Moving Target Search Using Theory of Mind

In both the domain of virtual agents and robotics the problem of moving target search is highly relevant. Hereby, algorithms such as A*, IDA*, or more advanced improvements are used to enable an agent to find a moving target in a certain world. One of the elements not utilized in these algorithms is a dedicated model of the behavior of the moving target, which for instance humans would use to tackle such a problem. In this paper, an approach is introduced that enables the agent that searches the target to reason using a Theory of Mind model of the moving target. The model is used in a number of scenarios and compared approaches to solve the problem.