Macroscopic quantitative observation of multi-robot behavior

Abstract

It is very difficult to estimate behaviors of multiple autonomous robots or mutual interactions of them in real time. Therefore, we propose a quantitative observation approach of multiple robots behaviors. This approach introduces thermodynamic macroscopic state values to the multi-robot systems. The advantage of this approach is that it enables to observe the behaviors of autonomous robots in real world and can be mapped to characteristic values in another conceptual state space. Thermodynamic macroscopic state values, such as temperature, pressure and entropy, are defined in mobile robots systems. In our definition, each mobile robot is supposed to have a particle in thermodynamic systems. The experiment shows that the states of robots system can be classified by thermodynamic macroscopic state value. This verifies that the macroscopic quantitative observation is efficient and applicable to control multi-robot systems.