Digital twin for optimising pesticide application in crops

Overview

This solution is a digital twin designed to predict how pesticide applications influence crop yields. By integrating mathematical and computational submodels across multiple biological scales, the platform simulates pesticide absorption, distribution, metabolism, and excretion within crops. This enables researchers and agricultural stakeholders to evaluate how the timing and level of pesticide application affect plant physiology and to identify strategies for maximising the benefits of pesticide use while minimising waste and environmental impact.

The digital twin addresses a critical challenge in modern agriculture: balancing the need for crop protection with economic efficiency and sustainability. By providing a predictive, data-driven framework, the platform supports more informed decision-making around pesticide selection, dosage, and timing, ultimately helping to improve yields across a range of crop species.

Technical specifications

Core modelling approach:

• Multiscale mathematical modelling integrating processes from genetic and cellular levels through to whole-plant architecture and resource allocation
• Integration of pesticide ADME (absorption, distribution, metabolism, excretion) into established crop-growth frameworks
• Built upon OpenSimRoot, an established 3D crop-growth modelling framework previously validated for evaluating drought and nitrogen fertilizer effects on rice, maize, and bean yields

Building-block submodels include:

• ODE-network models of gene regulation
• Vertex-based multicellular models of cell-to-cell transport
• Biomechanical models of cell growth
• Graph-based models of plant architecture and resource allocation

Data integration capabilities:

• Incorporates transcriptomics, proteomics, biosensor data, GFP-marker lines, transporter assays, and phenotype data
• Uses parameter optimisation, machine learning, and model-based experimental design to refine predictions
• Customisable framework adaptable to multiple crop species

Technology readiness level

The digital twin builds on a strong foundation of previously published and validated multiscale plant development models developed over 15 years of research. These mechanistic models have been demonstrated in high-profile publications and have been successfully applied to evaluate environmental factors such as drought and nitrogen fertilisation in rice, maize, and bean. The proposal outlines a clear plan to integrate pesticide ADME processes into these existing frameworks using OpenSimRoot. The platform is currently at a conceptual-to-prototype stage, with development requiring funded research time and a research assistant to assemble and validate the integrated digital twin.