AI uses satellite imagery to detect plastic in ocean

The new AI developed by the Swiss Federal Institute of Technology in Lausanne and Wageningen University does the analysis task more successfully than previous models, according to its developers.

According to estimates by the World Wide Fund for Nature [WWF], from 2020, 4.8 to 12.7 million tons of plastic find their way into the oceans every year. / Photo: Getty Images
Getty Images

According to estimates by the World Wide Fund for Nature [WWF], from 2020, 4.8 to 12.7 million tons of plastic find their way into the oceans every year. / Photo: Getty Images

An artificial intelligence (AI) developed by a Swiss technology institute and a Dutch university uses satellite imagery to detect plastic in the ocean.

The new technology, developed by the Swiss Federal Institute of Technology in Lausanne and Wageningen University in the Netherlands, could help to remove plastic waste from the oceans using ships systematically, the Dutch university announced in a press release on Thursday.

The newly developed AI estimates the probability of the presence of waste for each pixel in satellite images, according to the statement.

It also said that although the accumulation of waste in the sea is visible on the freely available Sentinel-2 satellite images from the European Space Agency, it requires the help of AI for the analysis, as it involves terabytes of data.

According to the developers of the AI, the new technology solves the task more successfully than previous models.

One of the reasons for this is that it can detect plastic even in difficult atmospheric conditions with clouds and haze.

This is particularly important as plastics are often washed into open waters after rain.

"Detecting plastics in marine debris under difficult atmospheric conditions with clouds and haze is particularly important, as often plastics are washed into open waters after rain and flood events," the press release said.

Read More
Read More

Plastic litter in oceans overestimated, study suggests

Double view

Aside from the more accurate predictions, the detection model will also notice debris in daily-accessible PlanetScope images.

This, according to researchers, will help "close the gap towards continuous daily monitoring."

"This double view of the same object at two locations reveals the drift direction due to wind and ocean currents on the water," Marc Rußwurm, Assistant Professor at Wageningen University and former researcher at EPFL, said.

"This information can be used to improve drift estimation models for marine debris," he added.

According to estimates by the World Wide Fund for Nature [WWF], an environmental organisation, from 2020, 4.8 to 12.7 million tons of plastic find their way into the oceans every year.

Only a small proportion of the plastic waste floats on the surface, the rest is transported to deeper waters or the seabed and is almost impossible to retrieve.

According to the WWF, it is estimated that around 80 million tons have accumulated there.

Read More
Read More

'High and low-ambition' countries split over plastic pollution treaty

Route 6