Development of Walking Assistance Orthosis by Inducing Trunk
Rotation Using Leg Movement:
1
st
Report on Prototype and Feasibility Experiment
Harutaka Ooki
1
, Sho Yokota
1a
and Akihiro Matumoto
1b
, Daisuke Chugo
2c
,
Satoshi Muramatsu
3
and Hiroshi Hashimoto
4d
1
Dept. of Mechanical Engineering, Toyo University, Saitama, Japan
2
School of Engineering, Kwansei Gakuin University, Sanda, Japan
3
Dept. of Applied Computer Eng., Tokai University, Japan
4
Adv. Instisute of Industrial Tech., Shinagawa, Japan
Keywords: Walking, Assistance, Trunk, Human Powered.
Abstract: Walking is a whole-body movement including an upper body (trunk) and a lower body (pelvis and lower
limbs), which is called the Spinal Engine Theory. Furthermore, there is a finding that the stride length and
walking speed increases with the amount of trunk rotation. Based on this insight, there have been existing
study that promotes rotation of the upper body to assist in walking. However, motors are used to assist the
rotation of the upper body, which requires maintenance of the power supply and complicates the system.
Therefore, this research aims to develop an orthosis that promotes gait by increasing the amount of trunk
rotation without any active actuators. In particular, this paper reports on the basic experiment to confirm that
the prototype can apply assistive torque to the trunk when the leg is raised.
1 INTRODUCTION
In contemporary society with a low birth rate and an
aging population, the independence of the elderly has
become a critical challenge. Extending healthy life
expectancy is considered one solution to address this
issue (
Ministry of Health, Labour and Welfare
, n.d.).
According to a survey by the Ministry of Health,
Labour and Welfare Japan, it has been shown that the
physical functions that decline relatively early among
the daily activities of the elderly is the abilities to
move, such as walking and getting up from a seated
position (
Ministry of Health, Labour and Welfare
, n.d.).
Therefore, actively engaging in walking exercises in
daily life is effective as an early preventive measure
against activities of daily living disabilities among the
elderly. Furthermore, it is beneficial for physical
health and contributes to improvements in subjective
well-being, life satisfaction, and sense of purpose
(Murata, 2009). Due to these reasons, various
a
https://orcid.org/0000-0002-8507-5620
b
https://orcid.org/0000-0002-3004-7235
c
https://orcid.org/0000-0002-3884-3746
d
https://orcid.org/0000-0003-2416-8038
walking assistive devices have been researched and
developed. Examples include: the "Power Suit HAL,"
(
CYBERDYNE
, n.d.), which measures muscle activity
and assists lower limb movement during walking.
The "aLQ," (Imasen, n.d.) incorporates passive
walking and uses springs to assist lower limb
movement. "TPMAD" (Hashimoto, 2018), which
utilizes motors to assist trunk rotation and walking.
These devices aim to support walking and range from
practical applications to those with proven efficacy in
rehabilitation.
However, devices such as the "Smart Suit HAL"
and "TPMAD" require power sources, such as
batteries, due to the use of motors and sensors. In
contrast, the "aLQ" utilizes passive materials attached
to the legs without power sources, leveraging their
elastic force to assist leg movements. However, this
device does not consider upper body movements.
Since walking involves full-body activities, including
740
Ooki, H., Yokota, S., Matumoto, A., Chugo, D., Muramatsu, S. and Hashimoto, H.
Development of Walking Assistance Orthosis by Inducing Trunk Rotation Using Leg Movement: 1 st Report on Prototype and Feasibility Experiment.
DOI: 10.5220/0012235200003543
In Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2023) - Volume 1, pages 740-746
ISBN: 978-989-758-670-5; ISSN: 2184-2809
Copyright © 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
the upper body, it is also necessary to consider the
upper body's activities.
Therefore, this research aims to develop a walking
assistance orthosis that can be easily used daily and
induces trunk rotation by using leg movements
without motors. In this paper, we model the trunk
movement and conduct simulations to examine the
maximum torque the proposed orthosis assisting the
trunk rotation. Additionally, we prototype the
orthosis and verify its capability to provide assistive
torque to the trunk by leg movements.
2 THE CONCEPT OF THE
PROPOSED ORTHOSIS
Generally, walking is a whole-body movement
involving the lower and upper limbs. This movement
is governed by the Spinal Engine theory
(Gracovetsky, 1987), in which energy exchange
occurs between the trunk and the legs through the
spinal column. During this process, the trunk rotates
opposite to the pelvis rotation, thereby controlling
balance during walking. It is also considered that
increased trunk rotation leads to longer stride lengths
and higher walking speeds (Nishimori, 2006).
Therefore, based on these insights, this study aims to
indirectly assist walking by using lower limb
movements to assist trunk rotation. Furthermore, we
propose to simplify the system and eliminate battery
management by assisting trunk rotation without the
use of motors or other actuators.
Therefore, we consider using leg force as a source
of force to induce the trunk rotation. In other words,
the proposed orthosis converts the power generated
by leg movements during walking into trunk rotation
torque and utilize it to assist trunk rotation. Thus, the
proposed walking assistance orthosis uses leg
movements to assist trunk rotation. The specifications
of the proposed orthosis can be summarized as
follows．
1) Assisting trunk rotation using leg movement
2) Targeting users capable of independent
walking, aiming to maintain and improve
walking.
3) Lightweight, providing a comfortable fit to
minimize user burden.
The primary feature of this orthosis is 1): Assisting
trunk rotation by leg movements enhance stride
length and improves walking speed. Existing walking
assistive orthosis typically rely on external forces
such as actuators or springs to supplement muscle
strength during walking. In contrast, this orthosis
aims to promote walking only using the user's own
leg muscles without any actuators. Regarding feature
2), the orthosis is designed for users who want to
maintain their health. The feature of 3) is an essential
specification to ensure practical usability.
The conceptional design of the orthosis is depicted
in Figure 1. The user wears the orthosis, including
attachments for securing thigh wires. These wires are
connected to a bobbin located at the waist via pulleys.
The tension in the wires rotates the bobbin, and the
resulting torque is transmitted through shoulder plates
to induce trunk rotation. The motion of the orthosis
can be summarized in the following four steps．
(i) The legs’ swing motion pulls the wire.
(ii) The pulled wire rotates the wire bobbin via
the pulley.
(iii) The torque of the bobbin is transmitted to the
plates.
(iv) The torque of the plate is transmitted to the
shoulders, inducing trunk rotation.
It is important to note that excessive assistance torque
can increase the risk of injury. Therefore, to ensure
safety, a torque limiter is inserted between the bobbin
and plates to prevent the transmission of torque
exceeding the estimated maximum required for
assistance.
Figure 1: Conceptual image of the proposed orthosis.
3 ESTIMATION OF THE
TORQUE OF THE TRUNK
The leg movements generate the assistive torque for
the trunk rotation through wire, pully, and bobbin. We
simulate the trunk rotation to study the torque that
Development of Walking Assistance Orthosis by Inducing Trunk Rotation Using Leg Movement: 1 st Report on Prototype and Feasibility
Experiment
741
amplifies the effective trunk rotation angle for
walking.
Figure 2 is a top view of a user. The x-axis
indicates the walking direction of user, the grey circle
represents the head, and the yellow oval represents
the trunk. The trunk and head are simply regarded one
rigid body as shown in Figure 3, and the torque
around z-axis is considered. The torque around the z-
axis is composed of the
(
)
meaning the rotational
torque of the human muscle, and
(
)
being the
torque around the center of rotation of the bobbin of
this orthosis, which induces trunk rotation. This trunk
rotation is formulated based on a (Irving, H. P., 2007)
as (1).
(
)
+
(
)
+
(
(
)
)
−
(
)
=
(
)
＋
(
)
(1)
Here J is the moment of inertia when considering
the trunk and head as unified one rigid body, D
represents the viscous friction resistance, and K
1
to K
5
are coefficients related to elasticity. Hereafter,
(
)
is referred to as assist torque. The values of each
coefficient in the equation were taken from the
research (Aoki, 1998) (Yamazaki, 2006). The
moment of inertia J was calculated using the trunk
mass (Irving, 2007) with a body weight of 69 kg.
These values are shown in Table 1.
Figure 2: Coordinate setting (top view).
Figure 3: Trunk as a rigid body and torque around the axis
of rotation (top view).
We investigate the maximum value of
(
)
the
assist torque required to induce trunk rotation. For
this purpose, it is necessary to identify the rotation
torque
(
)
by the user's own muscles during
walking in
(
)
=0 in (1). For this purpose, we
referred to a Japanese typical trunk rotation angle
during walking (Hashimoto, 2018), and exploratively
determined the torque
(
)
so that the trunk rotation
angle θ(t) in (1) would be the same as this reference
angle. Figure 4 shows a reproduction and drawing of
the approximate trajectory of the trunk angle
(Hashimoto, 2018).
(
)
was identified
exploratively so that the trunk rotation angle θ(t) in
(1) to be same as Figure 4.
Figure 4: Trunk rotation angle during walking for
()
plotting the data from the study (Hashimoto, 2018)).
Table 1: Various coefficients.
[kg ∙ m
]
0.221
K
3
[rad] 30.0
D [kg ∙ m
/s]
0.020
K
4
[Nm] 94.75
K
1
[Nm] 0.679
K
5
[1/rad] 0.181
K
2
[1/rad] 0.181
K
6
[rad] -30.0
3.1 User’s Own Torque of Trunk
Since the angular frequency of the reference
trajectory in Figure 4 was 5.23 rad/s, the angular
frequency of
(
)
is also 5.23 rad/s. Therefore,
(
)
is expressed as
(
)
= sin(5.23 +
). (2)
The amplitude A and phase
in (2) were obtained
exploratively, and the reference trajectory in Figure.
4 coincides with θ(t) in case of (3) without assist
torque, i.e.,
(
)
= 0, which is the only torque by
user’s own muscles.
(
)
= 6cos(5.23) (3)
Figure 5 shows the trajectory of the rotation angle θ(t)
in the state (without the orthosis) without assist torque
and only rotation torque by the user’s muscles,
superimposed on the reference trajectory in Figure 4
． The red dashed line in Figure 5 is the reference
trajectory, and the blue solid line is the numerical
solution of (1) with (3) and
(
)
= 0. As this Figure
shows, the two lines are fitted, thus (3) represents the
torque by the human muscles for trunk rotation during
walking in the model of (1).
y
x
y
x
ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics
742
Figure 5: Reference trajectory (red dashed line) of trunk
rotation angle and trunk rotation angle with
(
)
= 6 cos
(5.23t),
(
)
=0 (blue solid line).
3.2 Required Assist Torque
we find the approximate maximum assist torque
(
)
required for assisting trunk rotation. When
determining the assist torque, there is a question as to
what degree of rotation angle is sufficient to facilitate
walking safely. In this paper, we focus on gait speed
among the gait parameters and find the maximum
value of assist torque
(
)
to satisfy the trunk
rotation angle required to increase the gait speed.
According to the research (Nishimori, 2006), the
walking speed is maximized when the trunk rotation
angle is 20 deg. The angular frequency of the assist
torque,
(
)
, is the same as the frequency of user’s
own toque,
(
)
, shown in (3). Thus, the assist
torque,
(
)
, is expressed as (4) with the angular
frequency as in (3).
(
)
= cos(5.23) (4)
The coefficients of B was obtained in an exploratory
manner so that the amplitude of θ(t) is 20 deg when
(
)
is in (3) (during walking). As a result, in case
of
(
)
= 3.25cos(5.23), (5)
the amplitude of θ(t) became 20 deg. Figure 6 shows
the trajectory of θ(t) when inputting
(
)
(5) and
(
)
(3) to (1). As Figure 6 shows, the amplitude of
the rotational angle of trunk, θ(t), reaches 20 deg in
the steady state. Compared to the rotation angle of 13
deg without assist torque (blue line in Figure 6), the
rotation angle increased by 7 deg with the addition of
assist torque. Based on these results, the guideline for
the maximum assist torque of this orthosis is 3.25
Nm.
Figure 6: Comparison of trunk rotation angle without
assisted torque(blue) and trunk rotation angle with assisted
torque
(
)
(green)．
4 PROTOTYPE
Figure 7 shows an overall view of the prototype. The
prototyped orthosis consists of the Top Part and the
Bottom Part connected by the flexible shaft; the
Bottom part extracts torque from the leg movements
and the Top Part transmits that torque to the trunk. At
the Bottom Part, both ends of a single stainless-steel
wire 1.25 mm are attached to the back of the each
thigh via nylon band. The centre of the wire is wound
around a bobbin. When one leg is raised, the wire is
pulled, and the bobbin rotates. This rotational torque
becomes the assist torque and is transmitted to the
Top Part via the flexible shaft. In other words, the leg
movement assists the trunk rotation by using the
orthosis.
Figure 7: Overview of orthosis.
Development of Walking Assistance Orthosis by Inducing Trunk Rotation Using Leg Movement: 1 st Report on Prototype and Feasibility
Experiment
743
Figure 8 shows the mechanism of the Bottom Part
where the bobbin is installed. To ensure that the
torque transmitted from the bobbin to the Top Part via
the flexible shaft does not exceed the maximum assist
torque, 3.25 Nm shown in (5), a torque limiter is
inserted between the bobbin and the flexible shaft.
The cut off torque of the torque limiter is adjustable
to individual physical characteristics in the range of
X to Z Nm. The radius of the bobbin was set to 0.04
m, in order to exceed the maximum torque of the
torque limiter. The thigh flexion moment during
walking is about 30 Nm (Yamamoto, 2003). The
distance from the root of thigh to the nylon band is
about 0.3 m. Therefore, the bobbin rotates a
maximum of approximate 4 Nm. Since there is
always slack in the wire, the tensioners are placed
along the wire's path. In Figure 8, the wire paths are
drawn with red line to easily identify it.
Figure 8: Bottom part of the prototype.
5 FEASIBLITY EXPERIMENT OF
FORCE TRANSMISSION
This experiment aimed to confirm the conversion of
leg movement to trunk rotation torque, which is the
primary function of the proposed orthosis. Although
it is naturally essential to verify the walking
promotion and safety, the first step is to verify the
generation of rotation torque of the trunk by the wire
drive. The reason is that without the generation of
rotational torque, the evaluation of gait promotion
and safety cannot be performed.
5.1 Method and Preparation
In the experiment, subjects were asked to wear the
prototype and flex their thigh to an arbitrary angle
from a standing position. Then, the generation of the
rotation torque is examined. In other words, the
relationship between the leg flexion angle and the
rotation torque is measured. A motion capture system,
PERCEPTION NEURON PRO by NOITOM, was
used to measure the leg flexion angle. The flexion
angle,
, was determined with the hip joint as the
origin and the direction of leg flexion as positive
shown in Figure 9.
Figure 9: Definition of flexion angle
.
The rotation torque is calculated based on the
measured force on the shoulder pressing against the
subject's upper body using the FSR-406 pressure
sensor. Figure 10 shows the sensor layout and jig. The
pressure sensor was placed at a distance of 0.1 m from
the centre of rotation of the gyration. In addition, a
hemispherical jig with a diameter of 45 mm was
attached so that the sensor's sensing area always
touched the subject's shoulder. A sensor was
calibrated in advance to convert sensor output
(voltage) and physical quantity (force).
The subject was asked to flex the right leg at any
time and any angle while supporting their posture by
touching the wall with one hand, as shown in Figure
9. The flexion angle and torque were measured by
performing this operation three times. There were two
subjects shown as Table 2.
Figure 10: Pressure sensor and its jig.
ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics
744
Table 2: Information of subjects.
Subject A Subject B
Age 23 22
Weight [kg] 62.6 60.0
Height [mm] 1653 1662
5.2 Result
The results are shown in Figure 11, Figure 12. Figure
11 shows an example of the results for subject A. The
orange line is the leg flexion angle, and the blue line
is the rotation torque calculated from the force of the
orthosis pressing on the shoulder. This figure shows
that the rotational torque increases with increasing leg
flexion. Figure. 12 shows an example of the results
for Subject B. As in Figure 11, the rotation torque
increases with leg flexion. Similar trends were
observed in other measurements. The results show
that the wire is pulled, the bobbin rotates, and the
rotation is transmitted to the shoulder, generating a
rotational torque that pushes the upper body.
Figure 11: Example of subject A’s rotational torque and leg
flexion.
Figure 12: Example of subject B’s rotational torque and leg
flexion.
Figure 13: Peak of flexion angles and torque of both
subjects.
Figure 14: Orthotic and upper body mechanics model.
Figure 13 plots the maximum torque versus leg
flexion angle for Subjects A and B. The green and
purple dots represent the results for subject A and
subject B, respectively. The dotted lines represent the
linear approximations of these data. Although this is
insufficient data for meaningful analysis, but
indicative trend. The trend is that the torque is
proportional to the flexion angle of the leg. Namely,
the slope of this approximate line can be regarded as
the equivalent spring coefficient, k, between the Top
Part of the orthosis and the subject's shoulder, as
shown in Figure 14.
Let
() be the rotation angle of the orthosis,
and
() ∝
(), because the angle
()
varies with the flexion of the leg
(
)
via wire,
pully, bobbin and flexible shaft. If the elasticity
between the orthosis and the upper body of subject is
expressed as an equivalent spring coefficient k, the
torque also increases in proportion to the angle
() ∝
() , based on the torsion spring
principle. This trend can be roughly seen in Figure 13.
Therefore, it was confirmed that the basic function
of the orthosis, i.e., to convert leg movement into
trunk assist torque, is feasible.
6 CONCLUSION
This research aims to develop an orthosis that
promotes gait by increasing the amount of trunk
rotation without any active actuators. In particular,
this paper reported on the prototype and basic
-10
0
10
20
30
40
50
60
70
-1.00
-0.50
0.00
0.50
1.00
1.50
2.00
2.50
024681012
Time [s]
Torque
Angle
: Flexion angle of leg
[deg]
:
Torque for trunk rotation [Nm]
-5
0
5
10
15
20
25
30
35
-0.50
0.00
0.50
1.00
1.50
2.00
02468101214
Time [s]
Torqu e
Angle
: Flexion angle of leg
[deg]
:
Torque for trunk rotation [Nm]
0
0.5
1
1.5
2
2.5
0 10203040506070
: Peak Flexion angle of leg
[deg]
Peak Torque for
trunk rotation [Nm]
Subject B
Subject A
Top Part
100[mm]
Development of Walking Assistance Orthosis by Inducing Trunk Rotation Using Leg Movement: 1 st Report on Prototype and Feasibility
Experiment
745
experiment to confirm that leg movement can be
converted into an assist torque to the trunk.
The experimental results confirm the feasibility of
the proposed orthosis to apply assist torque to the
trunk. Limitations of this paper include the limited
number of subjects and that torque was measured in a
standing leg flexion rather than in a walking position.
The next phase of the study will involve a
comprehensive experiment with a larger number of
subjects. We will employ a treadmill to quantify the
assist torque during walking, assessing the validity of
the torque thresholds. Furthermore, leveraging the
capabilities of both the treadmill and motion capture
systems, we will determine the rotation angle of the
trunk, the flexion angles of the legs, and the arms
during gait. This experiment will highlight the
effectiveness of our proposed orthotic.
REFERENCES
Aoki, K., & Yamazaki, N. (1998). The Role of Joint
Resistance by Passive Tissues in Human Bipedal
Walking. Society of Biomechanism Japan, 14, 59–68.
CYBERDYNE, (n.d.). What’s HAL?.
https://www.cyberdyne.jp/products/HAL/index.html
Gracovetsky, S. A., & Lacono, S. (1987). Energy Transfers
in the Spinal Engine. Journal of Biomedical
Engineering, 9(2), 99–114.
Hashimoto, K., Tanaka, T., & Kusaka, T. (2018). Trunk and
Pelvic Rotation Assist for Walking Assist and Running
Augmentation. Transactions of the JSME, 84(864), 18–
64. https://www.jstage.jst.go.jp/browse/transjsme/-char/
ja/
Imasen electric industrial Co., LtdCO, (n.d.). alQ.
http://www.imasen.co.jp/alq.html
Irving, H. P. (2007). Physics of the Human Body:
Biological and Medical Physics, Biomedical
Engineering. Springer-Verlag GmbH & CO.KG.
Nishimori, T., & Ito, A. (2006). The Cross-Correlation
Analysis of Trunk Rotation with Increased Gait Speed
during Walking and Running. Rigakuryoho Kagaku,
33(6), 318–323.
Murata, S., Murata, J., Ota, H., Matsunaga, H., Oyama, M.,
& Toyota, K. (2009). The Effect of Walking on
Physical, Cognitive and Psychological Functions of
Community-Living Elderly: A Randomized Controlled
Trial. Rigakuryoho Kagaku, 24(4), 509–515.
Ministry of Health, Labour and Welfare, Japan. (n.d.).
Physical Activity or Exercise. https://www.mhlw.go.jp/
www1/topics/kenko21_11/b2.html
Yamamoto, S. (2003). Biomechanics of Human Movement.
Rigakuryoho Kagaku, 18(3), 109–114.
Yamazaki, N., & Tanaka, T. (2006). Development of a
Human Joint Imitated Dummy. Biomechanism, 18,
175–185.
ICINCO 2023 - 20th International Conference on Informatics in Control, Automation and Robotics
746
