Finite State Machine based Vehicle System for Autonomous Driving in Urban Environments

In autonomous driving systems, many approach have been developed based on high-precision sensors, high computing power, development of various networks in academia and industry. The key point of an autonomous car is that it needs to know what is happening in the driving situation and an action corresponding to this situation must be created and operated. In this paper, we propose the system structure and method of a autonomous driving car consisting of three main parts:perception, planning, control. Each main part is designed to recognize the real-time driving environment, make action plan based on finite state machine, and activate kinematic model based control. The proposed vehicle system copes with six major situations of urban environment driving, and we verified system through the simulation.