RESTORE AI

Pattern-based Risk Estimation of Remaining Explosives

RESTORE AI is a machine learning solution that supports the survey and clearance of landmines. The solution predicts the risk of finding landmines at a given location by learning from topology data and landmine patterns. These patterns can be linear, curved, or more complex. Ultimately, the solution can help to improve the efficiency and effectiveness of landmine clearance.

Demo

Try RESTORE AI in the live demo below by clicking on the map to add locations of found landmines. The app uses RESTORE AI's machine learning algorithm to dynamically update the landmine patterns and calculate landmine risk. You can toggle between a linear pattern and a curved pattern version of the solution.

Curved patterns Linear patterns

Click on the map to place landmines

Place landmines close to each other to create patterns

Place landmines closer to each other to create a pattern!

Landmine probability:

low

high

Restart demo

Team

RESTORE AI is being developed by a team from the UCL Centre for Artificial Intelligence.

Björn Kischelewski

Dr. Benjamin Guedj

Publications

RESTORE AI is embedded in an academic project based at University College London. Several publications lay the foundation of the solution and explain its approach.

AI for EO Detection in Clearance Operations: The State of Research

coming soon

Contact

Contact us to get access to RESTORE AI, support the development, or ask a question.