Detecting Marine Plastic Pollution with Edge AI Technology

Around eight to ten million tonnes of plastic end up in the ocean each year. For context, the Great Pyramid of Giza, the largest Egyptian pyramid, weighs approximately six million tonnes.

A global effort is underway to clean-up plastic from the world’s oceans, but finding and tracking it to then remove it is complicated. 

The Edge SpAIce project is enhancing on-board satellite data processing by developing more accurate yet compact artificial intelligence models. 

When using satellites to monitor marine plastic pollution, an extraordinary amount of data is collected. This data then needs to be sent back to Earth to be processed and analysed. But transferring this quantity of data can be time consuming. That’s where the edge AI system comes in – to help intelligently sort the data into what is useful and what is not, and send only the necessary elements back. 

This technology could mean near real-time detection and monitoring of marine plastic litter, helping to find it and then extract it before another pyramid-sized pile builds up. 

Edge SpAIce is an EU-funded collaborative endeavour involving CERN’s Experimental Physics Department, Bulgarian aerospace manufacturer EnduroSat and the National Technical University of Athens. The project is coordinated by France-based edge-AI technology specialists AGENIUM Space.