![\"\"](\"https:\/\/www.esi.fau.de\/files\/2025\/10\/IMG_2257-819x1024.jpg\")
On Friday, October 10, 2025, our 19th Embedded Talk on the topic of \u201cDeep Learning on Narrow Resources\u201d took place at Friedrich-Alexander University Erlangen-Nuremberg. The event, which was attended by around 60 participants, provided a platform for experts and researchers from the industry to discuss current developments and challenges in the field of machine learning methods for systems with limited resources. The event began with a welcome address by Prof. Dr. J\u00fcrgen Teich, spokesperson for the FAU Research Center ESI. Prof. Dr. Jian-Jia Chen, TU Dortmund University\/Lamarr Institute for Machine Learning and Artificial Intelligence, then gave a keynote lecture on \u201cResource-Aware Machine Learning for Cyber-Physical Systems.\u201d In his presentation, he highlighted the challenges and opportunities of machine learning methods in cyber-physical systems. Mark Deutel, researcher at Fraunhofer IIS and FAU Erlangen-Nuremberg, then presented his research findings on \u201cOn-Device Training of Deep Neural Networks on Cortex-M Microcontrollers.\u201d He demonstrated how deep neural networks can be trained on Cortex-M microcontrollers.<\/p>\n\n\n\n
After a short coffee break, Jan Seyler from Festo SE & Co. KG gave a presentation on \u201cReal-World Challenges of Deploying Embedded AI.\u201d He highlighted the challenges that arise in practice when implementing artificial intelligence in embedded systems. The event concluded with a panel discussion with all speakers and a poster session, where participants had the opportunity to exchange ideas with the speakers and other attendees. The event was a great success and provided a valuable platform for the exchange of ideas and experiences in the field of machine learning methods for cyber-physical and embedded systems.<\/p>\n\n\n\n