InSeKT - Development of intelligent sensor-edge technologies

With the InSeKT project - “Development of intelligent sensor edge technologies” - a new, interdisciplinary research project has been launched in Brandenburg that aims to revolutionize the integration of artificial intelligence (AI) at the so-called edges of IT networks. The aim is to anchor AI where the data is generated: directly at the sensors. Applications in industrial electronics, medical technology and environmental monitoring in particular will benefit from this.

Decentralized intelligence for greater efficiency and safety

Currently, large volumes of data are usually processed centrally via cloud servers - a process that costs both time and data protection. InSeKT takes a different approach: AI-based data processing is to take place directly at the source of the data, i.e. at the sensor itself. This not only increases the reaction speed, but also significantly reduces the risk of data leaks. Real-time capability and data protection go hand in hand here.

Strong together: expertise from science and research

The project combines the expertise of the Technical University of Applied Sciences Wildau, the Leibniz Institute for Innovative Microelectronics (IHP) and the Fraunhofer Institute for Photonic Microsystems IPMS. Together, the partners are developing new hardware and software solutions as well as advanced sensor technologies to significantly improve the performance of edge AI systems.

Fraunhofer IPMS: Miniaturized sensor technology for intelligent systems

The Integrated Silicon Systems branch of Fraunhofer IPMS in Cottbus contributes its extensive expertise in miniaturized sensor technology and the integration of electronic components. The focus is on three central fields of development:

Gas analysis with miniaturized ion mobility spectrometers (IMS): based on a FAIMS approach with flexible electrode spacing, an IMS demonstrator is being developed that can detect the smallest gas concentrations in the air - compact, sensitive and directly at the point of use.

Photodetectors in the near-infrared range: The aim is the data-supported optimization of Al-TiN-Si Schottky detectors for cost-effective and scalable applications, for example in material analysis and recycling.

Ultrasonic transducers for high-resolution imaging: Capacitive micromechanical ultrasonic transducers (CMUTs) are being adapted and further developed to enable precise image analysis in real time - from gesture recognition to non-contact blood glucose measurement.

From sensor technology to intelligent data processing

The sensor data obtained is used by TH Wildau and Leibniz IHP as the basis for training AI systems that are specially designed for use on network edges. This creates an end-to-end development process - from the signal source to intelligent evaluation - with high industrial relevance.