www.medical-devices.tech
03
'26
Written on Modified on
Fraunhofer Deploys AI Sensor Platform for Bioreactor Quality Control
The system enables continuous, real-time quality assurance of 3D cell production in bioreactors for biotech and pharmaceutical manufacturing.
www.fraunhofer.de
© Fraunhofer IBMT, Bernd Müller: AI-supported monitoring of 3D cell aggregate growth in suspension-based bioreactors.
Fraunhofer research teams have developed an intelligent, modular sensor platform that enables continuous quality control during the production of stem cell-based 3D tissue models in bioreactors. The solution replaces end-of-process inspection with real-time monitoring, reducing material losses and improving process reliability in biomedical and pharmaceutical applications.
The system was developed at the Fraunhofer Center for Sensor Intelligence ZSI through collaboration between the Fraunhofer Institute for Biomedical Engineering IBMT and the Fraunhofer Institute for Nondestructive Testing IZFP. It combines standardized cell culture expertise with a multimodal sensor and edge AI platform to monitor both the biochemical environment and physical characteristics of cell aggregates throughout cultivation.
© Fraunhofer IBMT: Bioreactor with integrated sensors for online process control and optical characterization of the 3D cell aggregates produced.
Within the SpheroSense research project, the platform integrates electrochemical sensors measuring glucose, lactate, dissolved oxygen, pH, and temperature, alongside optical monitoring using dual camera paths. One camera captures spheroids directly within the rotating bioreactor tubes, while a second inspects spheroids passing through a microfluidic channel. These imaging streams are analyzed in real time by AI models that classify spheroid size, quantity, and quality within milliseconds.
All sensor and imaging data are processed locally using the Multimodal Autonomous Sensor Platform (MAUS), which performs on-site data fusion, AI evaluation, and database transfer without requiring a cloud connection. This edge-based architecture ensures data sovereignty and compliance with stringent security requirements, while remaining compatible with optional cloud integration if needed.
The system supports label-free operation, avoiding chemical staining and preserving cell integrity. By enabling objective, automated quality assessment and potential adjustment of process parameters during production, the platform aligns with Process Analytical Technology (PAT) and Continuous Process Verification frameworks required by regulatory bodies such as the FDA and EMA.
© Fraunhofer IZFP: An NFC coil for MAUS manufactured using a 3D circuit board printer.
Designed with a modular architecture, the platform can be adapted for applications ranging from laboratory research to industrial cell production. Wireless sensor integration via Bluetooth and near-field communication allows monitoring inside sealed reactor tubes without additional cabling, supporting rapid deployment and scalability.
Fraunhofer researchers report that the system is already in productive use at the ZSI facility, supporting ongoing 3D tissue model manufacturing. The complete solution will be presented at Analytica 2026 in Munich from March 24–27, where visitors can view the integrated sensors, imaging systems, and real-time analytics in operation.
www.fraunhofer.com
Fraunhofer research teams have developed an intelligent, modular sensor platform that enables continuous quality control during the production of stem cell-based 3D tissue models in bioreactors. The solution replaces end-of-process inspection with real-time monitoring, reducing material losses and improving process reliability in biomedical and pharmaceutical applications.
The system was developed at the Fraunhofer Center for Sensor Intelligence ZSI through collaboration between the Fraunhofer Institute for Biomedical Engineering IBMT and the Fraunhofer Institute for Nondestructive Testing IZFP. It combines standardized cell culture expertise with a multimodal sensor and edge AI platform to monitor both the biochemical environment and physical characteristics of cell aggregates throughout cultivation.
© Fraunhofer IBMT: Bioreactor with integrated sensors for online process control and optical characterization of the 3D cell aggregates produced.
Within the SpheroSense research project, the platform integrates electrochemical sensors measuring glucose, lactate, dissolved oxygen, pH, and temperature, alongside optical monitoring using dual camera paths. One camera captures spheroids directly within the rotating bioreactor tubes, while a second inspects spheroids passing through a microfluidic channel. These imaging streams are analyzed in real time by AI models that classify spheroid size, quantity, and quality within milliseconds.
All sensor and imaging data are processed locally using the Multimodal Autonomous Sensor Platform (MAUS), which performs on-site data fusion, AI evaluation, and database transfer without requiring a cloud connection. This edge-based architecture ensures data sovereignty and compliance with stringent security requirements, while remaining compatible with optional cloud integration if needed.
The system supports label-free operation, avoiding chemical staining and preserving cell integrity. By enabling objective, automated quality assessment and potential adjustment of process parameters during production, the platform aligns with Process Analytical Technology (PAT) and Continuous Process Verification frameworks required by regulatory bodies such as the FDA and EMA.
© Fraunhofer IZFP: An NFC coil for MAUS manufactured using a 3D circuit board printer.
Designed with a modular architecture, the platform can be adapted for applications ranging from laboratory research to industrial cell production. Wireless sensor integration via Bluetooth and near-field communication allows monitoring inside sealed reactor tubes without additional cabling, supporting rapid deployment and scalability.
Fraunhofer researchers report that the system is already in productive use at the ZSI facility, supporting ongoing 3D tissue model manufacturing. The complete solution will be presented at Analytica 2026 in Munich from March 24–27, where visitors can view the integrated sensors, imaging systems, and real-time analytics in operation.
www.fraunhofer.com
