Pedro Dias

Pedro Dias

Former student (completed: Aug 24, 2018)


Master's Thesis Project: Development of PI-ETPU based sensors for prosthesis control

3D printing is not a new concept, this said, the capabilities of this technology have not been stretched to its full potential. Currently, surface EMG dry sensors are based in metals such as Ag, since these are good conductors, easy to use and do not require an electrolyte to get acceptable measurements. Furthermore, past work has proved the efficacy of 3D printed black-carbon doped PI-ETPU, a conductive thermoplastic used in 3D printing for the development of EMG based sensors. 

The first objective of this assignment will consist of the improvement of these sensors. Parameters like, shape, size and thickness of layers will be changed in order to find the optimal construction parameters of the electrode. Another possibility that will be looked at will be the development of bipolar or tripolar electrodes as an alternative.

The development of the electrodes will go through a wide variety of stages. First there is a need to understand the new material, Pi-ETPU. For that, the first aim of this model is to find an electrical model of the material. Furthermore, investigation will be conducted in order to see if the printing procedure inflicts differences between printed electrodes. 

After the understanding of the material and how the created electrode works, the aim will be to investigate several types of electrodes arrays. We are looking for a disposition of these in the arm in a way to decrease cross-talk between electrodes while allowing for the highest possible distinction of degrees of freedom. This in the end will allow for the distinction between more movements. The disposition of the electors will also be incorporated in a band that will also be 3D-printed. The material of this band should be stretchable in order to easy adapt to the contour of the body. 

After the development of the armband with the electrodes, the next stage of this work will be to test what has been developed so far. To do this, lower/upper arm EMG signals will be gathered and processed. After the processing a linear regression method or a Linear discriminant analyst method will be implemented in order to allow for the online control of a prosthesis using this armband. The implementation of this work, will serve a s a proof-of-concept of generating online and proportional control of prostheses using 3D-printed doped thermoplastic materials such a Pi-ETPU, taking full advantages of the capability of 3D printing to create a cheaper, easy to produce, readily available way of generating online proportional control of upper limb prosthesis.