Research Project Title:
Fast Multi-Vehicle Path-Planning with Decentralized Heuristics
abstract:NASA is interested in investigating the liquid water ocean hidden under the icy surface of Europa for conditions suitable for life using multiple underwater autonomous vehicles. However, concurrently planning paths for every vehicle using a single system is computationally difficult because the problem’s complexity increases exponentially with each additional vehicle. This project focuses on developing a distributed-path planning algorithm that plans each vehicle’s optimal physical path in a computationally reasonable manner while ensuring that vehicles will be safe and that maximum scientific knowledge is obtained during the mission. At a high level, this is accomplished by detecting when interactions between vehicles can be ignored, producing distinct problems to be individually solved.
“I am participating in the SuperUROP program to deepen my understanding of automated path-planning and activity-scheduling by taking on a multivehicle coordination project with direct application to a visionary under-ice exploration mission to Europa. I expect to gain intensive research experience and connect with world-class faculty, which are both critical to my goal of gaining admission to grad school and achieving a doctorate degree.”