The cAIge project aims to transform the way we monitor and assess salmon in fish farms. It plans to use advanced computer vision and artificial intelligence technology to automate the welfare monitoring and assessment of salmon. The project is organized into four distinct work-packages.
Work-Package one is focused on developing self-learning algorithms that can automatically detect high-level features of salmon, such as the eyes, fins, and body contours, from video streams. By identifying deviations from the standard appearance of these features, the system might detect anomalies that could indicate welfare issues, such as eye defects, wounds, or deformations.
Work-Package two is centered around the use of biometric techniques to identify individual salmon over time. This allows for long-term tracking of individual fish health and welfare status, a promising concept that could significantly enhance our understanding of individual and group health dynamics in full-scale fish farming.
Work-Package three aims to establish a correlation between fish welfare and various environmental and operational factors. By using AI reasoning strategies, the system will try to associate individual fish observations with overall fish population well-being, environmental conditions, and farming operations. This approach could help identify key contributors to fish health and enable the prediction of welfare scores under different scenarios.
Work-Package four is dedicated to the experimental validation and evaluation of the solutions created in the previous work-packages. This work-package also aims to ensure that the methods and algorithms developed are biologically accurate and relevant.
This project represents a significant leap forward in the area of computer vision and aquaculture. By automating the process of welfare monitoring, it contributes to enhancing the health and well-being of farmed salmon, while also improving the efficiency and sustainability of farming operations.
Technology development in cAIge will allow automated, machine-assisted monitoring of fish health and welfare in aquaculture farms at hitherto unprecedented precision levels. This is to be achieved by leveraging advanced AI-based computer vision technology along with individual fish re-identification algorithms for uninterrupted welfare assessment over time. In addition, the relationship between observed welfare status, environmental factors, and farming practices will be integrated into comprehensive models, thereby gaining a better understanding of how these factors interact to shape optimal holding conditions of fish in aquaculture. We emphasize that monitoring the welfare and responses to operational and environmental factors of individual fish over a prolonged period of time represents a significant advancement over the current practice to monitor the fish, thereby enabling researchers in the future to assess fish welfare and behavior on a scale never experienced before.