Design of a Novel Priority Mux Based Indoor Autonomous Surveillance Robot

Bank robberies have exacerbated with time. It is a daunting task to monitor the banks for human-based security, considering long hours. Autonomous surveillance robots, in such situations, are beneficial to patrol a particular environment by collecting data with an electronic payload. DRISHTI, the proposed robot model designed using Robot Operating System (ROS), includes facial recognition to raise an alert if any suspect is detected, and human tracking is implemented during night patrolling. An Intrusion Alarm Panel (IAP) is employed using external sensors to detect any unethical breaking of the vault in addition to the anti-theft feature. The robot surveillance is aided by autonomous patrolling using Cartographer-based Simultaneous Localization and Mapping (SLAM). The accuracy of the generated map of the prototype is found to be 95.31% with respect to the corresponding ground truth. The design of a novel priority mux is proposed to integrate the surveillance functionalities with the navigation node. The alert velocity command node is made to override the navigation velocity commands with a 10 times priority ratio using the novel priority mux with a latency of less than one millisecond.