A Modeling Methodology for Crop Representation in Digital Twins for Smart Farming

Digital twins of complex systems are operated by stakeholders from different domains, who typically do not work in the same language. This problem is exacerbated in digital twins where domain-specific representations are required to convey actionable results, such as in cyber-biophysical systems. Particularly, in controlled environment agriculture, agronomists devise seasonal production plans and run simulations to optimize the system in terms of crop phenology while growers maintain crops and ensure their optimal growth by assessing crop morphology. To breach this gap, we consider an optimization problem to reconcile the different users' points of views. We propose a modeling methodology that bridges the gap between crop phenology and morphology, generating visual representations of crops based on simulated phenological characteristics. To demonstrate the validity of our proposed methodology for digital twins in smart farming, we apply our approach to two case studies: a strawberry vertical farm and a smart canola field.