Together with Fraunhofer UMSICHT and Osnabrück University of Applied Sciences, Computomics received a grant this year for the project LightSaver AI. Computomics is responsible for the machine learning (AI) part of the project.
The central goal of LightSaverAI is to create the basis for an intelligent production system for indoor farms in urban areas. This system measures the chlorophyll fluorescence (ChlFl) as a measure of the photosynthesis rate as well as various environmental parameters online and evaluates them using AI approaches. In this way, the current light requirement of a plant is analyzed and an LED exposure module is adjusted via a controller so that the plant continuously receives the required exposure, which depends on the growth phase and environmental parameters.
The need for such systems is of particular interest in the global indoor farming sector - to produce high quality plants in a resource-efficient manner. The production system is able to optimally exploit the potential of LED technology by optimizing irradiance and spectral composition both temporally and spatially. This allows the maximum photosynthetic rate to be achieved with minimal energy consumption. It also offers the possibility of adapting other environmental parameters to it in the long term. To achieve this goal, the model plant lettuce (Lactuca sativa var. capitata) is exposed to different light scenarios. Fluorescence signals are detected and information about current processes in the plant is obtained.
In LightSaverAI, this application scenario is transferred to a laboratory-scale experimental setup for indoor farms. At the end of the project, the general suitability of the technologies developed here will be proven on a laboratory scale.
Computomics will evaluate the the relationship between the parameters of the PAM (Pulse-Amplitude-Modulation) fluorometer and the influencing factors such as temperature, CO2 or nutrient supply. Using AI, Computomics will find the optimal ranges of these parameters, and evaluate plant performances under different conditions.
Timeframe of the project: April 22 - March 2025
Funding amount: 27.497,96 €
Project homepage Hochschule Osnabrück (German language)
Dr. Rupashree Dass
Machine Learning Scientist
Photo: Fresh organic vegetable grown using aquaponic or hydroponic farming