Ocean plastic pollution poses a severe threat to marine life and ecosystems, yet monitoring it remains challenging due to the labor-intensive, time-consuming nature and limited coverage of traditional surveys. To address this, We developed an automated model to detect ocean floating plastics across large areas by combining Sentinel-2 satellite imagery (free for public use) with machine learning model. The machine learning model was trained and tested on annotated plastic data from selected sites. Results showed that our satellite-based AI technique achieved an overall detection accuracy of 92%. This technique offers an effective tool for monitoring of ocean plastic pollution over large areas. 

Impact of our AI plastic detecting model

• Enhances efficient cleanups

• Informs better polices

• Drives innovative solutions to combat ocean pollution and support ocean conservation

• Helps protect marine biodiversity, sustain fisheries, and support coastal communities’ economies that rely on clean oceans (the project can save 825 million – 19.8 billion dollars over the world per year).