Introducing SB-CoRLA, a schema-based constructivist robot learning architecture

We introduce the SB-CoRLA architecture that we have developed by extending our previously developed centralized ASyMTRe architecture (CA) to enable constructivist learning for multi-robot team tasks. We believe that the schema-based approach used in ASyMTRe is a useful abstraction for enabling constructivist learning. The CA algorithm only finds complete solutions for the entire team and is not well-suited for identifying useful schema chunks that can be used to find future task solution. Thus, we explore an evolutionary learning (EL) technique for the offline learning of schema chunks. We compare the solutions discovered by the EL algorithm with those that are found using CA, as well as with a third algorithm that randomizes the CA algorithm, called RA. Four different applications in simulation are used to evaluate the techniques. Our results show that the EL approach finds solutions of comparable quality to the CA technique, while also providing the added benefit of learning highly fit schema chunks.