Applying AI to cooperating agricultural robots

Abstract

We consider the experimental development of collaborating robots able to work within fields of crops. Specifically we investigate how AI principles can be applied to this agricultural domain and how the user should be involved in such a system. To support collaboration between agents a representation of responsibilities and dependencies is necessary. This is done by introducing groups and roles, from MAS theory, that the agent must adhere to, formalised by the Agent-Group-Role (AGR) model. To enable a human user to influence the system we adopt the principles declared by the VOWELS paradigm. We then show that the AGR model and the VOWELS paradigm, enable us to solve two practical agricultural problems, and lastly we argue that the obtained results can be transferred to other domains, such as pervasive computing. © 2006 International Federation for Information Processing.