A collision avoiding mechanism based on a theory of mind

Abstract

We develop a collision avoiding mechanism for a system of individual agents (pedestrians) that move in a crowd trying to reach their different goal points. The agents avoid collisions on the base of a model of the other agents’ behaviour, a “Theory of Mind”, which is realised at different levels through an iterative process (the first, or 0, level corresponds to ignore the other agents’ behaviour, level 1 to assume that the other agents will ignore each other, and so on). We show that our model reproduces some of the more simple organised behaviours of a system of pedestrians, and perform an evolutionary study of the parameters of the model (as the perception of the agent’s own size, the attraction to the goal, the radius and angle of view and the level of theory of mind)