METHODOLOGY

Research Challenge 1 (RC1): Observation and Analysis
The observation aims at understanding how the human simplifies motion control, regulate contact forces and modulates hand-arm impedance to successfully perform the task. To this purpose a new protocol for experimentation on humans aimed at the observation of the role of vision and touch in grasping and manipulation will be developed. The observation will be performed both in a surgery site and in a varied and unstructured scenario typical of everyday life.

Research Challenge 2 (RC2): Planning and control
By exploiting the results of RC1, neural networks and reinforcement learning techniques based on the concept of sensory-motor synergies will be useful to realize bio-aware planning and control modules. Feedback available from force and vision sensors will be used by the control module. Artificial vision will be used not only for object recognition and visual servoing but the information will serve to integrate force and position measurements in order to properly activate motion patterns on the basis of task demands. 

Research Challenge 3 (RC3): Design
Another objective of MUSHA is to use the results obtained from human observation in order to develop and sensorize new robotic hands inspired by the versatility, motion capability, strength and dexterity of the human hand. The hand devices will have a high number of degrees of freedom in order to allow complex and human-like motions but, at the same time, will be controlled with a very low number of signals, thus with the use of few motors. The idea is to design prototypes that can serve for studies on both robotic and prosthetic hand-arm systems, i.e. controllable with motors or physiological signals, and on smart surgical hands with reduced dimension and reduced number of fingers but still having bio-inspired design and sensory apparatus.