Challenge being tackled: Challenge 2 (Stereo event-camera on miniature drone), was focused on sensor integration, computer vision, design and development of miniaturised drones, advanced camera systems, and it was proposed by the University of Zurich.
Description of the solution: SPECTRUM introduces a compact aerial platform designed to autonomously navigate and inspect complex, low-visibility industrial environments, such as ship ballast tanks, using cutting-edge stereo event-based vision. At the core of the system is a stereo pair of Prophesee GenX320 event cameras and an NVIDIA Jetson Orin Nano onboard processor for low latency depth-processing. The mini drone features a lightweight but robust frame, high-RPM brushless motors for enhanced agility and manoeuvrability and open-source autopilot software for stability and reactivity. The system will run a custom vision stack, including spatiotemporal event filtering and noise suppression, stereo event correlation for disparity estimation, 3D edge-based depth reconstruction and SLAM.
This high-fidelity 3D pipeline offers excellent performance in visually degraded conditions, through the inherent resistance of event cameras to motion blur and variable lighting, maintaining perception in dust, glare or even total darkness. This leads to unprecedented agility and awareness during navigation in confined spaces and real-time obstacle recognition and avoidance. Onboard autonomy is provided by ROS2 handling sensor fusion, path planning and triggering failsafe behaviours, to enable safe and independent automated inspections.
About the provider: CogniSensus builds technologies that allow better sensing, data modelling, visualisation and/or actuation, supporting the operation, decision- making or training in complex environments.
Check their LinkedIn here.
The Team behind the scenes:
Georgios Tsimiklis – Project Overview
Vasileios Balntas – Project Lead
Fotios Konstantinidis – Smart Manufacturing leader, work on the drone and the integration
Valantis Tsiakos – Drone development and installation of the camera
Mary Papaefthymiou – Administrative and Quality
