Hands On II

Implementation of neural networks on resource-constrained microcontrollers for sensor applications

14.09.2022, 10:30 - 12:00 and 13:30 - 15:00
2 x 90 minutes
H 0.007


The workshop covers how machine learning with artificial neural networks (ANNs) can be used in sensor modules with low-cost, low-power microcontrollers. A sensor module with light sensors for the recognition of simple hand gestures with microcontroller ATMega4809 (6 kB RAM, 20 MHz) serves as use case. In addition to theoretical lecturing and practical demonstrations, the workshop includes a practical exercise on computers in which ANNs are trained and used in Python (with Keras / Tensorflow).

Due to the practical part the number of participants is limited to 12. Please bring your own laptop for the hands on session. Software does not need to be pre-installed, as the training and evaluation of the networks takes place in an online environment of the organizer.

Prof. Volker Turau
Institute of Telematics
Dr. Marcus Venzke
Institute of Telematics