EVPeriscope: Event-Camera Propeller Tracking Gives Air and Ground Robots Mutual Perception

· Editorial Team estimated
event-camera air-ground-teaming relative-localization heterogeneous-robot-teams uav occluded-navigation

EVPeriscope equips a ground robot with an upward-facing event camera that picks up the high-frequency visual signature of a quadrotor's propellers, enabling detection, localization, and closed-loop control of the aerial vehicle without fiducial markers and without the motion-blur and lighting sensitivity of frame-based cameras. The quadrotor can therefore act as an extended perception system for the ground robot when its own sensors degrade or are occluded, forming a marsupial air-ground team. The authors run field experiments in winds up to 15 mph, in daylight and at night, and demonstrate localization and closed-loop navigation through dense foliage where the ground robot's sensors are blocked; quadrotor control runs at 200 Hz entirely on onboard sensing and computation.

Paper · arXiv:2609.11920

Reliable relative localization is a precondition for tightly coordinated heterogeneous air-ground teams. Conventional solutions rely on frame-based cameras and fiducial markers, which are sensitive to motion blur and lighting changes and add payload and compute burden.

Core Innovation

The paper introduces EVPeriscope, an event-based perception system in which a ground robot carries an upward-facing event camera and exploits the high-frequency visual signature of a quadrotor’s propellers to detect, localize, and control the drone. Because event cameras record only brightness changes, they are naturally robust to motion blur and lighting variation and are well suited to capturing fast-spinning propellers. The quadrotor can then serve as extended perception for the ground robot when the ground robot’s own sensors degrade or are occluded — a marsupial ground-aerial system in which team coordination itself becomes a source of perception redundancy.

Results

  • Field experiments in winds up to roughly 15 mph, in both daylight and at night.
  • Localization and closed-loop navigation through dense foliage where the ground robot’s sensors are occluded.
  • Quadrotor control at 200 Hz, entirely with onboard sensing and computation rather than external positioning infrastructure.
  • Experiment videos are available on the project page: https://ongdexter.github.io/evperiscope

Limitations

The abstract is qualitative: it reports no localization accuracy, effective range, or number of trials, and offers no quantitative comparison against frame-based or fiducial-marker baselines, so the claimed advantage cannot be measured from the summary. Performance still depends on lighting and on the propellers remaining visible in the field of view; the demonstrations involve a single quadrotor and a moderate 15 mph wind, leaving more extreme conditions and multi-drone formations unexplored. The system relies on a custom event camera and onboard compute, and the abstract discloses neither code nor a hardware bill of materials.

Industry Implications

For inspection, search and rescue, and subterranean or under-canopy settings, the appeal is trading hardware for perception: ground platforms already need dependable air-ground relative localization, and event cameras are low-power and compact enough for robots already near their payload limits. The approach also gives the “drone as an external sensor” idea a concrete localization mechanism — when the host platform’s sensors are wet, occluded, or failed, the heterogeneous team’s coordination itself becomes the redundant source of state.