Research Analyzer
← Back ICRA 2026

Low-Latency Event-Based Velocimetry for Quadrotor Control in a Narrow Pipe

Leonard Bauersfeld, Davide Scaramuzza

PDF

AI summary

Key figure (auto-extracted from paper)
Real-time flow measurements via event cameras enable stable hovering and precise lateral control of a quadrotor in a narrow pipe by actively counteracting aerodynamic disturbances.
event-based vision quadrotor control flow velocimetry disturbance estimation reinforcement learning confined space flight

Problem

Autonomous quadrotor flight in confined spaces is hindered by unsteady, self-induced aerodynamic disturbances and poor state estimation, with prior solutions either requiring constant forward motion or suffering from unstable hovering.

Approach

The team developed a low-latency, event-based smoke velocimetry system to measure local airflow in real time, which feeds a neural disturbance estimator to inform a reinforcement learning-based controller.

Key results

  • Sub-millisecond latency event-based smoke velocimetry with 0.35 m/s mean error
  • First real-time aerodynamic disturbance estimator for quadrotors using live flow data
  • 29% reduction in hovering position deviation and 71% reduction in lateral maneuver overshoot
  • Monocular event-based motion-capture system achieving millimeter accuracy on embedded hardware

Why it matters

Enables safe, stable aerial robot operation in aerodynamically complex confined environments while advancing real-time fluid-robot interaction research.

Abstract

Autonomous quadrotor flight in confined spaces such as pipes and tunnels presents significant challenges due to unsteady, self-induced aerodynamic disturbances. Very recent advances have enabled flight in such conditions, but they either rely on constant motion through the pipe to mitigate airflow recirculation effects or suffer from limited stability during hov- ering. In this work, we present the first closed-loop control system for quadrotors for hovering in narrow pipes that leverages real- time flow field measurements. We develop a low-latency, event- based smoke velocimetry method that estimates local airflow at high temporal resolution. This flow information is used by a disturbance estimator based on a recurrent convolutional neural network, which infers force and torque disturbances in real time. The estimated disturbances are integrated into a learning-based controller trained via reinforcement learning. The flow-feedback control proves particularly effective during lateral translation maneuvers in the pipe cross-section. There, the real-time disturbance information enables the controller to effectively counteract transient aerodynamic effects, thereby preventing collisions with the pipe wall. To the best of our knowledge, this work represents the first demonstration of an aerial robot with closed-loop control informed by real-time flow field measurements. This opens new directions for research on flight in aerodynamically complex environments. In addition, our work also sheds light on the characteristic flow structures that emerge during flight in narrow, circular pipes, providing new insights at the intersection of robotics and fluid dynamics. MULTIMEDIA MATERIAL Video: https://youtu.be/ubHSlYZOeQQ

Index terms

Aerial Systems: Mechanics and Control Computer Vision for Other Robotic Applications Sensor-based Control Fluid Mechanics

Related papers