• Haptics
• Robotic Hands
• Variable Stiffness Actuator

This thesis focuses on novel algorithms and interfaces, arising from
inspection and comprehension of the human nature.
In the first part I deal with new mechanical designs and concepts
for building and controlling robotic hands. In particular I show how
the sensorimotor synergies of the human hand can be useful not only
for controlling but also for building robotic hands, suggesting novel
design paradigms.
Despite the synergy model is useful for designing and controlling
robotic hands, it is incomplete to explain the hand behavior during
grasp (both for humans and robots). To solve this problem, it is
needed to consider compliant articulations introducing the "soft synergies"
concept. Consequently, the compliance and the soft synergies
lead to the concepts of muscle and mechanical impedance. Thus, in
the second part of this thesis I present an observer for estimating the
time varying mechanical impedance of a Variable Stiffness Actuator
(VSA), i.e. a novel kind of actuator whose performances and capabilities
are very close to the human muscles.
Another important feature, both for human and robotic hands, is
the sense of touch. Indeed in the third and last part of this thesis, I
deal with the haptics and haptic interfaces. I show two new haptic
devices with their applications on humans. Moreover, I present a
tactile sensing algorithm toolbox for computing the contact point of
a robotic fingertip interacting with an object.