Design and Implementation of an Intelligent Autonomous Surveillance System for Indoor Environments

This paper aims to develop an intelligent autonomous surveillance system for the safety surveillance of indoor environments with the integration of techniques such as sensory information fusion, internet of things (IoT), and artificial intelligence (AI). We have developed the following four subsystems to implement the intelligent autonomous surveillance system: (1) an autonomous surveillance vehicle (ASV) to be placed on the track to detect environmental sensory information in indoor environments (such as environmental temperature, flame, CO concentration, and liquefied petroleum gas (LPG) concentration) and the movement information of the ASV (such as position coordinates and movement speed); (2) an intelligent fire detection algorithm to be used for identification of disaster statuses in indoor environments; (3) a visible human-machine interface to instantaneously display the environmental sensory information and ASV movement information, and show the environmental disaster information (such as disaster statuses, disaster images, image capture times, and disaster positions); (4) a remote server composed of a database server, file servers, a web server, and responsive web pages to store and display the recorded environmental sensory information, ASV movement information, and environmental disaster information. Furthermore, a 3D experimental track testbed was constructed to verify the effectiveness and feasibility of the intelligent autonomous surveillance system. Our experimental results have successfully validated that the intelligent autonomous surveillance system can detect disaster events in a variety of conditions.