Virtual Tomato Crops (VTC)
The current need for sustainable production of food with minimal environmental impact requires
high-tech solutions that facilitate and automate greenhouse crop management. We propose the
construction of the "Virtual Tomato Crops" (VTC) – a digital twin of real tomato greenhouse crop
systems that provides actionable knowledge for greenhouse tomato crop management. We describe the plan for the construction and testing of a prototype of the VTC.
Work on the VTC is divided in four work packages (WP). In WP1, the system boundaries and the scope of the VTC are defined, making use of a user committee of key players in the Dutch greenhouse industry. The VTC is further evaluated using key economic and sustainability indicators. In WP2, modern sensor technology, machine vision and deep learning algorithms are used to estimate parameters of plant growth, development, physiology, and architecture. In WP3, plant greenhouse climate models will be combined to simulate plant growth and development in unprecedented detail, and these will be parameterized using parameters from WP2. The combined plant-greenhouse climate model will further be evaluated for sensitivity to, and uncertainty of, changes in single parameters. In WP4, optimization and control theory will be used to evaluate the VTC, using three distinct use cases at short- (hourly), mid- (weekly) and long-term (annually) decision support. Experiments for data acquisition required for WPs 2 and 4 will be conducted in the newly built NPEC greenhouse at WUR, and a range of tomato varieties will be included.