Aquaculture technology: develop intelligent box net culture system, hope aquaculture system cost reduction and management efficiency

With the increase of the earth's population, the demand for aquaculture consumption is also on the rise; however, due to the land cost and water quality test of land aquaculture, marine cage network has become an important development trend of foreign aquaculture. Offshore cage culture in Taiwan has been developed for more than 30 years, but the cost of management and facility loss of cage culture remains high because of the many typhoons in summer. Therefore, Ocean University and Chengda artificial Intelligence Biotechnology and Medical Innovation Research Center work together to create an "intelligent box net culture management system", hoping to reduce facility costs and improve the efficiency of culture management. At present, some hardware equipment development has been completed and integrated with the Internet of things system. This summer vacation, we will go to Pingtung to collect and test the data of the outer sea box network.

At present, countries that have developed cage farming for a long time, such as Norway and Japan, have introduced automated management, said Zhang Zhong, a professor in the Department of Electrical Engineering at Ocean University, the project moderator. But the integration technology of AI identification and big data in the two countries is not yet mature, and this is Taiwan's opportunity. The project combines research teams in the fields of information and electrical engineering, aquaculture and AI artificial intelligence, as well as cage farmers, Chunghwa Telecom and monitoring system operators, and is confident that a system that meets the needs of the industry can be designed in the next two and a half years.

System concept map.

Xu Dehua, assistant professor of aquaculture department of Ocean University, explained that the biggest difference between marine cage culture and land culture is the "distance" of management. Depending on the difference in fish species, if the stocking location is in the sea area 2 to 3 kilometers away from the shore, it is impossible for farmers to inspect the culture area every day, so the ability to master the growth status of aquaculture is limited. If there is enough data to assist in decision-making, the survival rate of aquaculture can be improved. In addition, there are many typhoons in Taiwan, and tests on underwater cameras and sensing instruments will be more frequent, and the relative price of equipment abroad is relatively high. For farmers, a high replacement rate will be a great expense. Therefore, how to maintain the accuracy of detection and meet the needs of cost is also a problem that this plan hopes to solve.

The project has been going on for about a year and a half, and the research team has successfully integrated data from surface, underwater cameras and sensing instruments in the indoor culture pond of Gongliao Aquatic Biology Research Center so that AI can control the feeding system according to ripples and identify the body length of fish. In particular, in order to reduce the cost of arranging monitoring equipment in many places in the cage network, the research team made a low-cost IoT system, including temperature sensors, power supply systems, sensor control devices and other sets, as well as anti-theft gyroplane aircraft that can take off regularly. As soon as an unknown ship enters the breeding area, it will immediately notify farmers to reduce the chance of fish being stolen in the open sea.

Next, the research team will go to the Pingtung box net aquaculture farm to operate, collect culture data, and design box net floating and sinking devices; in addition, it will also strengthen the technology of underwater cameras to identify fish in wind and waves and their anti-pollution ability. it is not ruled out to combine sonar and other ultrasonic techniques to identify the physiological characteristics of schools of fish in order to increase the accuracy of decision-making.

The underwater camera and image monitoring system of the city proving ground. (photo / photographed by reporter Zhuang Xiaoping)