A Wearable Exoskeletal Rehabilitation Robot for Interactive Therapy - Biomechatronics in Medicine and Healthcare - page 13

• Exoskeleton machines - Exoskeleton robots, unlike the operational ma-

chines, follow the anatomy of subject's UE; the device is attached to the
subject's UE at different point, the humerus and the forearm etc. The
number of DOFs of the robot is at least equal to the number of joints in
the subject's UE that are targeted for robot-assisted therapy. The design of
these devices is more involved because of the need to align the subject's
joint rotation axis with the rotational axis of the robot's joint mechanism;
this is a non-trivial problem for complex human joints such as the shoulder
joint (Nef and Riener (2008)). Additionally, the need to fit subjects of differ-
ent body types compounds the complexity of the design. Exoskeletons, in
general, are specific to a particular side of the arm i.e. a device designed for
therightarm,inmostcases,cannotbeusedfortheleftarmandviceversa.
The major advantage of exoskeletons, however, is that they allow for the
possibility to measure and control individual joint movements, which is
not possible with operational machines. Also, the static arm posture of
the patient's arm is fully determined with an exoskeleton, but not with
an operational machine. Exoskeleton might be more suitable for subjects
with low levels of residual motor control, as the joint synergies of these
subjects are usually altered by stroke (Micera et al. (2005a)).

The second level of classification of these devices further divides the two
categories from the first level into two sub-categories, class I and class II. The
distinction between these two sub-categories is based on the engineering de-
tails/complexity of the robotic devices in the operational and exoskeleton cat-
egories. Class I devices are characterized by low mechanical impedance, high
backdrivability, use of advanced controllers such as impedance controllers, ability
to generate finely controlled force fields and quantify mechanical properties of the
arm, and high cost. Class II devices are in general relatively simpler devices. They
have a simple mechanical structure, no backdrivability, some of these devices have
active friction and inertia compensation, and low cost.
Some of the existing robots for training the UE are the MIT MANUS (planar
module (Krebs et al. (1998)), wrist module (Krebs et al. (2007)) and hand module
(Masia et al. (2007))), Mirror Image Motion Enabler (MIME) (Burgar et al. (2000)),
Assisted Rehabilitation and Measurement (ARM) Guide (Reinkensmeyer et al.
(2000)), Arm In (Nef and Riener (2008)), Gentle/S (Loureiro et al. (2004)), Pneu-
WREX (Sanchez et al. (2005)), MEMOS (Micera et al. (2005b)), NeRebot (Rosati
et al. (2007)), HWARD (Takahashi et al. (2008)), and Hand Mentor TM (Koeneman
et al. (2004)). The details of these rehabiliation robots are summarized in Table 2.2 .
The details summarized in Table 2.2 are mainly concerned with the engineer-
ing aspects of the robotic devices that determine a device's range of motion, its
workspace, and the different movements supported the robot. In addition to this,
the other crucial aspect of rehabiliation robots is the role played by the robot
during therapy. This is termed as the “mode of therapy” in the rehabilitation
robotics literature. The mode of therapy is mainly determined by the hardware