Source: Silicon Semiconductor
△ Scan to register for the conference
A newly launched interdisciplinary research project involving multiple universities and research institutions in Brandenburg is developing new technological approaches to better and more effectively integrate artificial intelligence at the edge of IT networks (the so-called “edge”).
These developments are expected to play a significant role in the future, especially in applications within industrial electronics, medical technology, and environmental monitoring. The Fraunhofer Institute for Photonic Microsystems (Fraunhofer IPMS) will contribute its expertise in micro-sensor structures and the integration of electronic components.
In the project named “InSeKT” (German: Development of Intelligent Sensor Edge Technology), the University of Applied Sciences Wildau, the Leibniz Institute for High-Performance Microelectronics (IHP), and the Fraunhofer Institute for Photonic Microsystems (Fraunhofer IPMS) are developing new hardware, software, and sensor solutions to better utilize artificial intelligence (AI) at the edge of IT networks. AI needs to process large amounts of data as quickly as possible. The goal of the project is to perform complex computations directly at the data generation point, such as the sensor itself.
Currently, AI data processing is typically handled through central cloud computing solutions. Data is computed on central servers, which means that large amounts of data need to be transmitted over long distances. This can pose risks of data leakage and create opportunities for unauthorized third-party attacks. Distributed data processing not only enhances data protection capabilities but also enables real-time capabilities by avoiding long-distance data transmission.
The project focuses on addressing key factors for market acceptance: developing system integration technologies, reducing costs, improving reliability, and enhancing miniaturization. The project is led by an interdisciplinary team from multiple institutions and various professional disciplines.
Advanced Sensor Technologies Addressing Material and Integration Issues
The “Integrated Silicon Systems” division of the Fraunhofer IPMS, located in Cottbus, is working on extending the functionality and integration of existing MEMS sensors for use in edge AI applications. By integrating signal processing capabilities directly into the sensors, data can be collected right at the point of data generation. The aim is to enhance the adaptability of sensors to different application scenarios without the need to replace the underlying hardware.
Fraunhofer IPMS’s initial core development area was gas analysis using Ion Mobility Spectrometry (IMS). IMS can directly detect ionizable analytes in the air, even at very low concentrations. Existing methods lack sufficient miniaturization. The first IMS demonstrator is based on the FAIMS (Field Asymmetric Waveform Ion Mobility Spectrometry) method, featuring flexible electrode spacing, which is expected to overcome this barrier.
Additionally, efforts are underway to achieve data-supported evaluation of photodetectors in the near-infrared wavelength range. For example, these photodetectors can be used for material analysis and recycling, and even for penetrating packaging for analysis. The focus is on improving the sensitivity of Al-TiN-Si-Schottky detector components with a cylindrical pyramid structure and enhancing scalability by using lower-cost materials.
The third area involves the adaptive use of capacitive micromachined ultrasonic transducers (CMUT) to improve imaging. Thanks to their size and capacitive working principle, CMUTs are highly sensitive ultrasonic receivers. Implementing signal evaluation close to the sensor can accelerate imaging speed. Dr. Sebastian Meyer, head of the “Integrated Silicon Systems” division at Fraunhofer IPMS, explains: “In the future, we expect to use ultrasonic signals inspired by bat echolocation to analyze hand movements very precisely and to measure blood glucose levels using ultrasound.”
Then, the University of Applied Sciences Wildau and the Leibniz Institute for High-Performance Microelectronics will use the generated sensor data to train edge AI systems for fast and accurate data processing. The project’s results will further advance the development of sensor systems towards intelligence and compactness.
Submissions Welcome
The Semiconductor Chip Technology media platform is open for submissions from readers!
There are no restrictions on topics or genres. If your company has the latest product releases,
cutting-edge technological advancements, or favorable news regarding financing and mass production,
you are welcome to submit via the following email~
Recommended by Industry Professional Media
Clickto read the original article
