Robust anticipatory stigmergic collision avoidance in multi-agent systems
Friedrich Burkhard von der Osten, Michael Kirley, Tim Miller · Adaptive Agents and Multi-Agents Systems · 2014
Anticipatory Stigmergic Collision Avoidance (ASCA) is a reciprocal collision avoidance model in which agents interact by indicating intended future paths to other agents in their local environment. Other agents can then incorporate that information to guide their path planning decisions. Additionally, constraints such as restricted movements and noise are incorporated. Our evaluation shows that ASCA is consistently more robust than Reciprocal Velocity Obstacles in noisy environments in which ransmitted information can be lost or degraded.