A MULTI-AGENT INTELLIGENT AUTONOMOUS SOFTWARE SYSTEM IN A DYNAMIC ADVERSARIAL ENVIRONMENT

Abstract

Automation of Multi-Agent systems is a relatively new field which deals with the development of Artificial Intelligence for robots to adapt in a rapidly changing environment accordingly. Automated systems are appreciated and preferred over manual systems due to many reasons such as labor cost, unpredictable human error percentage etc. In this project, we have worked on developing the software architecture of an intelligent autonomous team of robots. Robocup small size league is identified as a problem research domain to apply these concepts of multi-agent autonomous systems since it has been identified as a challenging research problem. The application of this project is to build intelligence for the robots on a server. The robots communicate with the server individually. The server contains the whole strategic intelligence of how to make a decision for the robots in a multi-agent and dynamic environment. This application would allow the robots to decide their best positions and tasks according to the current environment. It would also allow them to change their decisions rapidly where required and adapt to the changing environment quickly.