With the continuous spread of cheap 3D manufacturing technologies, especially 3D printers, those systems can be equipped with cheap sensors providing mechanical information with a high sample rate. Only recording this data is simply possible, but without interpretation these huge amounts of sample points are of not much use.
Modern high dimensional data analysis techniques provide supervised and unsupervised learning methods to be applied to recorded 3D printer acceleration data. The result should be a 3 dimensional visualization of the recorded data and print object as well an automated analysis of print quality.
This work can be done in either English or German, preferably as a guided research (Forschungspraxis).
Mandatory prerequisites are:
- good C/C++ skills
- experience in 3D printing (you have built you own 3D printer)
- electronics/soldering
- basic modeling of physical systems
For further information, please contact Johannes Feldmaier and Dominik Meyer