Data-Driven Modeling for Crop Growth in Plant Factories

Abstract

AbstractPlant factories, also known as vertical farms, are next-generation agricultural systems that integrate various modern technologies. They are almost completely insulated from the external climate and are able to precisely and automatically control internal environmental factors such as light, temperature, humidity, CO2 concentration, and nutrient solution. In such intensive agricultural production systems, massive crop-related data have emerged from equipped sensors and monitoring hardware. Data-driven modeling approaches can help describe the dynamic process of crop growth and provide information for plant factories’ decision-making. Improving energy efficiency and reducing the production costs of plant factories via the digitalization of the crop growth process represent essential research directions. In this chapter, the applications and prospects of data science in plant factories are described from three perspectives: environmental factor sensing, crop growth monitoring, and crop growth models. The physical and chemical properties of the environmental factors and their automated measurement procedures are described. In addition, the basic concepts regarding how the environment affects crop growth are discussed. In terms of techniques to automatically monitor the crop growth status in plant factories, we describe the concepts and applications of data-driven approaches from growth-related properties measurement and abnormal growth detection perspectives. Based on environmental factors and crop growth data, the growth processes of crops and their responses to the environment can be simulated quantitatively. The basic principles of crop growth models are illustrated, and applications of data-driven crop modeling in plant factories are outlined.KeywordsPlant factoryData-driven modelingEnvironmental monitoringGrowth monitoringCrop model