Israeli aquatech start-up GoSmart counting on NVIDIA AI platform to increase fish farm feed efficiency

Israeli aquaculture technology start-up GoSmart is leveraging artificial intelligence (AI) powered by NVIDIA’s Jetson platform to improve fish farm feed efficiency. 

Founded in 2020, GoSmart is using AI to help farmers optimize fish feed regimes. Aquafeed firm Skretting, is currently using GoSmart technology to create optimal fish feed regimes.

“Managing the right levels of fish feed saves a lot of money for the farmers and reduces organic matter from excessive debris in the aqua environment,” GoSmart Co-Founder and CEO Josef Melchner said in a release.

GoSmart's system, which is roughly the size of a soda bottle, is fully autonomous and can be attached to aquaculture cages, ponds, or tanks, according to a company release. The system, powered by the NVIDIA Jetson platform – a set of small-form-factor modules compatible with NVIDIA’s AI platforms – analyzes the weight and population distribution of fish, along with temperature and oxygen levels. The information is then fed in real-time to farmers to help determine the optimum time to feed and harvest fish, and to give an update on the health and welfare of the fish being farmed.

The GoSmart system uses lithium-ion batteries charged via solar panels and each module is equipped with power-management software enabling it to autonomously enter sleep mode, shut down, wake up, and conduct its work while managing its power needs. 

GoSmart is member of the NVIDIA Metropolis application framework – a set of tools that NVIDIA said brings visual data and AI together to improve operational efficiency. Melchner said the company chose NVIDIA Jetson specifically due to its small form-factor and affordability, and that it is currently training its systems to analyze fish behavior. 

By using AI, GoSmart said, the company can run multiple AI algorithms in parallel allowing a number of different characteristics to be analyzed simultaneously to create a more complete picture of fish health. The AI is currently being trained, with the initial step involving manually inputting data for thousands of fish. 

“There was a lot of diving and many underwater experiments,” Melchner said. 

According to Melchner, GoSmart could potentially help fish farmers reduce the amount of fish feed they need by up to 15 percent. It can also help aquaculture operators shorten fish growth-time by optimizing feed regimes for fish health.

Melchner said Go Smart hopes to expand its systems into a complete feeding solution, including a fully integrated autonomous feeding barge that could give fish the exact amount of food they need, when they need it, without the need for close human observation of the fish farm. 

“Our goal is to have our systems in every cage, every pond, every tank in the world – we want to cover the entire aquaculture industry,” he said.  

Photo courtesy of GoSmart