Page 339 - Handbook of Biomechatronics

332 Borna Ghannadi et al.

The interaction term in HRI for rehabilitation robots can be categorized into
two levels: physical and social. Mechanical upper extremity devices have
physical interactions, while noncontact upper extremity devices such as
“Microsoft Kinect” are considered to have social interaction. Most active
rehabilitation robots that provide different types of visual and auditory
feedback have physical-social interaction.
To study HRI in rehabilitation robotics, one should consider HRI
parameters: interaction arrangement, user interface, ability level, learning
and adaption, exterior design, and therapy time (Louie et al., 2017). In
robotic rehabilitation, interaction arrangement includes single-robot and
single-user, single-robot and multiple-user, and multiple-robot and
single-user; this arrangement can help to find the required mechanical
design. Robot user interface can be auditory, tactile, or visual; the type of
training can be distinguished by the user interface. Ability level indicates
the robot’s ability to perform a task, and this factor can have 10 levels varying
from no-assistance to independent control modes; these levels indicate the
form of rehabilitation. Regarding learning and adaption, both robot and user
should learn and adapt to each other’s performances, and this can motivate
the type of control scenario. Therapy time is each rehabilitation session’s
duration, and it is important to consider patient fatigue in control scenario
selection.
In addition to the HRI parameters, HRI metrics including user accep-
tance, user participation, user accompaniment, and user safety should be
considered. These metrics are used for postprocessing the results of a reha-
bilitation task with a robot. User acceptance indicates how much the user is
satisfied with the robot, user participation shows how long the user is
engaged in the robotic rehabilitation task, user accompaniment evaluates
how often the user is accompanying the robotic task (learning and adaption),
and the robot’s reliability is assessed by user safety (which is ensured by lim-
iting the robot’s ROM, kinetic variables, and motor torques).
To have a systematic and human-centered approach for optimal
mechanical design, these HRI metrics and parameters should be included
in the system requirements of the MBSE design process.

7 RECENT DEVELOPMENTS AND RESEARCH
OPPORTUNITIES

In previous sections, we categorized upper extremity rehabilitation
robots by mechanical design, type of training, form of rehabilitation, and