled opposite-crosswind experiments further showed wind-direction-dependent trajectory adaptation. In indoor fan-induced flight tests, WAND succeeded in 18 of 20 trials, demonstrating the feasibility of real-time onboard navigation.