Fraunhofer research scientists are optimizing the manufacturing of 3D tissue models using a modular sensor platform and AI. Their solution enables continuous quality control in bioreactors and opens up new perspectives for the biotech and pharmaceutical industries – for more than just animal-free drug testing.
Whether for drug screening or toxicity testing, stem cell-based 3D tissue models are key to biomedical research. However, producing cell aggregates in bioreactors is highly complex and cost-intensive. Until now, quality has only been checked at the end of the process. This entails a high risk of lost time and material.
Now, at the Fraunhofer Center for Sensor Intelligence ZSI, research scientists from the Fraunhofer Institutes for Biomedical Engineering IBMT and for Nondestructive Testing IZFP have jointly developed an intelligent sensor system that continuously monitors and optimizes the entire manufacturing process. Fraunhofer IBMT is contributing its expertise in developing optimized, standardized cell culture methods and innovative model systems for stem cell research. The Multimodal Autonomous Sensor Platform (MAUS) from Fraunhofer IZFP implements the groundbreaking collection and transmission of sensor data and of the information derived from it.
In the current SpheroSense research project, the Fraunhofer IBMT team installed two levels of sensors in a commercial bioreactor for 3D cell spheroids, i.e., spherical cell models that can mimic tissue. The cell environment is monitored by electrochemical sensors for glucose, lactate, dissolved oxygen, pH and temperature. Cameras provide for optical characterization of the cell aggregates. They continuously analyze the size, number, and quality of the spheroids in real time. This results in a growth curve, enabling conclusions to be drawn on optimal conditions.
The research scientists rely on two camera paths for visual quality control. An in-situ camera focuses on the rotating reactor tubes and detects spheroids in suspension, while another camera detects them as they are drawn through a microfluidic channel. The integrated AI recognizes and classifies cell aggregates within milliseconds. The MAUS edge AI platform bundles all sensor data, evaluates it directly on site and transfers the results to a database.
With their solution, the experts at Fraunhofer are integrating real-time quality control in production. This quality control is objective, automatic and cost-effective, as the media for stem cells are already expensive and the cultivation process is time-consuming. Project manager Thomas Velten from Fraunhofer IBMT explains: “We replace subjective assessment with objective AI-based evaluation, ongoing monitoring and potential adjustment of process parameters. Last but not least, we ensure connectivity with modern process control systems in the pharmaceutical and biotech sector. Our solution corresponds to modern concepts such as Process Analytical Technology (PAT) and Continuous Process Verification as required by the FDA and EMA.” System operation is label-free, meaning that no staining is necessary and the spheroids remain undamaged.
A further benefit: The solution works without any cloud connection and thus fulfills the most stringent data security requirements. A cloud connection is possible on request, but the default configuration keeps sensitive process data with the customer.
The system can be adapted to different application scenarios ranging from stem cell research to industrial cell production. “Our MAUS platform enables rapid prototyping for all monitoring applications. With its modular design, it can be adapted flexibly and expanded quickly and easily to meet any needs,” explains Christoph Weingard, sensor intelligence and microelectronics developer at Fraunhofer IZFP. “For example, we implemented radio modules with Bluetooth and near-field communication (NFC) in SpheroSense in order to read the sensors inside the tubes without any additional cables. The appropriate plug-in module is simply attached, and monitoring starts.”
The entire system is being used successfully in production at the Fraunhofer Center for Sensor Intelligence ZSI, with direct feedback from the colleagues who produce 3D cell spheroids and use them in further projects.
The specialists will present their solution at Analytica in Munich from March 24 to 27, 2026. Visitors can learn about the entire system, including sensors, cameras and real-time data analysis, at the world's leading trade show for laboratory equipment, analysis and biotechnology.
Fraunhofer-Gesellschaft