**6. Conclusions**

This work collects the results of the research activity on wearable robotics carried out at the Mechatronics and Dynamic Modeling Laboratory of the Department of Industrial Engineering of the University of Florence during the years 2015–2018. The activity focused on the study of design strategies for hand exoskeletons, starting from the analysis of state-of-the-art solutions and leading to the development of new robotic devices. Hand exoskeletons mechanical design constituted the main subject of the research, motivated by the issue that people who have lost or injured hand skills are deeply compromised in the possibility of an independent and healthy life with a sensitive reduction in the quality of their lives themselves. Since robotic devices, such as hand exoskeletons, may represent an effective solution for the patient affected by hand opening disabilities, or other mobility impairments, the focus was particularly given to the development of a robotic solution provider of an aid during prolonged and high-intensity rehabilitation treatments. Such systems also contribute to reduce costs and burden for the therapists, as long as being an effective aid in the execution of ADLs.

The overall research activity aimed not only to present the design of the hand exoskeletons developed by the University of Florence, but also, and actually mainly, to propose new user-centered and patient-based strategies which can be adopted in the production of wearable systems. The optimization-based scaling strategy (which, starting from the users' own gestures, leads to a tailor-made device) along with the mechanical choices (which guarantee the outline of a lightweight and compact overall system) center the end-user on the design process and the proposed wearable systems are fully designed around him. In all the presented hand exoskeleton prototypes, the system kinematics plays the main role for the design of a tool as much wearable as possible, which guarantees a comfortable feeling even during the prolonged use.

Starting from the same kinematic architecture, three different hand exoskeletons have been developed heading for four features mandatory for the actual use of the device: the comfortable feeling for the user, the adaptability of the kinematic chain of the finger mechanism to the patient's hand, the safe use of the device, and the user-friendly activation system and control of the exoskeleton. Table 2 summarizes the main results achieved in the design of each prototype.


**Table 2.** Results achieved in the design of the three developed prototypes. Symbol ✓ indicates the achievement of an acceptable level for the feature reported on the first column of the table, while symbol ✕ points out an open up point that had to be revised in the next version.

However, achieving the aforementioned feature was particularly difficult when the constraints of the single DOF had to be maintained. The proposed solution resulted in a tight (due the 1 DOF per finger, cable-driven, single-motor system) device capable of precisely replicating the hand gestures. The satisfying results highlighted the goodness of the derived solutions, which may constitute a suitable alternative to existing hand exoskeletons. Nonetheless, there is still room for improvement: for instance, clinical trials exploiting the reached device may allow for the assessment of the impact the hand exoskeleton would have during real rehabilitation sessions. A testing phase, employing a real actuation system, represents the first opened up point, which could lead to the necessity of improving the mechanical architecture of the whole exoskeleton in order to increase its usability and efficacy. Also the employment of different materials, whose structural features may benefit the robustness of the device, inspires the possibility for further investigations [42,43]. The last, but not least, important issue which is worth being explored in the near future is the assessment of the pose of the hand depending on the task it is required to accomplish. As reported in [44], the control of hand posture involves a few postural synergies leading to reduce the number of degrees measured by the hands anatomy and independently on the grip taxonomies. In this sense, considering the synergies occurred in some specific gestures [45] in the motion analysis assessment (Section 4.1) would allow the exoskeleton to guide the fingers with more effectiveness.

All these issues, whose resolution constitutes a natural continuation of the research activity carried out thus far, will be subjected to further investigation in the very near future.

**Author Contributions:** Design & Development, N.S. and M.B. (Matteo Bianchi); Supervision A.R. and Y.V.; Validation, F.V. and M.B. (Massimo Bianchini); Conceptualization, L.G. and B.A.

**Funding:** This work has been supported by Don Carlo Gnocchi Foundation (Italy) and by the HOLD project (Hand exoskeleton system, for rehabilitation and activities Of daily Living, specifically Designed on the patient anatomy), funded by the University of Florence.

**Conflicts of Interest:** The authors declare no conflict of interest.
