Digital twins for land-based aquaculture: A case study for rainbow trout (Oncorhynchus mykiss)

Abstract

The virtual, digital counterpart of a physical object, referred as digital twin, derives from the Internet of Things (IoT), and involves real-time acquisition and processing of large data sets. A fully implemented system ultimately enables real-time and remote management, as well as the reproduction of real and forecasted scenarios. Under the emerging framework of Precision Fish Farming, which brings control-engineering principles to fish production, we set up digital twin prototypes for land-based finfish farms. The digital twin is aimed at supporting producers in optimizing feeding practices, oxygen supply and fish population management with respect to 1) fish growth performances; 2) fish welfare, and 3) environmental loads. It relies on integrated mathematical models which are fed with data from in-situ sensors and from external sources, and simulate several dynamic processes, allowing the estimation of key parameters describing the ambient environment and the fishes. A conceptual application targeted at rearing cycles of rainbow trout (Oncorhynchus mykiss) in an operational in-land aquafarm in Italy is presented. The digital twin takes into account the disparate levels of automation and control that are found within this farm, and considerations are made on preferential directions for future developments. In spite of its potential, and not only in the aquaculture sector, the development of digital twins is still at its early stage. Furthermore, Precision Fish Farming applications in land-based systems as well as targeted at rainbow trout are novel developments.

A Novel Process-Based Digital Twin for Intelligent Fish Feeding Management Using Multi-Mode Sensors and Smart Feeding Machine

“…A digital twin serves as the digital representation of physical devices based on sensor data. While digital twins have been extensively integrated into various applications, such as smart cities and manufacturing processes [17][18][19], their full realization in aquaculture is still evolving, despite various attempts and preliminary integrations [6,[20][21][22][23]. In aquaculture, digital twins can be instrumental in monitoring parameters like temperature, light intensity, water flow, pH levels, salinity, and even surveillance videos.…”

“…In fish farming, digital twins can proactively monitor fish health, preventing diseases by simulating optimal farm conditions and applying them to physical aquaculture processes. For instance, in a previous study [23], a digital twin approach was explored in the context of land-based smart aquaculture. The successful integration of physical devices and digital twins relies on several key factors: (1) employing multi-mode sensors to capture high-quality data for the digital twin, (2) incorporating domain-specific knowledge, such as fish physiology, into the AI functions of the digital twin, and (3) endowing the digital twin with learning capabilities to calibrate crucial modeling parameters based on historical data from physical devices.…”