The U.S. federal government considers the timely detection and repair of public infrastructure a top priority for both national and financial security. Because of their ease of maneuverability and low personnel risk, small Unmanned Aerial Systems (sUASs) are an increasingly popular choice for assessing infrastructure. While there is plenty of work on efficient routing of vehicles through known networks, comparatively little work has been done in routing of vehicles where physical constraints, such as GPS or communications coverage, are dynamically incorporated. Our proposed solution combines previous work in Vehicle Routing Problem (VRP) and Orienteering Problem (OP) formulations to address the problem of near-optimal routing of sUASs along nodes that need to be inspected in communication constrained environments. This solution provides a novel model where the system is learning the real connectivity of the network as it is exploring it, allowing for the possibility of dynamic re-planning if needed.

I am participating in SuperUROP to obtain a more focused experience in providing clear deliverables as a result of research. I hope to apply my knowledge of systems engineering to a problem that I find very fascinating, which is the detection of faults in pipe networks. I also want to learn about the process of writing and publishing a high quality paper.