UC Berkeley

Agriculturally viable land has been the target of renewable energy production efforts, such as solar panels and wind turbines. The competition between energy production and agricultural production has led to restrictions on non-agricultural activity on agricultural land while pushing for sustainable agriculture and renewable energy production to reach the state’s carbon-neutrality goals. Next-generation food systems are needed to alleviate such problems. The goal of this study is to analyze next-generation food systems from an optical modeling standpoint. This study develops a reduced-order geometric raytracing model to evaluate the performance of various food production systems, namely solar greenhouses, open-field agrophotovoltaics, and indoor pod farming systems. A digital-twin approach, where a digital replica of the physical system is modeled, is used to quickly and efficiently evaluate designs and optimize them using a genomic-based optimization algorithm. The digital-twin consists of modeling the optical properties of the system to accurately simulate the power distribution within the food systems through the raytracing algorithm. In addition, power sizing analysis of a real-life indoor farming system is performed. Extensions of the digital-twin framework and how it can be coupled with other physics models are provided using a crop performance driven optimization case study of an open-field agrophotovoltaic system. This computational framework and optimization scheme aims to provide a foundation for understanding, evaluating, and optimizing the food systems of the future and prove a useful tool to efficiently and sustainably produce food and generate power, driven by innovation and cutting-edge technology.