Situated robot design with prioritized constraints

The constraint-based agent (CBA) framework with prioritized constraints is a simple and effective methodology for designing and building situated robots. This methodology can be seen as a formal development of the subsumption approach. It prevents ad hoc layering of behaviors and supports modular development of the system. A situated robot called Ainia that repeatedly finds, tracks, chases, and kicks a ball is presented as an illustrative case study of this design methodology. Ainia is first modeled, simulated, and animated with the constraint nets in Java (CNJ) tool. Then, a prioritized constraint-based controller of the simulated Ainia is used to control the physical robot in the real world. The results show that the behaviors of the physical robot satisfy the requirements specification. Hence, this study provides evidence that the formal CBA framework with prioritized constraints is an effective approach for synthesizing situated robot controllers. In addition, it supports the claim that CNJ is an effective tool for designing and building situated robots operating in the real world.