Digital twins are virtual replicas of physical systems, continuously updated with real-time data. In DiAGRI, we develop digital twins of our growth chamber environments — enabling researchers and growers to simulate, predict, and optimise crop production without the constraints of physical experimentation.
What is a Digital Twin?
A digital twin is a computational model that mirrors a real-world system in real time. For CEA, this means creating a virtual growth chamber that reflects the actual temperature, humidity, light conditions, and plant status of its physical counterpart — updated continuously through IoT sensor data.
How DiAGRI's Digital Twins Work
- Data Collection: Environmental sensors and imaging systems in the growth chambers continuously stream data — temperature, humidity, CO₂, light spectrum, soil moisture, plant 3D structure, spectral signatures, and fluorescence responses.
- AI Integration: Machine learning models process this multi-modal data to estimate plant state (growth stage, stress level, biomass, photosynthetic efficiency) and predict future development.
- Virtual Environment: The digital twin platform combines sensor data and AI predictions into an interactive 3D virtual model of the growth chamber and its crops.
- Simulation & Optimisation: Researchers can test "what if" scenarios — adjusting virtual temperature profiles, light recipes, or nutrient regimes — to predict crop responses before implementing changes in the real environment.
Applications
- Scenario Simulation: Test different growing protocols virtually before committing growth chamber time
- Experimental Design: Optimise experimental parameters in silico to maximise information from limited physical experiments
- Growth Condition Optimisation: Find the optimal combination of environmental variables for target crop traits (yield, quality, resource efficiency)
- Remote Monitoring: Access a live virtual view of growth chambers from anywhere, enabling collaborative research across sites
- Education: Use digital twins in university courses on computational biology, plant modelling, and CEA engineering