Development of plant growth prediction model considering environmental history and evaluation of plant factory productivity

In recent years, artificial light-type plant factories have been attracting attention due to the demand for stable food production in response to food shortages and climate instability. Despite their high productivity and stable year-round cultivation, the high initial and operating costs are still a problem. However, with the recent development of AI technology and the third plant factory boom, research on yield prediction combining machine learning and image recognition has been conducted in the field of food production. This enables more accurate prediction than growth prediction based only on environmental conditions around the crop by measuring the shape and color of leaves through image recognition. On the other hand, it has been reported that the phenotypes that contribute to the growth rate of a crop are caused by environmental factors during its growth stage. In this study, crops were grown in a controlled environment from the seeding stage, and the relationship between environmental history and phenotypes was investigated. This study investigated the relationship between environmental history and phenotypes. The plant model was introduced into an actual growing environment, and the accuracy of growth prediction was confirmed.