IEEE Robotics & Automation Magazine - March 2015 - 98

hands due to a spinal
cord injury was able to
use the glove to grasp
objects of various shapes.

Wrist motion was used as
the control input for easy
and intuitive control.

A Wearable Robot
Without an
Exoskeleton
When people suffer paralysis of the hand and fingers in
the wake of injuries or diseases, such as spinal cord injury
[1], stroke [2], and cerebral palsy [3], they often cannot perform even the simplest activities of daily life. Tetraplegics [1]
are particularly disabled. A wearable hand robot that
restores basic hand and finger function would greatly
improve quality of life for people with hand motility problems. Such a hand robot should be natural looking and

Figure 1. The Exo-Glove allows the disabled to perform everyday
tasks, greatly improving their quality of life.

Actuation
Module
Fabric
ric Straps

Outer
Sheath
O
of Bowden
Cable

Flexor/
Extensor
Tendons

Tendon
Anchoring (TA)
Support

Figure 2. The Exo-Glove is easy to wear because the actuator can
be positioned remotely (for example, on the upper arm or on
the wheelchair).

IEEE ROBOTICS & AUTOMATION MAGAZINE

march 2015

simple to implement so that people feel psychologically
comfortable wearing it.
A number of wearable hand robots have been developed,
most of them consisting of serially connected rigid links and
joints that assist body motion by applying force to the body
part to which they are attached [4]-[8]. The positioning of
these links and joints differs with the body part to which the
robot is attached. The frames of hard exoskeleton robots for
the legs or arms can be positioned along the sides of the
limbs, but the closeness of the fingers means that the frames
for a hand robot must be placed along the finger tops. To
match the axes of the finger and the exoskeleton, many wearable hand robots use linkages [4]-[7] and rack-and-pinion
mechanisms [8].
One way to overcome the limitations of a rigid wearable
hand robot is to dispense with frames and, instead, attach
actuators directly to the fingers. In this design, the function of
the rigid frame of a conventional robot is performed by the
finger bones, and actuation can be accomplished by mechanical tendons, air, or other soft actuators. In the literature, these
robots are referred to as soft wearable robots, soft exoskeletons,
exosuits [9], or exotendons [10]. Such robots do not have joint
alignment problems, and they are compact and lightweight
because of their simple structure and the materials used to
make them. Several soft wearable robots for the hand have
been developed [10]-[12], the majority of which use tendon
drives [10], [12].
The tendon drive system does, however, pose some design
challenges as follows.
● Tendons that are attached to the body without the use of an
exoskeleton apply shear forces to the attachment points. If
the tendons are attached by means of a soft material, the
shear forces can move the attachment point and obstruct
force transmission.
● In conventional tendon drive systems, pretension is necessary for good performance of the robot. However, in soft
wearable robots, pretension can cause discomfort or injury,
as well as hampering efficiency because of increased friction along the tendon path.
● Conventional adaptation mechanisms that enable objects
of differing shapes to be grasped by simple control use
multiple rigid mechanical components. Soft wearable
robots for the hand require a new adaptation mechanism
to accomplish the same task.
The Exo-Glove [13] employs a soft tendon routing system
inspired by the human musculoskeletal system (see Figures 1
and 2). The developed force transmission components are
firmly attached to the body even though they consist of fabrics, with the exception of a thin anchor. All the elements of
the routing system, including the actuator, are designed for
operation without pretension, resulting in improved safety,
comfort in use, and system efficiency. We developed a new
adaptation mechanism for the soft tendon routing system by
modifying the conventional differential mechanism. This
reduced the complexity of the system and enhanced the handling of various objects. We verified the robot function

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - March 2015