Fisherman in boat cabin viewing fishing data on tablet.
An arrow pointing leftHome

How Oregon fishers are embracing AI

  • Mike Pearl

A short documentary features tech from Oregon- and B.C.-based startups.

With fisheries all over the world in very real danger of imminent collapse, marine fish need all the help they can get, and an AI-based answer to overfishing might be one such tool. Fishers can be quick to lash out at restrictions and bureaucracy they deem onerous or intrusive, like extremely low-tech “at-sea” monitors working for the federal government. But ideas for high-tech — and hopefully more elegant and efficient — controls on bycatch and overfishing could be a welcome change, according to a recent short documentary.

According to the trade publication National Fisherman, the companies providing the technology featured in the 20-minute film include Medford, Oregon-based CVision AI, as well as Teem, which is based in British Columbia.

Perhaps in no small part due to the involvement of these PNW companies, the short film “Fishing Smarter,” created by the Environmental Defense Fund, makes fishermen working off the coast of Oregon into an intriguing case study in AI and fishing. It’s worth noting at the outset that the EDF has been criticized for promoting policies that favor large fishing corporations over mom-and-pop ones, but this use of AI may deserve a look, regardless of who is touting it.

The video depicts multiple uses of computer vision on fishing boats.

Fishing Smarter: How new technology can help our oceans thrive

One concept for an AI monitoring system shown in the video would use Teem’s technology to record time lapse footage of fishermen doing their jobs, which would then be analyzed by a trained AI that can tell what the crew is up to. That information would then be synced with GPS data, creating a map of various fishing activities. “They’re setting at this point. They’re hauling at this point. They’re sorting at this point,” says Teem CEO Amanda Barney in the video, explaining the concept.

This prospect does at least seem less irritating than having an actual human monitor looking over your shoulder while you do your job.

Another computer vision application would replace burdensome manual analysis and handwritten recordkeeping of fish species. Each rejected fish would be fed into a “photo booth” on a conveyor belt and recorded by an AI camera system. The AI would analyze each fish by size and shape, and a determination could be made about which fish species are being extracted where. Armed with this knowledge, managers could reduce bycatch, and stave off the depletion of fish stocks.