GCSP research stipend | Spring 2026

Neuromorphic Vision for Reliable Robot Navigation in Challenging Conditions

Data icon, disabled. Four grey bars arranged like a vertical bar chart.

This study investigates whether neuromorphic, or event-based, vision can improve the reliability and performance of mobile robot navigation in visually challenging environments such as underground mines and tunnels. Using the Gazebo simulator and Robot Operating System 2 (ROS2), a mobile robot is tested under varied lighting, speed, and visibility conditions using three perception modes: Red-Green-Blue (RGB)-only, event-only, and RGB-event fusion. It is expected that the fusion approach will demonstrate faster reaction times, higher navigation success rates, and greater robustness compared to RGB-only perception. Future work will involve validating these simulation results on physical robotic platforms in real-world environments.