The “AI Underwater Monitoring and Management System” utilizes underwater cameras to obtain real time videos of shrimps day and night with visible and infrared lights. The images are enhanced (defogging and contrast enhancement, etc.) to face the challenges of different lighting, watercolor and turbidity. Next, the system automatically detects the number, size, activity and eating status of shrimps and measure the residual amount of feed pellets. Therefore, it can intelligently control the feeding quantity and schedule, improve water quality, resulting better growth, water quality and lower costs. 
These technologies are implemented with software running in the cloud server. On the other hand, some key technologies are implemented as AI circuits directly at the edge (pond) sides to further improve system performance, reduce power consumption, network bandwidth and data storage requirement. It is conducive to large-scale deployment and protection of intellectual property rights.

The “Artificial Intelligence Underwater Monitoring and Management System” utilizes underwater cameras to obtain real time videos of shrimps day and night with visible and infrared lights. The images are enhanced (defogging and contrast enhancement, etc.) to face the challenges of different lighting, watercolor and turbidity. Next, the system automatically detects the number, size, activity and eating status of shrimps and measure the residual amount of feed pellets. Therefore, it can intelligently control the feeding quantity and schedule, improve water quality, resulting better growth, water quality and lower costs.

As for industrial application potential, computer technologies and aquaculture are both the important industry in Taiwan.  Our team consisting of professors who specialize in aquaculture, computers, microelectronics and AI integrates growth knowledge of shrimps and aquaculture technology into smart underwater monitoring and managing system.
It is the trend of future aquaculture to use AI shrimp farming techniques to increase production, decrease risk, save manpower and reduce costs. We have submitted over 6 patent applications for our system. The possible business models include sale, licensing, rental, monthly cloud service, branding, etc. Consequently, our system will become a bright spot of aquaculture industry in Taiwan and have the opportunity to entry global markets.