Archipelago, a global provider of marine resource management products and services and Ai.Fish, a software startup developing artificial intelligence (AI) tools for marine conservation and ocean science, are pleased to provide an update on the ongoing development of AI-based tools through their partnership formed in July 2021.
At the onset of the partnership, Archipelago and AI.fish looked to address the challenge of accurately identifying fishing activity on longline fishing vessels. “This has been a challenge since the inception of EM as the longer the fishing trip and the more cameras carried by the vessel, the more hours of footage to review just to identify locations that a fishing event begins and ends,” Archipelago President and CEO, Gord Snell. “This results in more review time and more cost to the vessel owner to meet their reporting requirements and we wanted to find ways to reduce the effort and in-turn, the cost to our clients.”
Utilizing data from longline vessels on multi-day trips, an algorithm was developed, tested, and refined that allowed the system to automatically detect gear set and gear haul events with an accuracy level greater than 90%. As the algorithm continues to learn and improve, this number will move even higher. This advancement alone will reduce the review time by an estimated 8-10% for the type of longline trip utilized in the development.
“We are thrilled to see this kind of success in such a short window and look forward to continued success as we build on our partnership in the coming months. We are actively looking for industry partners who are interested in addressing their fisheries monitoring requirements and want to explore the use of AI in their data analysis,” stated Mr. Snell.
“We formed this partnership three months ago to move AI development efforts forward more rapidly within fisheries monitoring than we have seen in recent years,” said Jimmy Freese, CEO of Ai.Fish. “We wanted to showcase the knowledge of our two organizations and how quickly we could start to make an impact on the efficiency of data review within operational EM programs that Archipelago is involved in throughout North America and Australia. Our early results are very encouraging, and our hope is these efficiencies can be adopted across other electronic monitoring pilots and programs throughout the world.”
The next step in this development will be to implement the algorithm into Archipelago’s industry-leading EM data analysis tool, EM Interpret (EMI) and allow clients to begin to benefit from the cost savings this AI algorithm can deliver.
Archipelago pioneered EM in the late 1990’s and have continued to be a global leader in the advancement of the technology for more than 20 years. They currently support 500+ vessels carrying their EM products. Through their data review software, EM Interpret (EMI), clients review more EM fisheries data each year than any other software tool in the world. In 2017, Archipelago partnered with Marine Instruments in Spain to develop and deliver the next-generation of EM hardware solutions and together they continue to work towards the advancement of the EM Ecosystem through existing and future clients.
Ai.Fish is a software startup aiming to harness the latest artificial intelligence and machine learning research to address problems in marine sustainability and conservation. In 2020, with support from the National Fish and Wildlife Foundation, Ai.Fish developed machine learning and computer vision algorithms for analyzing EM video, demonstrating the efficacy of AI technology in streamlining the EM review process through initial user trials in Hawaii. This partnership will enable them to deploy and test their AI processing at scale with feedback from users in a broad range of fisheries.
In an effort to advance the utilization of AI within fisheries, Archipelago and AI.fish will continue to support other AI advancement efforts throughout the industry such as The Nature Conservancy’s Fishnet initiative (fishnet.ai).
Visit www.archipelago.ca or ai.fish for more information or contact Archipelago at info@archipelago.ca or AI.fish at contact@ai.fish
by archipelago
Nov 4, 2021 - 12:41 am
Share This Post: