elmug4future
KI für energieeffiziente Sensorsysteme – Effiziente Überwachung von Maschinen und Anlagen
Sebastian Uziel
Abstract:
The presentation shows how neural networks have to be adapted to run on microcontrollers. Moving such algorithms from the cloud to the sensor reduces the energy consumption for the transmission and calculation of the data in the cloud. The procedure is illustrated by an example from predictive maintenance:
At IMMS, algorithms for embedded systems were designed, optimised and implemented in a demonstrator for AI-based fan monitoring using the machine learning approach TinyML: Motor bearings or fan blades can be directly monitored with the compact retrofittable solution. Defective parts are indicated directly via a red LED.
Related content
Project
SME Digital
Under the name ”Smart Sensor Systems Model Factory“ we are supporting enterprises in their gradual adoption of Industry 4.0 technology.
Press release,
Gemeinsam durch die Krise: Thüringer Netzwerke starten branchenübergreifende Initiative
Informationen, Förder- und Unterstützungsmöglichkeiten für KMU in Thüringen
Press release,
Reinhard Bütikofer zu Besuch im „Mittelstand 4.0-Kompetenzzentrum Ilmenau“
Lösungen zur Vernetzung und zu Energie- und Ressourceneffizienz an TU Ilmenau und IMMS vorgestellt
Press release,
Ramelow besucht „Mittelstand 4.0-Kompetenzzentrum Ilmenau“ am IMMS alias „Modellfabrik Migration“
Sommertour durch Digitalisierungsthemen – dritte Woche #ZukunftThüringen startet am IMMS
Press release,
Bundestagsabgeordneter Tankred Schipanski zu Besuch am IMMS
Am 12. Juli 2018 besuchte der für den Landkreis Gotha und den Ilm-Kreis gewählte Bundestagsabgeordnete Tankred Schipanski das IMMS. Zur…
This might also be interesting for you
Lead application
Adaptive edge AI systems for industrial application
We are researching solutions for adaptive edge AI systems to make AI possible on low-consumption embedded systems in industry and to network them in real time.
Service for R&D
Development of embedded systems
We develop embedded systems for you as complete solutions consisting of sensors and actuators, signal processing and communications technology as well as open-loop and closed-loop controls.
Core topic
Embedded AI
The numerous existing AI algorithms and methods for high-performance computing are unsuitable for embedded systems. We are researching to optimise AI algorithms and methods so that they can be used on embedded systems.
Research field
Smart distributed measurement and test systems
Integrated sensor ICs are the heart of sensor and measurement systems like wireless sensors, stationary or handheld devices. We are researching solutions for ever more powerful sensors with more intrinsic intelligence and task allocation in the network.
Contact
Contact
Dipl.-Hdl. Dipl.-Des. Beate Hövelmans
Head of Corporate Communications
beate.hoevelmans(at)imms.de+49 (0) 3677 874 93 13
Beate Hövelmans is responsible for the text and image editorial work on this website, for the social media presence of IMMS on LinkedIn and YouTube, the annual reports, for press and media relations with regional and specialist media and other communication formats. She provides texts, photographs and video material for your reporting on IMMS, arranges contacts for interviews and is the contact person for events.