A platform design for distributed computational intelligence

Abstract

To date, ordinary robotic systems have always operated individually, without any knowledge that other similar systems are present in their environment. However, recent developments in artificial intelligence and swarm robotics have shown that collaboration amongst robots can have many benefits and possible practical applications. This project investigates and addresses the hardware and software limitations of current robotic swarms. The designs, tests, and implementations of a robotic platform are also presented, along with demonstrations of the robotic platform operating in a small group of robots, computing simple behavioural algorithms. The project involves designing and building versatile robots with a small footprint. These mobile robots must be capable of sensing the environment composed of different obstacles, communicate wirelessly together, and be able to perform tasks as a team in an efficient and constructive manner. The design includes the mechanical and electronic aspects of these robots, which are presented as a final product. Following testing, calibration and identification of all sensors, two robots were tested in an obstacle detection, and path following application. The results obtained through local and global interaction of the robots verified the successful implementation of the platform in a small swarm. Moreover, the platform was fully characterized, and the objectives were successfully achieved, therefore, the robot can be used in a swarm to model distributed control systems and real-world scenarios.