For decades, scientists knew there was plastic in the ocean but had no reliable way to measure how much, where it was concentrated, or how it moved. Estimates ranged wildly, from 5 million to 150 million metric tons. The ocean is simply too vast to survey by boat.
Then came satellites. And then came the algorithms that could read them.
Seeing Plastic from Space
The Ocean Cleanup, started by Boyan Slat in the Netherlands, has grown into the world's largest ocean plastic removal project. But before you can clean up plastic, you need to find it. That turned out to be the harder problem.
Their team developed AI models that analyze satellite imagery and aerial photographs to detect plastic concentrations on the ocean surface. The challenge is that plastic in the ocean does not look like plastic on land. It is often partially submerged, covered in algae, broken into small pieces, and mixed with natural debris like seaweed and driftwood. A human looking at a satellite image cannot tell the difference. But a neural network trained on thousands of labeled images can.
The models achieve detection accuracy above 86 percent for concentrated plastic patches, which is high enough to guide cleanup operations. Before this technology existed, the cleanup vessels were essentially guessing where to deploy.
The Great Pacific Garbage Patch
The Ocean Cleanup's primary target is the Great Pacific Garbage Patch, a zone of concentrated plastic debris between Hawaii and California that covers an area roughly twice the size of Texas. The AI mapping revealed that the patch contains approximately 100 million kilograms of plastic, a number far higher than previous estimates based on boat surveys.
As of 2026, The Ocean Cleanup has removed over 15 million kilograms of plastic from oceans and rivers worldwide. Their latest system covers an area the size of a football field and operates continuously for weeks. AI guides every deployment, predicting where plastic will concentrate based on current patterns, wind data, and historical accumulation models.
Rivers: Stopping Plastic at the Source
Research shows that 80 percent of ocean plastic enters through rivers, with just 1,000 rivers responsible for roughly 80 percent of that flow. The Ocean Cleanup developed Interceptors, autonomous solar powered barges that sit in rivers and collect plastic before it reaches the sea.
AI plays a critical role here too. Machine learning models predict plastic flow based on rainfall, river levels, population density upstream, and waste management infrastructure. This helps the team decide where to place Interceptors for maximum impact. They currently operate in Indonesia, Malaysia, Vietnam, Jamaica, the Dominican Republic, and Guatemala.
A Companion Effort: Global Fishing Watch
While The Ocean Cleanup focuses on plastic, another organization is using similar AI techniques to protect marine life. Global Fishing Watch, a partnership between Google, Oceana, and SkyTruth, uses machine learning to monitor the movements of over 65,000 fishing vessels in near real time. The system detects suspicious behavior patterns that indicate illegal, unreported, and unregulated fishing.
Their AI has revealed that an estimated 20 percent of the world's catch comes from illegal fishing. By making vessel tracking data publicly available, they have given governments, journalists, and conservation groups the tools to hold the fishing industry accountable. Indonesia alone has used this data to crack down on hundreds of vessels fishing illegally in its waters.
The Bigger Picture
Neither satellite imagery nor machine learning will solve the plastic crisis on their own. Production of new plastic continues to grow at about 4 percent per year, and no amount of cleanup can keep pace with that unless production is reduced. The Ocean Cleanup acknowledges this openly. Their goal is not to replace policy change but to prove that cleanup is technically feasible while those policy changes work their way through governments.
What AI has done is make the invisible visible. We can now see exactly how much plastic is in the ocean, exactly where it concentrates, and exactly which rivers are carrying it there. That clarity is itself a form of action. Because once you can see a problem with that level of precision, it becomes much harder to ignore.
Sources: The Ocean Cleanup progress reports, Science Advances (2023), Global Fishing Watch transparency platform, Nature Sustainability (2024), MBARI research publications.
