Robotic fish revolutionize aquaculture with AI precision

At the National Innovation Center for Digital Fishery in Beijing, two robotic fish with a sleek, metallic finish swim gracefully in a breeding tank, showcasing cutting-edge research aimed at revolutionizing fish farming.

"These are the robotic tuna and robotic dolphin we developed," said Liu Jincun, an associate professor at China Agricultural University (CAU). "They represent different swimming characteristics, but their core purpose is the same – to assist in fish farming."

Liu is part of an underwater bionic robotics team dedicated to simplifying the labor-intensive tasks of fish farming.

"We want to make the work of fishermen less difficult," said Wei Yaoguang, another team member with more than 20 years of experience. Wei recalled how fishermen used to inspect vast seafood farms, with divers spending three to four days checking a 400-meter-diameter net cage for sea farming. In addition to being time-consuming, the process was also costly.

To address these challenges, the team turned to underwater robots equipped with sensors to monitor fish and nets. Their first robot could patrol the entire 400-meter-diameter net cage in just four hours, a significant improvement over the manual process.

However, the team soon encountered a new challenge: fish stress response. "Fish are easily stressed, and if the robots get too close, they will jump out of the water," Liu explained.

To minimize disturbance, the team designed a smaller, more streamlined bionic fish that blends seamlessly into its environment without disrupting aquatic life. "When our fish robots swim, live fish often follow them, creating a natural dynamic where larger fish lead smaller ones," Liu said.

Another challenge arose with the robotic tuna's tail, which swayed from side to side, causing its head to move unpredictably and complicating underwater inspections. To solve this, the team developed a visual stabilization system, improving the robot's ability to clearly observe underwater conditions.

According to experts, these bionic robots incorporate advanced technologies, including fish-like propulsion systems and flexible dual-drive motors, which reduce disturbance and enable smoother swimming. They are also equipped with high-precision sensors to monitor water quality and fish behavior in real time.

Meanwhile, leveraging artificial intelligence, the team developed autonomous motion control algorithms and lightweight AI chips, allowing the robotic fish to swim, analyze and adapt autonomously in various underwater environments.

Experts believe the robotic fish opens new possibilities for fisheries management, such as guiding target fish to designated areas for more efficient harvesting and enhancing feeding strategies.

"Traditional feeding methods can result in waste or uneven distribution, but the robotic fish, with an integrated bait bin and sensors, can deliver food more precisely by analyzing bio-indicators such as size, number and activity level," Liu said. This reduces bait waste, lowers aquaculture costs and optimizes the growth environment for fish.

Beyond aquaculture, robotic fish have potential applications in deep-sea exploration and marine resource monitoring. "Traditional underwater equipment is often bulky, energy-intensive and disruptive to marine life. Bionic robotic fish, with their quiet, inconspicuous nature, can collect deep-sea data, monitor marine ecology, and assist with topographic mapping without interfering with aquatic ecosystems," Wei said.

So far, the team has collected over 200 terabytes of video data on more than a dozen fish species, along with 10 million photos and videos. "This data is essential for training AI models and will contribute to smarter, more efficient fish farming," Wei added.

The team has also integrated its intelligent aquaculture technology with the Fanli Big Model, an AI-powered system developed by the National Innovation Center for Digital Fishery. Deployed across 23 provincial regions, the technology now manages over 6.3 million square meters of land-based aquafarms and 5.5 billion square meters of pond and offshore aquaculture farms, cutting labor costs by 50 percent.

In addition to the robotic fish, the center has developed a range of "AI plus fishery" technologies, including water quality monitoring, fish behavior analysis, intelligent feeding and disease early warning, offering comprehensive technical support for modern fisheries.

Li Daoliang, director of the center and a professor at CAU, emphasized AI's transformative role in aquaculture. "AI's impact on fisheries will be profound, enhancing efficiency while also promoting sustainability."

"With the global population growing and food safety concerns rising, traditional fish farming faces challenges such as water pollution, resource waste and drug misuse. AI can address these issues and help achieve precise fisheries management," he said.

"Our team will continue exploring innovative solutions in underwater robotics to create a more efficient and intelligent future for fish farming," Li added.

(Cover: The robotic fish developed by the CAU team. /CMG)