Swimming Robots? Yes, to Clean the Oceans

January 28, 2025

Maybe you’ve seen the videos: an aquatic robot, in the ocean or a river, gobbling up enormous quantities of floating garbage.

They’re impressive. And so is the goal. Removing debris from the world’s water. In 2019 there were an estimated 5.25 trillion pieces of plastic litter in the ocean weighing more than 268,000 tons. But plastic isn’t the only material choking oceans and streams today. When all materials are taken into account, recent estimates say there may be more than 171 trillion pieces of trash floating in the ocean.

Small pieces of ocean trash could be eaten by marine animals. If they remain in the animal’s body, they could create a platform for infection or travel farther up the food chain as smaller animals are eaten by larger ones. Larger pieces of marine debris could impact shipping — about 10,000 shipping containers fall into the ocean each year.

Robots can help identify this debris and even remove it from the ocean, but there are limitations.

Patrolling Rivers

Deployment of garbage-eating aquabots is limited, mostly by a lack of funding, but there are a couple of prominent examples. First, there’s the Jellyfishbot, which IEEE Spectrum has called a “Roomba for Rivers.” Some local governments, like Long Beach in California, use them.

Another example is the River Interceptor, developed by the Dutch nonprofit The Ocean Cleanup. It stays anchored in the river. As garbage floats down the river, it is guided to the solar-powered bot’s mouth along a floating boom, and is then removed from the water with a conveyor belt.

These two bots have some differences. Jellyfishbot is relatively small; it would fit in the trunk of a car and is remote-controlled. The River Interceptor, on the other hand, is much larger, about twice as long and wide as a standard shipping container.

But both can help tackle a major problem: most of the trash in the ocean flows there from rivers. Which means that removing trash from those environments offers a bigger impact.

A Rapidly Improving Technology

Development of aquabots also faces some challenges. For one thing, not all garbage remains on the surface. Some sinks to the bottom, and some floats below the surface. And surface-dwelling robots can’t reach them.

Another challenge: visibility can be limited underwater. A robot operating below the surface may have trouble identifying marine debris. As a result, most prototypes for underwater garbage robots rely on human operators to identify garbage. There is, however, significant research into automatic, AI-based garbage identification methods. This article in IEEE Access provides an overview of research into that field, and shows a new method that increases accuracy.

“Most marine debris robotics projects are pilot projects. We need to sort out some of the core issues of marine debris to solve the problem of large-scale removal,” said IEEE Senior Member Wang Gang. “Most of the projects address surface litter, which is only one part of the debris. Most litter may be underwater, such as the fishing nets.”

Do We Need Aquabots to Clean the Ocean?

The amount of man-made debris in the oceans is staggering. And some estimates suggest that the volume of garbage could quadruple by 2040.

Robots, notes IEEE Member Paulo Drews, are expensive. They require maintenance, and operators. They also need somewhere to put the garbage.

“Using aquatic robots to deal with litter is an important direction for society,” Drews said. “Still, there are cost complexities to be considered as this technology improves. The robots can help, but humans need to reduce the amount of trash in the water bodies as soon as possible.”

Learn More: The ocean is vast, and so are the engineering activities. Check out these resources from IEEE TryEngineering to learn more about all of the opportunities found under the sea. And if you want to learn about the career path of an ocean engineer, check out this Q&A from the IEEE Oceanic Engineering Society.