This project seeks to innovate the University of Maryland community's public safety toolset by implementing an autonomous patrol system. This system would consist of multiple robots capable of patrolling established perimeters and taking escort requests to various destinations on campus. Programmed to analyze historic trends and current instances of crime in the community, the robots reach a level of autonomy that allows them to decide on the safest possible route on campus to escort users through, even accounting for real-time potential risks and subsequent rerouting. The goal of this autonomous system of robots is to promote peace of mind amongst the community and to add an extra level of versatile surveillance on campus.

The majority of my research contributions focused on the development of software simulations through Processing and Arduino; we wanted to use a real-life map overlay picture with visual markers to represent these patrol robots utilizing different pathing algorithms. By executing many simulation run-throughs, we could compile numerical data about the average time it would take for a patrol robot to safely find a student in need of guidance, as well as the safety "accuracy" of aiding that individual to their destination despite crime hot-spots generated live during the simulation. This project challenged me to code various graph traversal/pathing algorithms in an object-oriented manner; by doing that, it could be utilized as part of the behavior of patrol robot objects in our code simulations. I also gained experience with basic robotics circuitry and hardware as I learned how to program robotics commands in Arduino. Lastly, I learned how to employ teamwork and leadership skills within a research setting; in order to make progress on this topic, we needed to communicate various deadlines and deliverables with each other in an online setting due to COVID-19, which was a major challenge in itself.