Microtechnologies for compliant interfaces between manipulated objects, hand prostheses and the nervous system

Some factors in the field of upper limb prosthetics impede the usability of most advanced devices. The main limitations deal with two challenges, namely: i) the restoration of a bidirectional comunication channel with the Central Nervous System (CNS), which would allow the direct control of the prosthesis; and ii) the provision of a rich sensory feedback for providing the necessary information about the contact conditions to the low level control of the prosthesis, thus enabling dexterous manipulation tasks. This PhD thesis investigates novel solutions for overcoming the mentioned limitations, for reliably interfacing the prosthetic hand and the human neural system as well as the former to the external environment. To this aim, aspects of neurophysiology, microtechnology, robotics and material science have been integrated. Starting from the study of neural interfaces state-of-the-art, the theoretical feasibility analysis of a novel bidirectional interface between upper limb prosthetic devices and the CNS, based on electromagnetic stimulation of the neurons and electrical recording of their responses, is presented. Results show that microprobes, integrating microcoils and planar electrodes, could represent a promising alternative to existing solutions, nowadays based on pure electrical means and whose performace, especially during stimulation, are prone to be affected by fibrotic reactions. As it regards exteroceptive interfaces, the focus is on tactile sensors, based on thin film technologies, easily integrable on artificial skins and with a low computational cost. Capacitive pressure sensor arrays and thermal slip sensors, which, from a comparative analysis of the state-of-the-art, appear among the most promising solutions for the specific application, have been developed and tested. Finally, friction enhancing mechanisms have been investigated for improving the grasp stability, especially when wet objects are contacted, without increasing the cognitive burden of the user. In particular, a novel mechanism, named electrowet adhesion, based on electrowetting and wet adhesion techniques, has been investigated and its feasibility has been proved by means of tests performed using an ad hoc engineered surface. Tests show the capability of this device of actively modulating the adhesion force at the interface with wet substrates by acting on external electric fields. This surface can be further miniaturized and developed on a flexible substrate, thus being technologicaly compatible with the tactile sensors, with which it would constitue an artificial skin. Special attention has been paid on choosing soft materials (flexible or stretchable) for all the technological solutions presented, in order to guarantee compliance, to avoid mechanical mismatches at the interface and to allow the foldability of the developed devices around complex surfaces. The work is organized in seven Chapters, starting with the analysis of the state-of-the-art. The main outcomes are analyzed and discussed in each Chapter, wherease conclusions and the impact of the work presented are discussed in Chapter 8.