Recent Articles and Multimedia
Paper Title: Vision-Based Collision Avoidance for Multi-Agent Systems with Intermittent Measurements
Authors: Mia Scoblic, Camilla Tabasso, Venanzio Cichella, and Isaac Kaminer
This paper presents a novel decentralized vision-based collision avoidance algorithm for multiagent systems that operates without inter-vehicle communication and can handle intermittent sensor measurements. By re-planning to circular detour paths and maintaining a safe phase shift between agents, the method guarantees collision avoidance, with its effectiveness validated through simulations and real-world drone experiments.
READ MORE ABOUT THIS PAPER HERE
Four drones fly toward a central collision point. Upon detecting each other’s presence at the red dots, they identify a potential collision, trigger the algorithm, and execute a coordinated circular detour to maintain safe separation.