Fishing for Data

Countless industries are being transformed by modern advances in artificial intelligence (AI). One of the more visible impacts is in the area of smart agriculture, where intelligent systems are helping to increase crop yields, detect diseases in livestock before they become widespread, optimize watering and fertilizing schedules, and much more. However, there is still an area of agriculture that has been underserved by technological advances to date — aquaculture.

More than one-third of the world’s population relies heavily on seafood to meet their daily protein requirements. Accordingly, developing effective and efficient aquaculture practices is essential for the health of both humans and the oceans. But the underwater environments found in this area present engineers with many challenges not found on traditional farms. And that makes developing intelligent monitoring systems exceedingly difficult.

A company called TidalX AI is working to develop monitoring systems that can meet the challenges of these environments and give fish farmers the same advantages as their landlubber counterparts. They are building a suite of AI-powered hardware and software tools designed to optimize fish farming operations. Their work is already transforming how the aquaculture industry monitors fish health, manages feeding, and ensures sustainable practices.

One of TidalX’s earliest developments was a computer vision system capable of analyzing live underwater video to estimate the biomass — weight and population distribution — of fish in real time. Traditionally, this has been a time-consuming and laborious manual process. But by training AI models using data captured from a specially designed fish run — a circular saltwater tank where salmon swam continuously, TidalX was able to automate the job.

To deploy these models in the field, TidalX created a rugged hardware platform featuring stereo cameras, lighting, and environmental sensors, all housed in gear tough enough to withstand frigid waters, strong currents, and the corrosive effects of salt. These systems are suspended in the fish pens and autonomously moved by a smart winch to scan fish populations day and night, collecting tens of thousands of observations daily.

The team also put AI to work to identify feed pellets in the water and monitor how fish respond. This works in conjunction with an autonomous feeding system that adjusts feed rates in real time. This minimizes waste, reduces pollution, and ensures the fish receive just the right amount of food to promote healthy growth. Given the importance of proper feeding schedules, this tool was initially met with skepticism by the farmers. But by giving users control over feed parameters, the team found that users gradually gained trust in the AI.

Beyond fish farming, TidalX’s technologies are already expanding into other sectors of the blue economy, including mapping seagrass for carbon sequestration and inspecting underwater infrastructure. With over 700 systems deployed globally and billions of data points collected, TidalX AI is proving that smart aquaculture is not only possible, but also essential to the sustainable future of our oceans.