Using underwater cameras, imaging sonar, and artificial intelligence, Nova Scotia-based Innovasea claims to have successfully created technology that instantly detects and counts fish, which could help avoid fish kills at hydroelectric dams.
95% Accurate
As reported first by CBC, The prototype was put to the test in the field last spring at a hydroelectric dam owned by Nova Scotia Power outside of Kentville, Newfoundland, where it counted approximately 900,000 fish during an upstream migration of a species of herring called gaspereau.
Compared to hand counting done to verify results, the aquatic technology company claims that the artificial intelligence system was 95% accurate.
"This is a new capability that gives them a lot more data, a lot more resolution of data, and potentially a lot more deep insights as to how fish are moving around the system," Innovasea Vice-President of research and development Jean Quirion said.
A fish ladder that allows fish to pass the 3.4 megawatt White Rock dam upstream was observed by an optical camera. A separate camera has also watched over the dam bypass that lets fish move downstream, according to CBC's report.
To ascertain if the fish were migrating upstream or downstream, two imaging sonar devices were positioned in the fish ladder's outlet canal. With minute-by-minute data on a mobile app, machine learning enabled a computer to count passing fish continuously throughout the day.
Since the insights are supplied in real-time, one is suddenly aware of them and has the chance to take action, according to Quirion.
Read also : Stanford Scientists Create 'OceanOneK' - A Humanoid Robot That Can Dive In Ocean Depths!
Fish-Detecting Technology
Nova Scotia Power agreed to contribute $50,000 to a federal environmental damages fund four years ago for triggering a major fish kill at the dam in 2017.
According to the company, there are already precautions taken to ensure fish passage. The White Rock dam is currently turned off by the corporation during the spring migration.
An acoustic tracking study was conducted in the spring to track the movement and behavior of fish marked near the White Rock dam to further validate the AI technology.
The next iteration of Innovasea's AI technology is now being refined to recognize specific species, including endangered ones like the Atlantic salmon.
CBC noted that camera-based AI systems work better in confined areas like fish ladders and rivers than in the ocean, where cameras have limited range.
The federal Ocean Supercluster innovation fund supported the $29 million Ocean Aware project, which includes around one-third of the fish detection technology created by Innvoasea. With $15.7 million from industry partners, the Ocean Supercluster contributed $13.74 million.
Related Article : Scientists Create A 'Robot Fish' That Eats Microplastics - Will It Solve Plastic Pollution?
This article is owned by Tech Times
Written by Joaquin Victor Tacla
