At AlgaEurope 2025, GreenSight presented an AI-enabled sensor platform for real-time monitoring and optimisation of photobioreactors. The work demonstrates continuous, non-invasive process control under industrial conditions and was awarded the Silver Poster Presentation Award.
Advanced Sensors and AI-Driven Data Analysis for continuous Monitoring and Optimisation of Photobioreactors
Presenter: Dr. Franz-Josef Schmitt
Schmitt, F.J.1,2
Löber, J. 1, Deininger, F. 1, Rieder, F. 1, Lentz, L. 1, Hackenberg, E. 1, Ranga, A. 1, Rosenlöcher, L. 1, Vitali, M. 2, Grehn, M. 2, Duci, A. 2, McGee, D. 3, MacDonald, J. 3
1Martin-Luther-University Halle-Wittenberg, Germany
2Sensoik Technologies GmbH, Germany
3AlgaeCytes, United Kingdom
ABSTRACT
Efficient, large-scale algal biomass production relies on the precise monitoring and control of critical process parameters. We present a fully integrated monitoring approach that combines distributed optical sensors, advanced data acquisition hardware, and AI-assisted analytics to enable a non-invasive real-time process optimization in photobioreactor (PBR) systems.
Our sensor platform employs low-cost, high-stability optical detectors capable of continuous optical density (OD) measurements, determination of biomass, Chl a and carotenoid content in 18 wavelength sections across the visible spectrum. Time resolved signal acquisition based on time-of-flight allows for the separation of reflections and signal distortions and probes for humidity, CO₂ and temperature complement the acquired data. The sensors are attached to the surface of closed tubular glass PBRs. Data are collected via WiFi or wired interfaces and streamed into a cloud-based data infrastructure.
Recent installations at the AlgaeCytes facility (UK) demonstrate the system’s ability to operate continuously under industrial conditions. Over several months, the sensor network captured high-resolution datasets that revealed momentary growth speed and reproducible circadian oscillations in AlgaeCytes Eustigmatophyte, strain ALG01. Machine learning models identified the green/blue (G/B) absorption gradient as a primary indicator of physiological state, and infrared absorption as the parameter most strongly correlated with OD and growth rate. The findings enable targeted adjustments of spectral illumination regimes, nutrient dosing, and aeration strategies. Our results underline the importance of continuous, multi-parameter monitoring for both scientific understanding and industrial optimization of algal cultivation. The combination of dense sensor networks, cloud-based analytics, and interpretable ML provides a pathway to predictive, self-optimizing algal production systems. Beyond production efficiency, the technology offers potential for environmental monitoring applications, supporting the broader concept of an “Internet of Nature” through scalable, autonomous sensing solutions.