Press Releases

The business unit Mobility of the Fraunhofer-Institute for Microelectronic Circuits and Systems IMS presents its latest development TimestampsAI for environment recognition at the »embedded world 2022« in Nuremberg from 21 to 23 June. TimestampsAI provides a faster and data-reduced solution for high-resolution LiDAR systems to capture complex scenes in 3D in any environment.

The business unit Health of the Fraunhofer-Institute for Microelectronic Circuits and Systems IMS presents its contactless vital parameter measurement as a demonstrator at the »embedded world 2022« in Nuremberg from 21 to 23 June. The contactless measurement of vital parameters offers a way to quantify health parameters quickly, accurately, hygienically and conveniently.

The Embedded Software and AI core competence of the Fraunhofer-Institute for Microelectronic Circuits and Systems IMS will present a customizable AI-based gesture recognition algorithm from June 21 to 23. The algorithm was developed using the Fraunhofer software AI framework AIfES, which supports highly efficient implementation of neural networks on embedded systems. In addition to applications in the field of device and machine operation, this gesture recognition method is also suitable for gaming applications, in the smart home as well as in the medical field.

The Smart Sensor Systems core competence and the unit Industry of the Fraunhofer-Institute for Microelectronic Circuits and Systems IMS will present the AIRISC-processor from June 21 to 23. It is able to evaluate sensor data up to 80 % faster.

Three business units and two core competences of the Fraunhofer-Institute for Microelectronic Circuits and Systems IMS present their applications at the »embedded world 2022«. Contactless vital parameter measurements, gesture recognition with the AIfES software framework, the RISC-V processor AIRISC and TimestampsAI for environment recognition: You can experience two demonstrators, an exhibit and the institute's latest LiDAR development at the Fraunhofer-Joint Booth in Hall 4-422 from June 21 to 23.

The RISC-V processor AIRISC, a 32-bit RISC (Reduced Instruction Set Computer) architecture, now features instruction set extensions and coprocessors for efficient neural network computation. For a medical application to detect atrial fibrillation in ECG data, we achieved a speed advantage of over 80% in evaluation with only 10% higher area and power requirements.

Mit dem AIRISC core stellt das Fraunhofer IMS seinen leistungsfähigen RISC-V Embedded-Prozessorkern für Sensorik-Aufgaben unter eine Open Source Lizenz, welche auch die Verwendung für kommerzielle Produkte erlaubt. Mit dem leistungsfähigen 32-Bit AIRISC-Kern lassen sich Produkte mit FPGAs schnell und kostengünstig entwickeln, die Umsetzung in einen (Mixed-Signal-)ASIC ist anschließend aufgrund der identischen Codebasis in Rekordzeit möglich und erfordert für den Basiskern keine kommerzielle Lizenz.

Since the beginning of the year an AIfES® update (version 2.1.1) is available. It can be downloaded directly from the Arduino IDE using the Arduino Library Manager or alternatively from GitHub. Here is an overview of the new features: - The new AIfES Express API - Q7 weight quantization - Arm® CMSIS integration enhanced by the NN functions.

The AIRISC enables efficient machine learning and AI in sensors, IoT devices and other embedded applications. The free RISC-V processor instruction set is ideally suited to implement custom extensions in a short time to provide optimal performance for specific applications. In combination with the AIfES software library developed by Fraunhofer IMS, the AIRISC processor family supports neural network inference and training directly on the embedded device.

The BMBF-funded "ARTEMIS" consortium will develop a small, portable detector under the leadership of Getemed AG. The core innovation is an artificial intelligence-enhanced semiconductor circuit in the ECG electronics that reliably identifies atrial fibrillation using machine learning.