AI-based modeling of photosynthesis as a function of radiation intensity, wavelengths, and - pulse modulation.

• Applicants: Professor Dr. Ralf Kaldenhoff; Professor Dr.-Ing. Tran Quoc Khanh
• Subject Area: Measurement Systems ; Plant Physiology
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 500805487

Photosynthesis is considered a well-described process. It is defined at a molecular basis and many influences, especially those of light, seem sufficiently well known. The findings were used to create mathematical, single- or multi-factor-based models that allow to predict photosynthesis performance under various defined light conditions. The established models apply only under the given conditions, which do not necessarily always correspond to the real ones. By using modern LEDs and AI-based control of lighting programs we will explore, a) the possibility of using these techniques, that are only available in recent years, for the study of photosynthesis and b) a new holistic model describing photosynthesis effectiveness will be established. Unlike conventional models, it does not work with individual factors integrated into a mathematical model. Rather, the entire plant system is considered a black box and the resulting data is represented by means of big data, AI, deep learning etc. in a multi- variant model of photosynthesis. This model is modified and variable. It is influenced by radiation intensity, wavelengths and pulse modulation reflecting plant species, plant age (developmental stage) and light history. For this purpose, a high-throughput measurement method is created, which includes, among others, the use of thermal cameras, reflection and absorption measurement, LED-based variable illumination and chlorophyll fluorescence measurement. Competent AI algorithms are developed with data from high- throughput screening. After model development, these are validated for defined plant species at different stages of development and tolerance ranges are defined. The description of photosynthesis in the targeted AI-based model will change our image of the process that is so important to all of us, refine it, and make it possible to create a novel overall picture. The model will have an impact on all areas of plant cultivation; Agriculture and greenhouse culture included. It will be generally accessible through server-database provision and scientific publications.

• DFG Programme: Research Grants