Smart Innovative System to Optimize Salmonid Feeding in Chile

Avoiding food losses for both economic and environmental reasons has been a challenge, which has essentially relied on trained and experienced feeder operators, and then on operators capable of detecting the moment of satiety in the delivery of rations, with the support of cameras and ad hoc screens. But there were still challenges, such as identifying clear and advanced signals, not ex post, that allowed operators objectively identify the disposition and desire of the fish to initiate feeding, as well as the moment of satiety. Therefore, and for almost 5 years, a group of qualified academics and professionals linked to the industry and the subject in Chile, addressed this challenge, from the point of view of the behavior of the fish and its identification and processing, with artificial vision and artificial intelligence support. Thus the company 3se was born, which gave rise to “Intelligent Feeding”, a system that monitors online the appetite of fish in salmon on – growing centers, allowing to adjust this process in real time. On the other hand, the recorded behavior is useful for subsequent analyzes that permanently improve the feeding system. The development of friendly and easy screens supports the operators in charge with the application of indicators and alarms that prevent insufficient or excessive food delivery. The company decided that its system should act on any type of cameras and sensors and that it would benefit, as much as possible, the installed capabilities of the companies.

From the innovative company they point out that the great value of this technology is that it has place emphasis on the behavior of the fish, developing a sophisticated and complex system of algorithms that interpret numerous individual and collective reactions of the fish under observation, optimizing their feeding and providing information that is also useful for other applications such as early detection of diseases.

The intelligent Feeding System has two essential components: On the one hand, Feeding Assistant that supports operators in real time in their feeding function; and on the other, Feeding Evaluator, which generates daily and long-term reports for analysis about the efficiency of feeding and the procedures associated with it.

The System considers the use of high resolution cameras, usually present in farming cages that detect individual and collective behaviors of interest, along with other environmental parameters that deliver sensors ordinarily used in the centers. The information captured is directed to an analytical phase, where the observed behaviors are interpreted by more than seven sophisticated algorithms. Then, the fundamental outputs that the System delivers are: the activity of the fish that indicate whether or not they are ready for feeding as well as the condition of satiety.

The technical – academic team behind 3se is integrated by, Dr. Alfredo Illanes, a Chilean researcher, trained at the Santa María University and then PhD in signal processing at INRIA (France), and currently a professor at the OvGU University in Berlin ; Cristian Wente, Computer Science Engineer, with experience in neurobiological simulation and who has been residing in Chile for the past 7 years; Victor Valerio, its General Manager, with vast and recognized experience in the salmon industry in Chile and Andre Flem, a recognized and experienced executive and entrepreneur in the salmon industry.

For Flem, “this is a very reliable and efficient system with multiple projections, because it learns and improves based on artificial intelligence. For this reason and its enormous potential to deliver useful information on fish health and the effects of environmental challenges we have set out to promote it in Chile and other markets, such as Norway. ”

On the other hand, Victor Valerio points out that “3se group is also promoting other innovations based on integrated digital technologies, such as: early detection of diseases based on fish behavior; induction of behaviors in fish to optimize essential processes, respecting the welfare of farmed fish; Caligus detection system through images that allow more objective evaluations of epidemiological evolution and treatment effectiveness; modeling of carrying capacity of culture sites, considering oxygen demand, physiological requirements of fish and oceanographic conditions, and biomass estimators based on stereoscopic cameras for 3D vision. For this executive, “the digital transformation of the industry is an urgent process and in full swing, which should expand very quickly. We know that this is a challenge that not only implies technical changes but also cultural changes in companies to understand integrated technologies and their effective potential along the value chain. We are prepared to work with them in this context, and by the way, beyond the strictly technical fields”.