Resilient System Architecture for an Autonomous Robotic Patrol System

As research endeavors in autonomous robotics and multi-robot systems grow, more real-world applications of these technologies are identified. The coordination of multiple types of robots to achieve a common objective can increase the ability of the system. Defining resilience as the system's ability to absorb shocks, adapt, and restore following a disruption/disturbance, this study outlines a resilient system architecture for an autonomous robotic patrol system meant to utilize the various functions of current robotic platforms to act as surveillance and complete tasks in both military and industrial environments. Many existing multi-robot systems use a centralized control method which has a central compute system to coordinate the movement and actions of all robots within the system. Research and development of decentralized control methods exist where robots have a certain degree of decision-making ability. The proposed system architecture allows the robotic fleet to operate under decentralized control with decision-making abilities while also having the ability to transfer over to a centralized control method under certain conditions. The architecting of this type of hybrid system is significant as it allows the robots to perform autonomous patrolling and other tasks entirely based on sensing the environment around them while automatically transitioning into a centralized control for more involved and significant tasks.