{"slug":"sartech","name":"SARTech","hackathon":"pennapps-xxvi","tagline":"Predicts where a lost person will move, using terrain and behaviour models, to guide search-and-rescue drones.","placement":{"status":"track-winner","awards":["Most Technically Complex Hack"],"track":null},"team":{"size":3,"members":[]},"problem":"Drone searches usually follow a lawnmower pattern that assumes the missing person stays still.","solution":"A Monte Carlo model that incorporates how lost people behave (for example, following a river) over real geographic data to produce a probability map that steers the drone.","demo_moment":"Team-described: run the simulation over a map and see the drone's search area shift.","build_style":"hardware-software","tech":{"stack":["python","fastapi","react","google-earth","yolo","gemini"],"ai_patterns":["ml-model","computer-vision"]},"domains":["emergencies-disaster","robotics-hardware"],"analysis":{"why_it_stood_out":[{"claim":"Won “Most Technically Complex Hack” among 85 submissions.","basis":"organizer-stated","source":1},{"claim":"Models human psychology instead of assuming a stationary target.","basis":"team-stated","source":0},{"claim":"Drone hobbyists applied their own hardware experience.","basis":"team-stated","source":0}],"mechanisms":["inverted-approach","visible-engineering-depth","human-stakes"],"judging_lens":["technical-depth"],"lessons":["Challenge the default assumption of the existing method and show the difference."],"confidence":"low"},"links":{"devpost":"https://devpost.com/software/sartech","repo":null,"demo":null,"video":"https://www.youtube.com/watch?v=6qLmTkxwUps"},"sources":[{"url":"https://devpost.com/software/sartech","kind":"devpost-project","retrieved_at":"2026-10-13"},{"url":"https://pennapps-xxvi.devpost.com/project-gallery","kind":"devpost-gallery","retrieved_at":"2026-10-13"}]}
