Reinforcement Learning Based Production Planning in the Aquaculture Industry

Production planning and decision-making processes in the aquaculture industry are becoming increasingly complex. Decision-makers must take into consideration a large number of factors - economic, biological, environmental, etc. - and, in addition, the increasing size of enterprises and intensive production systems have exponentially increased the number of decisions to be made. The aim of this article is to evaluate the possibilities of applying a novel technique, such as reinforcement learning, to decision-making processes in aquaculture. To this end, its application in comparable industries, also part of the agricultural sector, that are more technologically advanced, is reviewed in order to use them as a reference in the short and medium term. As can be observed in this article, reinforcement learning not only offers an interesting alternative for dealing with this complexity but also makes it possible to tackle problems that had not been solved until now.