This chapter presents a newly developed mobile robot based multi-agent system with capabilities of robust motion control and intelligent collision avoidance. The system consists of three mobile robots. One main robot acts as a master and the other two act as slaves. The master intelligently takes decisions as to which action to perform to avoid obstacles and collisions. The master mobile robot has the capability to swerve around a static or moving object when necessary. All possible conditions have been coordinated in a fuzzy knowledge base which is used to make a decision on the required maneuver to avoid a collision with a slave robot that the mobile robot may encounter on its driving lane. The proposed research has been carried out to simulate a real car driving regime on roads where the driver may not react properly. The system is implemented on a robot experimental test bench and some experimental results are presented and discussed.