Research
Coverage Path Planning Optimization
A NASA and Mosaic ATM collaboration — a coverage path planning optimization model built in Python with Gurobi, producing feasible plans across multi-facility scenarios.
Coverage path planning asks a specific question: how should an aircraft’s route be laid out so that it fully covers a given area?
I was a core developer on this collaboration between NASA and Mosaic ATM. I built the coverage path planning optimization model in Python with the Gurobi solver, able to generate feasible plans for scenarios involving multiple facilities.
What I built
Urban scan path generation. For a small vertical takeoff and landing fixed-wing aircraft, I modelled scanning missions in an urban environment and automatically generated scan paths that cover the target area.
Fastest route through multiple waypoints. A city mission often has several waypoints the aircraft must pass through; I solved for the fastest flight path through them.
Validation and integration
To find out whether the model held up at realistic scale, I also built an automated test framework and ran more than 20 large-scale scenarios, tracking feasibility rate, objective value, and run time for each. More than 80% of the approach was adopted in technical review.
The whole project was managed in Bitbucket, a code hosting platform similar to GitHub. I led version control for the code, and it was through Bitbucket that the modules I developed were integrated into the overall system.