*2.1. General Considerations*

Taking into consideration a multitude of existing developed robotic hand exoskeletons, there are a few design factors that need to be taken into account. There are a number of structural designs that can be implemented in order to transmit motion to human fingers. These designs can be divided into three types, one of which is based on rigid structures that implement classical mechanical actuation. The second one is based on soft robotics [20,21], a new field of robotics that uses various types of soft actuators [22–25], and the third type, is based on hybrid actuation, which implements a combination of rigid, soft, and compliant [26] actuation systems [24,27–29]. In this paper, the structural type used belongs in the rigid structure category. This type of construction using a rigid exoskeleton implies that some parts must be customized for each wearer [30], in other words, a good design would permit interchangeable components to be easily swapped out and reconfigured. One alternative for customizing for each wearer is to use standard sizes similar to clothing and footwear. Having a kit of standardized sizes of elements for the fingers and palm region can speed up the process of configuring the exoskeleton for the wearer.

The device's level of complexity is increased due to the asymmetric distribution of the fingers since each finger has a unique set of anthropometric dimensions. Not only does each finger have unique anthropometric dimensions, but also the phalanges have unique anthropometric dimensions for each of the subjects, following a Fibonacci dimensional ratio [31–33]. As seen in Figure 1, a visual rendering of the human hand mechanical model is presented, where the asymmetric nature of the human hand can be observed. The phalange area represents the finger segments that comprise the thumb, index, middle, ring, and little finger. The blue segments represent the phalanges of the fingers, while the red segments represent the joints of the fingers, where the cylindrical joints have one degree of mobility (DOM) (comprised of 1 rotation) and the universal joint has two DOM (comprised of two rotations).

Considering a variety of mechanical designs present in this field of research [34], there are a few distinct constructive types that stand out. A solution with direct matching of the finger joint centers (DMFJC) was developed by Chiri et al. [6] which can be observed in Figure 2a. This type of design is a good type of construction due to its behavioral similarities to the biomechanics of the human hand. A limitation for this design is that the direct matching of the joints is possible on the Distal Interphalangeal (DIP) and Proximal Interphalangeal (PIP) joints, but mechanically it cannot be implemented on the Metacarpophalangeal (MCP) joint due to the hand anatomy. For the MCP joint, there are a number of other solutions that rely on more complex mechanisms to obtain an actuated or

underactuated movement. The example shown in Figure 2a is an implementation of an underactuated MCP joint using a rotation translation mechanism controlled by a single actuator.

**Figure 1.** Proximal Interphalangeal (PIP), Distal Interphalangeal (DIP) and Metacarpophalangeal (MCP) connected to Metacarpals I joint have one axis of rotation, while the MCPs connected to Metacarpals II-V and Carpo-Metacarpal (CMC) joints have two axes of rotation.

**Figure 2.** (**a**) Direct matching of the finger joint centers by Chiri et al. (**b**) Linkages for remote center of rotation (Shields et al.). (**c**) Underactuated redundant linkage (Wege et al.).

A significant issue that is critical to an exoskeleton mechanical structure is the orthotic shell. One of the most crucial dimensional characteristics is the thickness in the lateral areas of the fingers that is coaxial to the finger joints. This dimension is highly dependent on the wearer and the hand anthropometrics. In other words, a wearer that has thinner and longer fingers will have more space between the fingers, which will permit the mounting of an exoskeleton. In comparison, a wearer that has thicker and shorter fingers will have less space between the fingers, thus resulting in the need for a thinner orthotic shell. The finger segments of the exoskeleton must be designed in such a way as to not hinder the natural motion of the finger or even produce discomfort.

Another type of construction, developed by Shields et al. [35], implements a more complex structure based on a mechanism with linkages for remote center of rotation (LRCR), as seen in Figure 2b. Although it seems like a bulky design, considering the large and complex mechanism, it has the considerable advantage of saving a lot of space between the fingers, an essential factor to take into account when designing the orthotic shell of the exoskeleton fingers.

The hand compliance is also an important factor, so to produce a flexion and extension motion of the biological fingers some designer such as Wege et al. [36] utilized an underactuated mechanism for all joints by implementing an underactuated redundant linkage (URL) structure as seen in Figure 2c. The size of the mechanism is considerably larger than the one implementing a mechanism with the

direct matching of the joints. While this design does not have the precise control of each phalange individually as the mechanism with LRCR structure [35], it has the advantages that it can be operated using fewer actuators and can provide a more natural movement of the wearer's hand due to its underactuated mechanism and the compliance of the biological hand.
