Inspiration
We got the idea for solving the mine problem from a discussion we had with LMT.
What it does
Landmine detection is slow, dangerous, and expensive; large areas stay blocked for years. Minesweep ingests multi-sensor drone data and fuses it with AI to highlight likely hazardous zones on a map with clear confidence scores and an audit-ready evidence stack.
How we built it
We used google cloud platform and Ultralytics YOLOv11 models to train a fine tune a model that can work outside the RGB spectrum.
Challenges we ran into
At first we didn't have an idea or know what to do but by the end of the second day we had already developed most things necessary and caught up.
Accomplishments that we're proud of
Actually getting it done.
What we learned
How to find problems and solutions as well as how to develop the solution with a very small time window. (Less than others.)
What's next for Minesweep
Getting more business people on the team, continuing the AI development and mapping algorithm, testing it with drones and getting a full solution by 2027.
Log in or sign up for Devpost to join the conversation.