• Haptic Human-Robot Interaction
• HPR devices
• Human Hand Synergies Framework
• Motor synergies
• Sensory synergies
• Sensory-Motor synergies
• Touch and proprioception integration

From a biological point of view, a sign of the main role of the hand in daily life activity can be found in its marvelous evolution, nowadays not yet surpassed by any artificial hand. The human hand is characterized of a complex biomechanical structure, indeed a complete description of the hand should consider more than 20 Degrees of Freedom (DoFs). Therefore, it is not surprising if many researchers have spent their efforts to investigate the human hand as cognitive organ and end-effector in object grasping and manipulation. Despite the aforementioned complexity in the description of the hand, there are some biomechanical constraints limiting the number of DoFs controlled independently by the Central Nervous System (CNS). We can look to these limitations as "enabling constraints" to the hand sensory-motor organization. The sensory-motor system components subject to such biomechanical constrained arrangement work together in a synergistic fashion (from the Greek term "synergia" meaning "working together"). A synergy can be defined as a collection of independent DoFs that behave as a single functional unit in a way that preserves the function integrity of the collection. Then, it is possible to define a synergistic control as a method to control a multi-DoFs system within a lower dimensional space than the available number of dimensions. Such dimensionality reduction based control of the sensory-motor system is also extensively supported from a neural point of view, which suggests the synergies idea as a natural solution to describe the organization of the nervous system. Therefore, it is possible to define the concept of human hand synergies framework as the overall of motor (muscle) synergies, sensory synergies, and sensory-motor synergies. If we focus on the motor output generates the movement, the synergies can be defined as covariation patterns of the angular excursions of many joints, i.e. motor synergies. Similar dimensionality shrinking can be also found in the sensory domain, where the high amount of dynamic sensory inputs (e.g., in touch from the mechanoreceptors) are organized into a low-dimensional set of manageable cues representations of the external environment, i.e. sensory synergies. Moreover, whenever we interact to the external world the CNS integrates the manifold input signals from the musculoskeletal system and the mechanoreceptors of the skin to produce a reliable and coherent estimation of our hands position, orientation, motion, contact characteristics and touched object properties, i.e. sensory-motor synergies.
For the purpose of my Ph.D., I focused on the threefold aspect of human hand synergies framework: motor synergies, sensory synergies, and sensory-motor synergies. In Chapter 2, I combine the theoretical foundations of the optimal design of hand pose reconstruction (HPR) devices based on hand synergies information with textile goniometers based on knitted piezoresistive fabrics technology, to develop an under-sensed glove for measuring hand kinematics. I use only five sensors optimally placed on the hand and complete HPR leveraging upon synergistic information. The reconstructions I obtained from five different subjects were used to implement an unsupervised method for the recognition of eight functional grasps, showing a high degree of accuracy and robustness. Then, I present an integrated sensing glove that combines two of the most visionary wearable sensing technologies to provide both hand posture sensing and tactile pressure sensing in a unique, lightweight, and stretchable device. I report on the real-time software integration of both modalities, and a qualitative evaluation experiment analyzing hand postures and force patterns during grasping. Moreover, I discuss the performance enhancement of an HPR technique based on a commercial RGBD camera data, which is affected by self-occlusions, leveraging upon postural synergies information. More specifically, I ignore joint angle values estimated with low confidence through a vision-based HPR technique and fuse synergistic information with such incomplete measures. Finally, I present a real-time software integration for augmented teleoperation where wearable hand/arm pose undersensing and haptic feedback devices are combined with teleimpedance techniques for the simplified yet effective position and stiffness control of a synergy-inspired robotic manipulator. In this work, I tested the usefulness of the HPR device discussed in Chapter 2 during a drilling task as an illustrative example of a dynamic task with environmental constraints. In Chapter 3, I present two solutions to extend Intrinsic Tactile Sensing to deformable surfaces, relying on force-deformation characteristics of the surface under exploration. I have tested both solutions using ellipsoid silicone specimens, with different softness levels and indented along different directions. A combination of two methods, using one to produce the initial guess for the other, turns out to be very effective. Indeed, in the validation this solution showed convergence under 1 ms, attaining errors lower than 1 mm. The proposed approaches were implemented in a real-time toolbox, integrating both solutions. The realized software provides both geometrical and dynamic contact information that can be used in soft-robotic sensing. In Chapter 4, I test the hypothesis that touch provides auxiliary proprioceptive feedback for guiding actions. I use a well-established perceptual phenomenon to dissociate the estimates of reaching direction from touch and musculoskeletal proprioception. Participants slid their fingertip on a ridged plate to move towards a target without any visual feedback on hand location. Tactile motion estimates were biased by ridge orientation, inducing a systematic deviation in hand trajectories in accordance with the formulated hypothesis. In particular, for such work I developed the hardware and software necessary to perform the experiments, then I collected the subjects' data to perform a statistical analysis based on the Generalized Linear Mixed Models. Furthermore, I developed an ideal observer model based on Kalman filtering, and I compared its prediction to the empirical findings. I found that cutaneous information systematically biases reaching movements and the results are in line with the prediction of the Kalman filter, thereby demonstrating that cutaneous touch is indeed an auxiliary cue for proprioception.