System and software architecture for autonomous surveillance robots in urban environments

In this paper, we propose a system of the security robot and its software architecture. The software architecture consists of two parts, which are an autonomous navigation part and a semantic perceptron part. Autonomous navigation software can drive a robot autonomously. An semantic perception software can perform the security missions. The security patrol scenario compromises three steps. In the first step, a robot collects information about the environments. Then the perceived structures are compared with previous information. If there is something to be changed or to be strange, a robot alarm to a supervisor. The proposed system and software architecture will make a robot to perform security missions.