· 4 min read
Plastic pollution is one of the most pressing environmental challenges of our time, with millions of tons entering the world’s oceans and rivers each year. The urgency to deal with this is ever growing, particularly in light of the increasing body of scientific evidence linking exposure to plastic and the particles into which it breaks down to a range of conditions impacting human health and biodiversity.
To address this urgent, global issue and achieve its mission of ridding the world’s oceans of plastic, The Ocean Cleanup, a nonprofit organisation based in The Netherlands, is leveraging cutting-edge artificial intelligence (AI) as a catalyst to help restore ocean health.
By strategically integrating AI, drones, machine learning, and remote sensing, The Ocean Cleanup is revolutionising the detection, tracking, and removal of plastic debris from not only the Great Pacific Garbage Patch (GPGP), a vast plastic soup floating between the west coast of the US and Hawaii that covers an area the size of Texas, but also from rivers where our strategic interceptor barriers stop the trash ‘tsunamis’ entering our seas in the first place.
Pioneering AI-driven solutions are helping transform its efforts, accelerating positive impact on marine ecosystems, coral reefs, mangroves, and impacted communities worldwide, and showcasing how new and innovative applications of science and technology can help clean our oceans.
For example, AI has enabled real-time monitoring, predictive modelling, and automated plastic detection, using advanced machine learning techniques to analyse vast amounts of environmental data, including automated waste classification and plastic and organic material differentiation to ensure that cleanup efforts are strategic and efficient.
Quantifying floating plastic debris at sea using AI
The Ocean Cleanup’s Automatic Debris Imaging System (ADIS), an AI-driven tool that analyses surface plastic debris in oceans leverages image recognition and deep learning algorithms, classifying different types of plastic waste and quantifying pollution density. This real-time data helps optimise cleanup operations by targeting areas with the highest plastic concentrations. Meanwhile, its River Monitoring System (RMS) employs cameras to capture images of plastic debris which are then used to train AI models to identify and categorise plastic pollution. Understanding how plastic emissions change over time, due to geographical and seasonal fluctuations, allows us to more efficiently and precisely identify debris accumulations so that we can better target ‘hotspots.’
Plastic forecasting
Beyond detecting existing pollution, AI will also play a crucial role in our ability to forecast plastic accumulation in the future. By integrating oceanic currents, wind patterns, and river outflow data into predictive models, it will be possible to anticipate where plastic waste is likely to accumulate, speeding up future operations.
Additionally, drones can provide a cost-effective and scalable solution for monitoring pollution over large and remote areas. Capturing high-res images with AI-enabled Unmanned Aerial Vehicles (UAVs) allows us to detect and differentiate plastic debris from natural elements, even in the most hard-to-reach riverine and oceanic locations. This drone-collected data is fed into machine learning algorithms, improving plastic recognition accuracy over time.
Remote sensing
Remote sensing technology, including satellite imaging and aerial surveys, provides us with a comprehensive view of plastic pollution on a global scale. Combining AI with satellite data can identify pollution hotspots, along rivers and coastal areas, and assess cleanup impact over time.
Satellite imaging helps detect and track macroplastic movements across oceans, but the quantification of individual objects, under 50 cm, remains challenging. To address this, we have trained an object-detection algorithm by selecting and labelling footage of floating plastic debris recorded offshore with vessel-fixed GPS-enabled action cameras. Results are consistent with macroplastic densities predicted by global plastic dispersal models, and reveal insights into how camera-recorded offshore macroplastic densities compare to concentrations collected through other methods.
Challenges and future outlook
Despite significant advancements, challenges remain in leveraging AI for ocean health restoration, not least in acquiring the funding to realise most ambitions. For example, plastic detection limitations (distinguishing bio-fouled or submerged plastics from natural materials) are an ongoing challenge, and AI models require extensive datasets to improve accuracy, necessitating continuous data collection and international collaboration. Looking ahead, The Ocean Cleanup plans to refine its AI models, integrate more sophisticated drone and satellite capabilities, and expand partnerships with technology companies and research institutions worldwide.
Conclusion
By harnessing the power of machine learning, drones, and remote sensing, we are revolutionising how to detect, track, and remove plastic waste from marine ecosystems. And, as AI technology continues to evolve, innovative approaches like that demonstrated by The Ocean Cleanup will continue to serve as a blueprint for the future of environmental conservation, demonstrating that science and technology can be amongst our most powerful allies in restoring the health of our planet’s oceans.
illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.
