Satellite imagery is commonly used by the ICRC to monitor the devastating impacts of armed conflicts on populations. However, the prevailing approach relies on manual, on-demand assessments of very high-resolution (VHR) commercial satellite images, which are both time-consuming and costly.

The Remote Monitoring of Armed Conflicts (RMAC) project aims at leveraging machine learning and open source satellite imagery to help the ICRC scale up their monitoring effort and making it more cost-effective. The ultimate objective is to design a screening algorithm capable of automatically detecting changes indicative of armed conflicts, such as damaged buildings. This is envisioned to serve as an early-warning system, setting the stage for more in-depth manual investigations when necessary.

To date, the team has developed a tool that maps building destruction over time using pixel-wise time series of Sentinel-1 SAR data. This fully open-source tool operates on the cloud-based Google Earth Engine platform. A manuscript describing the methodology and its application to the Ukraine conflict is currently under review.