June 16, 2023 — Researchers around the world are using AI and machine learning to identify and avoid bycatch.
Call it inevitable. If doctors are using artificial intelligence (AI) to read X-rays and MRIs, how long could it be before fishermen could use AI to identify bycatch?
While AI assisted bycatch control systems are not yet fully operational and on the market, they are close, and a number of U.S. and European players are collaborating on development.
“We’ve got a network of people all working on this,” says Noelle Yochum, a former NOAA bycatch reduction expert, now working for Trident. “We’re sharing information in the hopes of getting at least one over the finish line.”
While Yochum’s primary concern has long been halibut and salmon bycatch in the Alaska pollock fishery, the problem is worldwide. The current estimate of global discards in commercial fisheries is 27.0 million metric tons, with a range of from 17.9 to 39.5 million mt.
“I read that the value of that was somewhere around $84 billion US dollars,” says Hege Hammersland, Business Development Manager for Scantrol Deep Vision, a Norwegian company using AI to help reduce bycatch.
“Scantrol Deep Vision started about 10 years ago,” says Hammersland. “We developed a big box camera system that we commercialized for research in 2017.” According to Hammersland the company collected millions of images from the research model and used those to teach the unit’s computer how to identify size and species. “We’re working with a Spanish company, Gerona Vision Research, that specializes in machine learning.”