AI route-planning supercharges ocean cleanup, boosting plastic collection by 60 percent

The world's oceans hold millions of tonnes of plastic, and the organization The Ocean Cleanup has spent years dragging it out with large U-shaped barriers towed between ships. The hard part is knowing where to go: plastic drifts on currents and wind, scattering and concentrating in ways that are difficult to predict. A new AI route-planning tool tackles that question directly, and the gains are substantial.

Researchers from the University of Amsterdam, led by Dick den Hertog, and the London Business School, with Jean Pauphilet, working in collaboration with The Ocean Cleanup, developed a nonlinear path-optimization algorithm. Published in the journal Operations Research under the title "Optimizing the Path Towards Plastic-Free Oceans," the system computes the best collection routes for cleanup ships in real time, adapting to the dynamic, shifting marine environment. The result is a roughly 60 percent increase in plastic collection efficiency, achieved without raising operational costs.

“The hard part is knowing where to go: plastic drifts on currents and wind, scattering and concentrating in ways that are difficult to predict.”

The principle is simple even if the math is not. Instead of sweeping the ocean on fixed or intuition-based paths, the algorithm rapidly evaluates vast numbers of possible routes across enormous ocean areas and identifies the ones that intercept the most plastic per pass. "By optimizing routes in real time, we ensure every sweep collects as much plastic as possible, making cleanup operations significantly more effective," said den Hertog. The Ocean Cleanup has already integrated the solution into its operational software for Pacific Ocean deployment.

This is a quietly powerful example of AI for good: not a flashy new robot, but an optimization tool that makes existing cleanup hardware dramatically more productive with the same fuel, crews and time. Because the gain comes from smarter decisions rather than bigger machines, it also means less fuel burned and fewer emissions per tonne of plastic recovered, an environmental bonus on top of the headline efficiency number. The broader fight against ocean plastic still depends most on cutting how much enters the water in the first place, through better waste systems and less single-use plastic, and a 60 percent efficiency gain does not solve pollution at its source. The algorithm also relies on good forecasts of where plastic will drift, so its real-world performance tracks the quality of that underlying data. But for the slow, costly work of recovering what is already out there, smarter routing means more plastic pulled from the sea for every voyage, a clear and immediate win for the ocean and the people working to clean it.