2025/01/01 by Katarína Smoleňová, Nastassia Vilfan, Daniela Bustos‐Korts +5 · 1 voice
Agricultural and Biological Sciences · Engineering · #Greenhouse Technology and Climate Control #Smart Agriculture and AI #Digital Transformation in Industry
paper · doi:10.1093/insilicoplants/diaf022
openalex publication_date 2025/01/01 · openalex created_date 2025/11/28 · openalex updated_date 2026/07/29
Abstract Digital-twin technology is a promising decision-support tool in controlled-environment agriculture, that can be applied to optimize agronomic production, improve management decisions, and formulate data-driven breeding strategies. We developed a concept for a digital twin of a tomato crop in a high-tech greenhouse with the aim to help increase resource-use efficiency of greenhouse tomatoes. At the core of the digital twin is a functional–structural plant (FSP) model that can simulate 3D architectural development and growth of young tomato plants and can be used to predict plant responses to environmental factors and management interventions. In this study, we present the newly developed tomato FSP model and procedures to use data from climate sensors and image data of single plants from high-throughput phenotyping, to feed the FSP model to ensure synchronization with the real-world crop development. A greenhouse climate model is used to calculate the indoor climate above the crop based on outdoor weather data, greenhouse properties, and climate control settings. Synchronization of plant architecture between the real and the virtual crop is demonstrated for plant height adjustments, by performing model calibration based on Bayesian optimization. The tomato FSP model is designed to study the role of individual organ traits and assess the effects of architectural manipulations, such as leaf pruning, and lighting strategies. The presented framework addresses a vital component of a digital twin representing the flow of information from the real crop to the virtual crop.