In this course, a circuit board and software will be developed for an existing mechanical drink mixing machine. The machine has various actuators. We expect the group to split up so that everyone can do what they are comfortable with. Ideally, the following tasks will be done by someone in the group:

• Designing of a PCB for IO (motor drivers, sensors)
• Designing of a PCB for a controlling MCU and connection to the IO PCB
• Programming of the controlling MCU
• Create an idea for the user interface on the Raspberry PI (optional)
• Programming the RPI and the user interface for the existing touch screen (optional)
• Connecting the RPI and the controlling MCU (optional)

Organisation:

The kickoff meeting for this project will take place on 10.10.2023 at 11:00-11:30 in room BC 013. Attendance at the kick-off meeting is mandatory for participation in this project.

Medical exoskeletons are an important topic e.g. for medical rehabilitation of stroke patients. One challenge is to identify what kind of movement a user wants to do, e.g. lifting an arm. This can be supported using biosignal analysis such as EEG (i.e., neuron activity in the brain) or EMG (i.e., muscle activity). In this project, we will explore how embedded machine learning can be used to implement better exoskeletons through such biosignal analysis. 

• Experience with C/C++ programming

• Prior knowledge of time series analysis and / or digital signal processing
• Experience with designing / training machine learning solutions
• Experience with embedded systems development
• Python or VHDL programming experience

The kickoff meeting for this project will be on the 10. Oct. 2023 at 10:00 am in room BC 013. Attendance in the kickoff meeting is mandatory for participation in this project.

As part of this practical project, students are to optimise the methods for neurosignal processing of extracellular action potentials, which are recorded using microelectrode arrays.

A Python framework already exists for this purpose, which is to be expanded with additional functions for AI-based methods, additional functions for synthetic data generation and for neuronal data analysis (incl. representation). For this purpose, the classification tasks are to be validated using deep learning techniques and with neuromorphic networks via spiking neural networks. In addition, there is the possibility to further optimise the internal hardware setup for the playback of neurosignals from digital source to analogue signal.

Possible subjects:

• Data set creation with MEArec
• Data analysis with MEAnalyzer
• Preparation of data sets for autoencoder training (Dense NN, CNN, Denoising, ...)
• Use of the elastic AI.Creator to generate neural networks
• Use of neuromorphic networks
• Preparing the Neurosignal Player (C code for playing the signals)

Organisation:

The kickoff meeting for this project will take place on 10.10.2023 at 10:30-11:00 in room BC 013. Attendance at the kick-off meeting is mandatory for participation in this project.