Applying artificial intelligence (AI) techniques to implement a practical Smart Cage Aquaculture Management System

Project Intro
The present research project aims to apply state-of-the-art artificial intelligence (AI) techniques to implement a practical Smart Cage Aquaculture Management System (SCAMS) useful for minimizing the cost of farming and maximizing the harvesting, and to further facilitate our country establishing a near-shore aquaculture industry of world-class competitiveness. Economic benefits of SCAMS include: to trigger the transformation of Taiwan’s fishery, enabling fishermen to obtain real-time information for achieving the best aquaculture performance, as well as promoting Taiwan’s current saturated fishing production. In addition, an AI farming services company will be established upon the completion of this project to provide fishermen with various technical services. R&D personnel concerning this project includes: an AI team, an automation team and an aquaculture team, all members of which are experienced scholars associated with NTOU (National Taiwan Ocean University), the aforesaid three teams will be joined by related enterprises such as the Taiwan-headquartered shrimp feed and farming giant Grobest group. NTOU also owns the oldest farming department in Taiwan and the various advanced research centers, Grobest group has both near-shore and offshore aquaculture experiences and skilled professional team, all these advantages make us superior over other competitors in completing the proposed research project. This four-year project is planned to accomplish the following two tasks: (1) to apply state-of-the-art AI techniques to the construction of an optimal aquaculture model, which is characterized by five sub-systems: Smart Feeding, Smart Image Behavior Monitoring and Analysis, Smart ROV, Smart Drone, and Big Data Analysis, the first four sub-systems collect data in real-time through various types of sensing components, these data will be uploaded to NTOU Cloud center and selectively labeled by the aquaculture team for the subsequent AI training at NCHC (National center for High-performance Computing) using state-of-the-art machine learning methods (e.g. deep learning) (2) enabling users to use mobile devices for monitoring the status of cages in real-time, herein an IoT architecture will be employed to integrate the Big Data for realizing the abovementioned practical benefits of intelligent cage aquaculture.