Introduction: Reducing production risks, increasing transparency to consumers, and ease of use solutions are the primary priorities for large aquaculture producers considering further or new investment in recirculating aquaculture (RAS) units in their production structure. Operators must focus on fine-tuning their energy, water, and feed to reduce their OpEx, often with constrains to the CapEx investment at start-up. Technological innovations to achieve better monitoring and control performance in various elements RAS, particularly in the areas of intelligent automated feeding (feeding costs can be up to 80% of production costs) and integrated farm monitoring, will have vital impacts on the operational expense and sustainability of RAS production.
Method: Innovations of the ongoing iFishIENCi Horizon 2020 project will be the next step in the future of RAS development, to ensure that the increasingly diverse needs can be met for the 1.4 million tonnes planned RAS production in the coming years. The resulting technology will improve feeding technology based on machine vision and digital twin development to reduce feed loss and improve FCR. iBOSS cloud connected technology for intelligent design and control of water treatment solutions, oxygen injection, fine solid removal, and disinfection (including ozone optimisation) uses water quality predictive AI modelling based on continuous data provided by in tank sensors. Camera integration coupled with a smart automated feeding system, based on modelling calculations of fish metabolism and machine vision, will allow for recognition of satiation behaviours to optimise feed delivery. The iBOSS innovation will be fully integrated with RAS production technologies to deliver SmartRAS to fulfil increasingly diverse RAS production and research needs.
Results and Outlook: At current status, amount of feed dispensed and feeding pattern can be user determined to suit the needs of the specific fish species in RAS production. Feed is automatically dosed throughout the day while simultaneously taking water quality into consideration. Total dispensed feed is used to predict updated growth projections daily, based on tabled FCR, and the next feeding day automatically adjusted to meet production forecast. iBOSS applies an adjustment factor to the feeding model based on ‘Activity’ calculated from smart camera observation of the fish behaviour, further adjusting the daily amount of feed automatically dispensed. The iBOSS AI will self-evolve throughout the remainder of the project as changes in feeding intensity, water quality, fish behaviour, and growth feedback data continuously optimise the models. SmartRAS equipped with the iBOSS integrated management system will enable more precise and accurate research with more data, shorten trial duration, as well as lowering operational costs and time to market. In this increasingly competitive and regulated industry, extrapolative modelling and prediction of challenging water quality parameters and visualisation of production data will ensure farm managers have the tools needed to reduce production issues, manage disease and mortality, reduce input costs, monitor and valorise waste products, and demonstrate sustainability and transparency of production, resulting in a highly attractive consumer product with reduced investor risk.