Development of a Standing Assistance Walker for a Patient with Low

Level of Care

Daisuke Chugo

, Shohei Kawazoe

, Sho Yokota

, Hiroshi Hashimoto

, Takahiro Katayama

Yasuhide Mizuta

and Atsushi Kojina

School of Science and Technology, Kwansei Gakuin University, Sanda, Hyogo, Japan

Department of Mechanical Engineering, Toyo University, Kawagoe, Saitama, Japan

Advanced Institute of Industrial Technology, Shinagawa, Tokyo, Japan

Service Robot Division, RT.WORKS CO., LTD, Osaka, Japan

Keywords: Standing Assistance, Active Walker, Patient with Low Level of Care.

Abstract: This paper proposes a novel low cost robotic walker with standing assistance function. Our system focuses

on domestic use for elderly people who is low level of care and need nursing in their day-to-day lives.

Usually, these patients require a partial standing assistance only when they need it, not a full assistance

during standing motion such as a hanging by the lift. The widely and easily use of such assistance in daily

life will be successful in ensuring safety and providing an inexpensive manufacturing cost. These two

opposed requirements have been realized with our developed robotic walker. Our key ideas are two topics.

First is proposal of a mechanical design with minimum and smaller actuators. Proposed system uses a gas

spring which helps the up/down actuator and our system assists the patient with wheel actuators on a

powered walker for stabilizing its user as well as for lifting up the user. Second is assistance procedure

which leads the patient to suitable posture by the force guidance and voice instruction. We investigate what

factor enables the patient to stand up safety by preliminary experiment. The performance of our proposed

system is verified through experiments using our prototype with elderly and handicapped subjects.

1 INTRODUCTION

Activities such as standing, walking, and sitting may

be the most serious and important activities in the

day-to-day lives of elderly people as they lack

physical strength (Alexander et al., 1999; Hughes et

al., 1996). However, assisting elderly individuals in

these tasks can be difficult for caregivers and can be

a primary source of the lumbago that many of them

experience. Thus, developing a caregiving robot

capable of assisting the elderly when they stand,

walk, and sit is important, and many such devices

have been developed and reported in previous

studies (Nagai et al., 2003; Funakubo et al., 2001).

In Japan, elderly people requiring assistance in

their daily lives are classified into five different care

levels (Cabinet Office, Government of Japan, 2016),

where care level 1 is minor and care level 5

represents a serious condition such as bedridden life.

Elderly people within care level 2 or less are more

than 60% and voluntarily body movement in normal

daily-life activities is important in order to keep their

physical strength, and thereby preventing to become

worse care level (Hirvensalo et al., 2000). They have

difficulty in standing, walking or sitting on their own

but are otherwise able to perform routine activities if

partial assistance only for these motions is provided.

This paper calls these situations as low level of care.

In many previous researches, devices that can aid

in such activities are developed (Munro et al., 1998),

but these are designed for care houses and hospitals

because their motivation is reducing the caregivers’

burden. On the other hand, the assistive robot for

low level of care people should be widely used in

their homes. For realizing them, the robot is required

to be practical and low cost. The robot should be

compact for easy use because standing, walking and

sitting motion will be done in narrow space in daily-

life activities. Furthermore, the robot should have

enough assistive performance and fail-safe design

providing an inexpensive manufacturing cost.

However, no such robots have yet been developed.

In our previous studies, we developed an

assistive robot to continuously aid patients with

activities such as standing, walking, and sitting

(Chugo et al., 2015; Chugo et al., 2012). The robot

142

Chugo, D., Kawazoe, S., Yokota, S., Hashimoto, H., Katayama, T., Mizuta, Y. and Kojina, A.

Development of a Standing Assistance Walker for a Patient with Low Level of Care.

DOI: 10.5220/0006469501420153

In Proceedings of the 14th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2017) - Volume 2, pages 142-153

ISBN: Not Available

was based on a walker (a popular assistance device

for elderly people to use in normal daily life) and

had a manipulator with three degrees of freedom

(DOF) to assist patients in standing. We designed

the robotic walker for realizing enough performance

in standing, walking and sitting assistance with

safely, but however, we did not consider its

manufacturaing cost. Its body size was too big for a

typical toilet room in Japan, therefore, it was not

practical in home usage. Furthermore, this system

used many actuators and high-precision sensors, thus,

its cost was too expensive and not acceptable for

home usage of the patient who is low level of care.

For realizing the assistive robot which the many

elderly people can use in their daily-life activities,

the robot should practical and low cost, and of

course should have enough assistance performance.

Therefore, this paper present a novel standing

assistance walker.

For relazing practical robot, we mainly describe

two key topics. First is proposal of a mechanical

design with minimum and smaller actuators.

Proposed system can lift the patient’s body with a

smaller actuator force by combination of a linear

actuator and a gas spring. A gas spring generates

upper direction force when it lifts up. On the other

hand, it stores upper direction force from the

patient’s body weight when it takes down.

Furthermore, developed system enables standing

assistance with only one linear actuator for lifting up

the patient’s body by using wheel actuators on a

powered walker for stabilizing its user.

Second topic is proposal of assistance procedure

which leads the patient to suitable posture by the

force guidance and voice instruction. For realizing

safety standing assistance, the subject is required to

take a stable posture in standing with the robot.

However, it is difficult to guide the motion of the

patient because the assistance for the low level of

care should fir the patient motion based on his/her

will, should not assist all necessary force for doing a

standing motion. Thus, we investigate what factor is

useful to guide the patient’s motion by preliminary

experiment and with this result, this paper proposes

a standing assistance procedure with force guidance

and voice instruction.

The rest of this study is organized as follows:

section 2 explains the configuration of the proposed

system, section 3 describes the assistance procedures,

section 4 describes a practical experiment with

elderly and handicapped people, and section 5

concludes the study.

2 MECHANICAL DESIGN

PROPOSAL WITH MINIMUM

AND SMALLER ACTUATORS

2.1 Required Condition

2.1.1 Required Assistance Function

As mentioned in the Introduction, elderly people

who is low level of care can be considered to be the

main audience of our assistance robot. The

characteristics of these people are follows (Cabinet

Office, Government of Japan, 2016);

 The patient has dexterity to take suitable

posture if physical load is small.

 The patient can maintain his/her body balance

by grasping the handle on the assistive device.

In other word, he/she has enough force to

grasp it.

 The patient requires force assistance for

reducing physical load when he/she lifts up

his/her trunk in standing.

 The patient requires assistance for keeping its

body balance during standing, walking and

sitting assistance.

From these conditions, the assistive robot should

have 2DOF minimally, one is up/down direction for

lifting the patient’ body and the other is

forward/backward direction for keeping his/her body

stability. We do not consider the right/left direction

because we can approximate human standing motion

based on the movement on a 2D plane (Nuzik et al.,

1986).

2.1.2 Required Condition for Practical

Usage

Since the proposed walker is small and mobile, it

can be used in any situation in users’ homes. In

typical scenario, if a patient would like to go to the

toilet room from his/her bed via the corridor, he/she

can stand up with the assistance of the walker, walk

through the corridor without scratching the wall,

enter the bathroom, turn around into the sitting

position, and sit down with the device’s assistance.

In this scenario, the narrowest room in Japanese

typical home is a toilet room. A typical toilet door

with a standard width is 600mm (JIS - Japan

Industrial Standard – 1526:1997) and in the toilet

room, the width is 800mm minimally. Therefore, for

using the robot in daily-life activities, the robot

should have the following specifications.

 The robot can pass the entrance with 600mm

width.

Development of a Standing Assistance Walker for a Patient with Low Level of Care

143

 The robot can rotate with its patient in the area

with 800mm circle.

 The robot can approach a chair, a bed and a

toilet enough which enable the sitting patient

to use its standing assistance.

 The robot can pass the small steps between the

room and corridor floor. Usually, its height is

within 20mm in the typical Japanese house.

2.2 Developed Robot

2.2.1 System Overview

Considering these conditions in section 2.1.2, we

propose the assistive robot as Fig. 1. Our robot

consists of a powered walker and a standing support

manipulator, which moves the user in an upward

direction so as to be lifted. A standing assistance

manipulator has 1 DOF (up/down direction) which is

generated by a linear actuator and a gas spring (see

section 2.2.2). In standing, our robot assists the

patient cooperating with a standing assistance

manipulator and wheel actuator. (see section 2.2.3)

Fig. 2 shows a top view of our robot. Its width is

540mm and can pass easily a typical entrance in the

patient’s home. Our robot has two actuated wheels

in each side. Their axle is same position as the foot

center of its patient and he/she can turn easily within

the circle which diameter is 800mm.

As Fig. 1(b), our robot uses large casters at front

position for increasing the mobile performance on

the non-flat ground. Its diameter is 120mm and it

can pass easily the 20mm height step. On the other

hand, our robot uses small casters at the rear position

for preventing the conflict between the caster and

objects as legs of the chair which its patient sits on.

540mm

Gas Spring

Linear

Actuator

965mm

500mm

Actuated Wheel

Front Caster

Rear Caster

(a) Side view (b) Front view

Figure 1: Our developed robot for standing assistance.

540mm

500mm

340mm

UserRotation Center

Foot Position

Axle Position of Actuated Wheels

Figure 2: Top view and turning radius of our robot.

2.2.2 Standing Manipulator

A standing manipulator lifts up the patient body

directly and its load tends to be large. Generally for

this purpose, a high powered actuator is suitable,

however, its cost is expensive and there is its

malfunction risk. A smaller actuator with high

reduction gear is useful choice, however, maximum

lifting velocity will be reduced and the robot cannot

lift up the patient by a required velocity.

Thus, this study proposes a novel mechanism

combing a linear actuator and a gas spring as Fig.

1(b). A gas spring can output force almost constant

during its stroke, therefore, it helps the actuator

when the standing manipulator lifts the patient. On

the other hand, when a gas spring shrinks, it requires

down direction force. Usually, in this situation, the

standing manipulator assists in sitting, and a gas

spring shrinks by the body weight of its patient.

Therefore, this device is useful for this purpose and

furthermore, a gas spring is widely used and its cost

is inexpensive.

Generally, a gas spring generates the force

when it extends as (1), and it requires the external

force

when it shrinks as (2) as Fig. 3. Because of

its internal resistance

is larger than

as (3).

max

minmax

stroke

f +

−

−=

(1)

max

minmax

stroke

f +

−

−=

(2)

rlu

fff +=

(3)

where

is the manipulator position and

stroke

is its

stroke.

()

uorlif

max

is maximum force which

the gas spring can generate and usually, it can

ICINCO 2017 - 14th International Conference on Informatics in Control, Automation and Robotics

144

generate at lowest position

min

()

uorlif

min

is its minimum force at highest position

max

If the subject applies the maximum load

max

the robot at

lift

when it assists in standing, the

following conditions should be fulfilled.

 The total output by the linear actuator and the

gas spring is larger than the maximum load as

(4) when the robot assists to lift up the patient.

 For shrinking the manipulator without the body

weight of the patient, the output of linear

actuator is larger than the maximum force

which the gas spring requires to shrink as (5).

(

)

max

ffyf

aliftl

(4)

maxua

ff >

(5)

With our proposed mechanism, our robot uses

the linear actuator which can generate

400N

and the gas spring which specifications are shown in

Table 1. These selected devices are fulfilled these

conditions discussed above.

Fig. 3 is the output force of the gas spring and

the typical applied load when the 90kg body weight

patient stands up with our robot (Chugo et al., 2016).

During the lifting up the patient’ body (

is around

50mm to 130mm), the standing manipulator can

generate enough upper direction force (more than

650N) with a linear actuator which capacity is 400N.

Using our proposed idea, our robot can use a

smaller actuator, which means that its design can be

fairly inexpensive. Furthermore, the gas spring

prevents the standing manipulator from moving

suddenly when the power is down.

Table 1: The specifications of the gas spring.

()

minmax

yff

373N at y=0mm

()

maxmin

yff

270N at y=270mm

()

minmax

yff

313N at y=0mm

()

maxmin

yff

240N at y=240N

stroke

270mm

200

400

600

800

0 30 60 90 120 150 180 210 240 270

Position of the manipulator (mm)

Upper Force (N)

Extended Force

Shrink Force

Tota l O utput (Lifting)

Ap plie d F o rce

Shrink Direction

Extended Direction

Lifting Body

f−

ff +

max

min

max

Figure 3: The output force of the gas spring.

2.2.3 Powered Walker

We developed an assistive robot to continuously aid

patients with standing, walking, and sitting as Fig. 4

(Chugo et al., 2012). The movement pattern

in Fig.

4 refers to a ratio of the standing motion as

determined by (6).

is the time required to

complete the standing operation, and

is the present

time.

s =

(6)

Our developed robot had a standing manipulator

with 3DOF to assist patients in standing, because

standing motion consists of three phases.

 The first phase, the patient inclines his upper

body to the forward direction and moves the

center of gravity (COG) to the foot area as Fig.

4(a).

 The second phase, he lifts up his upper body

from the chair as Fig. 4(b).

 The last phase, he extends his knee joint

completely and ends the standing motion as

Fig. 4(c).

(a) 25% (b) 50% (c) 75%

Figure 4: Suitable standing posture guided by our previous

standing assistance system (Chugo et al., 2012).

Therefore, the standing assistance requires at least

more than 2DOF. For realizing low cost system, our

developed robot consists a standing assistance

manipulator which has 1 DOF (up/down direction)

and a powered walker which has also 1DOF

(forward/backward direction). In standing, our robot

assists the patient cooperating with a standing

assistance manipulator and wheel actuator. Using

this design, our robot realizes 2DOF with simple

standing manipulator.

By this idea, the powered walker is required to

assist not only in walking but also in standing. The

standing assistance requires the following function

in forward/backward direction (Chugo et al., 2012).

 The first phase, the powered walker should

guide the patient’s upper body to the inclined

posture as Fig. 4(a).

Development of a Standing Assistance Walker for a Patient with Low Level of Care

145

 During standing, the powered walker should

keep the body stability in forward/backward

direction.

Basically, these conditions require position

coordination function and generally, position control

is suitable. On the other hand, the robot should apply

the force to the patient like force feedback for

leading him/her. For this purpose, force control is

also suitable.

For realizing these functions, the developed

powered walker has an encorder and an ammeter on

each wheel. Using these sensors, it can measure its

movement distance and the applied force by its

patient in forward/backward direction. Each wheel is

actuated by the motor driver which can control the

wheel cooperating a standing manipulator with

position control mode. Using this hardware, we

proposes wheel control scheme which combines

position and damping control mode as (7).

()

(

)

ref

iixx

ref

xxKFFBxx −−−−=



(7)

() () ()

[]

ref

xsxx 1,,

,,0 



(8)

where

ref



is the velocity control reference

()

rightorleftj =

, which is a function of the

movement pattern

defined in (8). This reference is

calculated from the standing movement

recommended by the physical therapists in section

3.1.

is the applied force to the forward/backward

direction by its user.

ref

is the position reference

and

is the actual position.



is the updated

reference that proposed controller inputs to the

motor driver during standing assistance.

is the

coefficient and force that the patient applies to the

robot while he/she stands. Using (7), our developed

walker has both functions of the position control

mode and the damping control mode, and it can

fulfilled the required function for standing

assistance. B and K are constants that coordinate the

ratio between the damping and position controls. We

discuss on the parameter setting in section 3.3.2

2.2.4 User Interface

A handle, armrest, and controller are provided on the

top of the walker, as shown in Fig. 5(a). There are

force sensors inside the armrests which measure the

applied force to the vertical direction, and touch

sensors on the handles. When the patient wants to

move, he/she has to put his/her arm on the armrest

and grips the handles. Using the touch sensors and

the force sensors, our robot judges whether the

patient is ready to stand; if it judges him/her to be

ready, our device guides the patient to push a

gripping switch using a voice instruction (These

voice instructions will be explained in the section

3.3.).

A gripping switch is provided on each handle, as

shown in Fig. 5(b). This switch has two input steps

that can be changed by the strength used for the grip.

Usually, in emergency situations, elderly people

tend to release the control switch or push it strongly

because of the fear of falling (E. Maki et.al., 1991) .

Therefore, we use the two-step switch in such

conditions, as shown in Fig. 5(b), and our robot

provides assistance for standing only in the case of

the first step, whereas in the case of the second step,

our robot regards the user as being in an emergency

situation.

Force Sensor

Touch Sensor

Power Switch

OFF

STEP

OFF

STEP

(a) Handle and armrest (b) Gripping switch

Figure 5: Its user interface.

3 ASSISTANCE PROCEDURES

For realizing standing safety, the patient should take

a suitable posture during standing. Our main

audience is the low level of care patients, therefore,

they has dexterity to take suitable posture if suitable

guidance is provided. Therefore, we propose the

guidance scheme which leads the patient to take an

inclined posture using force guidance and voice

instruction.

3.1 Motion Recommended by Nursing

Specialists

In a previous study, different types of standing-up

movements were proposed. Kamiya (Kamiya, 2005)

proposed a standing-up movement that utilizes the

remaining physical strength of a patient, as

determined by their nursing specialist. Fig. 6(a)

shows an example of this movement proposed by

Kamiya.

In our previous study, we analyzed this standing

movement, and we found that Kamiya’s proposal

ICINCO 2017 - 14th International Conference on Informatics in Control, Automation and Robotics

146

was effective in enabling the patients to stand up

with minimum load (Chugo et al., 2012). We

assumed that the standing motion is symmetrical and

discussed the motion as a movement of the linkages

model on a two-dimensional (2D) plane as shown in

Fig. 6(b) (Nuzik et al., 1986). We measured the

angular values among the linkages as these reflected

the relationship between different parts of a body.

From the measured results, we can verify that to

achieve the motion proposed by Kamiya, a patient’s

trunk needs to incline in the forward direction while

getting up from a chair, as shown in Fig. 7(a). In this

figure, the Y-axis shows the angular values of the

pelvis and trunk, knee, and ankle), whereas the X-

axis shows a movement pattern (Nuzik et al., 1986),

which is the ratio of the standing-up motion, as

shown by (6). Fig. 7(b) shows the position of a

patient’s center of gravity (COG), which indicates

the body balance of the patient during the standing

motion.

(a) Assistance motion (b) Its coordination

Figure 6: Standing-up motion as described by Kamiya.

shows the angular value of the pelvis and the trunk.

and

show the angular values of the knee and the ankle,

respectively (Chugo et al., 2012).

To realize this motion, we derived the control

reference of our assistance system kinematically. We

assume the human model as Fig. 6(b) moves each

joint according to the measured values as Fig. 6(a).

For assisting this human model movement, we can

derive the position which the standing manipulator

should assist in standing (Chugo et al., 2012). In this

study, our robot uses them as the control reference.

For this derivation, the parameters were chosen from

the standard data of the body of an adult male

(Okada et. al., 1996), as shown in Table 2.

Fig. 8 shows the positions of the handle in

standing. In Fig. 8, the Y-axis shows the up/down

position (by the standing manipulator) or the

forward/backward position (by a moving function on

a powered walker) of the handle, and the X-axis

100

120

140

160

180

0255075100

Movement Pattern (%)

Angle (deg)

Pelvis/Trunk

Knee

Ankle

-0.25

-0.2

-0.15

-0.1

-0.05

0.05

0.1

0.15

0.2

0.25

0 25 50 75 100

M o ve m e n t P atte rn (% )

Center of gravity

Foot

Lifting up

-0.25

-0.2

-0.15

-0.1

-0.05

0.05

0.1

0.15

0.2

0.25

0 25 50 75 100

M o ve m e n t P atte rn (% )

Center of gravity

Foot

Lifting up

(a) Angular values of each joint (b) Its coordination

Figure 7: Analysis of the standing-up motion proposed by

Kamiya. The size of the foot of the human model was

0.26m, and the foot area is shown by the red arrows in (b).

At a 25% movement pattern, the subject lifts up his/her

body.

shows the movement pattern. The coordination of

Fig. 8 is defined as in Fig. 6(b). Using these tracks

as the position control reference, our robot can

realize the standing motion proposed by the nursing

specialist.

Table 2: Human Parameters.

Linkage Name Length [m] Width [m]

Head 0.28 0.21

Trunk 0.48 0.23

Hip 0.23 0.23

Humerus 0.39 0.12

Arm 0.35 0.08

Hand 0.2 0.07

Femur 0.61 0.17

Leg 0.56 0.16

Foot 0.26 0.11

-100

100

200

300

400

500

600

-50

100

150

200

250

0 20406080100

Moving Distance [mm]

(Forward/Backward)

Moving Distance [mm]

(Up/Down)

Movement Pattern [%]

Up/Down

Forward/Backward

Figure 8: The reference in standing.

3.2 Force Guidance in Standing

For guiding the patient to take the inclined posture

when the robot starts to assist in standing, our robot

moves to the forward direction according to the

reference as Fig. 8 and this movement tells the

patient that he/she should incline his/her upper body

to the forward direction. In our previous work, we

found the suitable force applying could tell its user

Development of a Standing Assistance Walker for a Patient with Low Level of Care

147

how the robot would guide him/her (Chugo et al.,

2015). Thus, in section 2.2.3, we propose the wheel

control scheme which has both position control

performance and force control performance.

Proposed controller changes both performance

by two coefficients, B and K in (7). B coordinates

force control performance ratio and K coordinates

position control performance ratio. In this paper, we

investigate the suitable ratio between two parameters

for guiding the patient to the suitable posture by the

preliminary experiment.

3.2.1 Preliminary Experimental Setup

In this experiment, subjects try three test cases as

Table 3. Subjects are 23 young students whose age

are 21 to 24. All subjects use our robot for the first

time and we request them to stand simply according

to the robot’s movement. After this experiment, we

ask them two questions. First question is “Did you

notice the robot tried to make you do what kind of

movement at the beginning?” Second question is

“How feel did you during standing assistance by our

robot?” Seven subjects try the standing assistance

provided by case1, another eight subjects are case2

and another eight subjects are case3.

Table 3: Test cases in the preliminary experiment.

B K

Case 1: Force mode 0.8 0.2

Case 2: Moderate Mode 0.5 0.5

Case 3: Position Mode 0.2 0.8

3.2.2 Preliminary Experimental Results

Table 4 shows the experimental results. By standing

assistance by our robot, in case2, almost all subjects

can stand according to the reference. On the other

hand, in case1, in some trials, the subject fails to

stand. Fig. 9(a) shows the typical failure. In this

failure, the subject noticed the robot tried to guide to

the forward direction by its force. However, the

subject could not find the suitable position because

the robot did not show the reference position clearly

because of the low position control ratio, and as the

result, the subject failed to stand as Fig.9 (a). In

case3, the subject also failed to stand in some trials

as Fig. 9(b). In this failure, the subject did not notice

the robot guided to the forward direction because the

guidance force was weak. As the result, the subject

did not move the position of COG to the forward

direction and his body balance was unsuitable.

From the questionnaire results as Table 4, in

case1 and 2, almost all subjects noticed the robot

guidance to the forward direction, thus, force control

approach seems to be effective for this purpose.

However, too strong force causes the subject felt a

fear and should be avoided by the results in Table 4.

In case3, some subjects did not notice the robot

guidance to the forward direction and it causes the

standing assistance was uncomfortable. This means

to provide the effective standing assistance, the

powered walker should have both position control

function and force control function.

From these results, our powered walker uses the

parameter settings as case2 which activates both

position and force control function in standing.

Table 4: Experimental Results and Questionnaire

Answers.

Case1 Case2 Case3

Success in standing by

our robot

5/7

71%

7/8

88%

4/8

50%

Question 1: The inclined

body posture

7/7

100%

7/8

88%

5/8

63%

Question 2: Fear or

uncomfortable

3/7

43%

0/8

3/8

38%

Too Forward

Not Enough to

Forward

(a) case1 (b) case3

Figure 9: Typical failure of standing.

3.3 Assistance Procedure with Voice

Instruction

For safety standing, our robot guides the patient by

voice instruction. At the beginning in standing, the

patient’s upper body needs to incline in the forward

direction. From the opinions of the physical

therapists, these information are required for the

patient to take this posture.

 The patient should incline his/her upper body to

the forward direction.

 The patient should face to the bottom of the

forward direction.

ICINCO 2017 - 14th International Conference on Informatics in Control, Automation and Robotics

148

 The foot should move to the back position,

should not take the posture which throws out

his/her leg.

Considering with them, we propose the

assistance sequence with voice instruction as Fig. 10.

Table 5 shows the voice instructions provided in

Japanese by the device as well as their English

translations.

When a user turns on the power of the walker, an

announcement (Message A) is spoken. After that,

the walker remains in a waiting state until both touch

sensors and force sensors are turned on. The user has

to touch the gripping switches and put their weight

onto the armrests, because the device must first

check whether the user is holding the walker

properly to decide whether it is safe to provide the

assistance.

If these sensors respond, a voice announcement

(Message B) tells the user to stand ready to move.

After this, when the user grips the switches on the

handle as the first-step input shown in Fig. 5, the

device initiates its standing assistance. The user has

to continue holding the switches on the first-step

input, as elderly people generally tend to release

their grasp or become stiff if they feel scared (Omori

et al., 2001). Thus, if the user releases its grip, the

second-step input or no input, the system stops the

assistance. When no further assistance is required,

the actuators stop moving and a voice announcement

encourages the user to walk.

During the standing motion, our device leads

user to a suitable standing posture using the two

DOF (i.e., the up/down direction and

forward/backward direction).

After the user stands up, they can use the device

as a powered walker (Hirata et al., 2007).

Table 5: Voice Announcements.

No Voice Message Its Objective

A I’ll do my best to

support you.

Saying hello to the

user.

B Move your feet back

and bend your body to

forward. Then, grip the

switches on the handle.

Telling the user to

ready his/her posture to

stand up soon.

C Let’s stand up together. Signal of start of the

standing up motion.

D Have done. Let’s walk

carefully with me.

Signal of end of the

standing up motion and

encouraging the user to

walk.

Figure 10: Standing assistance process flow.

4 ASSISTANCE PROCEDURES

To confirm the efficiency of the proposed assistive

robot, we conducted a practical experiment.

4.1 Experimental Setup

To verify its effectiveness, we used three test cases.

In case1, our proposed system assisted a standing

motion with all proposed technique. In case2, our

system assisted a standing motion without proposed

force guidance function (only velocity control,

B=K=0 in (7)), because it simulates standing

assistive devices traditionally provided by many

manufacturers (Funakubo et al., 2001). In case3, our

system assisted a standing motion with only the

standing manipulator, and it simulates the automatic

movable handrail equipped the bedside which is

widely used in care houses and hospitals.

We used four subjects. All subjects were elderly

or handicapped people with disabilities and required

standing assistance in their daily activities. All the

details about these four subjects are provided in

Table 6.

Development of a Standing Assistance Walker for a Patient with Low Level of Care

149

During this experiment, we measured the body

movement by the motion capture system and the

applied force to the up/down and the forward/

backward direction by equipped sensors on our robot.

Using measured values, we can estimate the traction

output of waist, knee and ankle joint as an index of

the physical load of the patient. For detail estimation

scheme, please refer our previous research (Chugo et

al., 2015).

All the experiments were performed by nursing

specialists and under the ethical rules and technical

safety measures provided by the Yokohama

Rehabilitation Center, Shin-Yokohama, Kanagawa,

Japan.

Table 6: Subjects.

Weight

/Height

Age

Care

Level

Remarks

60kg

/170cm

60 Level2

Peripheral

neuropathy,

Paraplegia

78kg

/178cm

52 Level2

Ataxic both sides

hemiplegia

68kg

/152cm

68 Level2

Limb paralysis,

Parkinson's disease

58kg

/178cm

34 Level1

Hypoxic

encephalopathy,

Limbs and trunk

ataxia

4.2 Experimental Results

Figs. 11–14 are visual descriptions of the

experiments. Fig. 11 describes a series of standing

scenes of subject A, whereas Fig. 12 is about subject

B, Fig. 13 is about subject C, and Fig. 14 is about

subject D. These pictures show that all the subjects

were able to stand up without the occurrence of any

accidents.

(a) 20% (b) 50% (c) 100%

Figure 11: Subject A (Case1).

(a) 20% (b) 50% (c) 100%

Figure 12: Subject B (Case1).

(a) 20% (b) 50% (c) 100%

Figure 13: Subject C (Case1).

(a) 20% (b) 50% (c) 100%

Figure 14: Subject D (Case1). For safety reason, a

therapist stays near the subject during this experiment.

All subjects evaluated our assistance robot using a

questionnaire after the experiment as Table 7. The

subjects A to C evaluated case1 is better and on the

basis of their responses, we were able to tell that

leading to the suitable posture was important during

the standing assistance. Subject D, meanwhile,

found case 3 to be better because he had limbs and

trunk ataxia caused by hypoxic encephalopathy and

leaned completely against our assistance system.

Fig. 15 shows the estimated torque on the ankle,

knee, and waist of subject A on each case.

Furthermore, we show the estimated torque when he

stands up only by his own physical strength using a

handrail equipped on the bedside. In Fig.15(a)

without assistance case, maximum traction is about

1.0 Nm/kg on a knee joint. In Fig.15(b), traction is

within 0.5Nm/kg in case1, in Fig.15(c), traction is

within 0.8 Nm/kg in case2 and in Fig.15(d), traction

is within 0.6 Nm/kg in case3. From previous

ICINCO 2017 - 14th International Conference on Informatics in Control, Automation and Robotics

150

research, maximum traction should be within 0.5

Nm/kg for safety standing motion by own physical

strength of elderly people (Fisher et al., 1990).

Case1 has a best assistance performance and this

result indicates the suitable posture during standing

motion is maintained using our ideas. Furthermore,

Fig. 16 shows maximum traction output of a knee

joint when the subjects A to D lift up his body.

According to these results, the subjects were

supported with the lowest burden in all of the three

cases, and case1 has best assistance performance in

standing.

Fig. 17 shows the position of the COG on the

forward/backward direction of subject A. These

results were calculated according to the linkage

model and the assumptions outlined in section 3.1.

As shown in Fig. 17, the COG movement in case 1

was closest to the reference. In case 2, the COG was

over 20 cm, which means that the traditional

controller led to the users learning too far forward.

In contrast, in case 3, the COG was less than 10 cm,

which implies that the users did not move forward

enough to bend their body or may be led in danger.

Moreover, Fig.18 shows the COG of subject A to D

at 60 % movement pattern. At this time, subjects

incline their trunk and lift up it to upper position.

According to this result, in all subjects, the COG fit

the designed reference and we can evaluate the body

balance is suitable in case1. In case2, the COG is too

far and in case 3, the COG is too close. These

unsuitable COG lead a risk of falling down and in

the questionnaire results as Table 6, some subjects

feels it. On subject D, COG tends to be large value

because this subject leaned completely against our

assistance system.

According to these results, our robot succeeds to

assist the subjects with the lowest burden and

suitable body balance during standing motion with

proposed robot system (case1). Moreover case3

(position fix version) may be effective when the

target user completely does not have dexterity to

maintain a body balance.

Table 7: Questionnaire results.

No Case1 Case2 Case3

A Good

Body balance is

bad.

Body balance is

bad.

B Good Fear of falling.

Body balance is

bad.

C Good

Acceptable, case

1 is better.

Body balance is

bad.

Acceptable,

case3 is better.

Fear of falling. Good

0.2

0.4

0.6

0.8

1.2

0 20406080100

Torque [Nm/kg]

Movement Pattern [%]

Ankle

Knee

Waist

(a) Without Assistance

0.2

0.4

0.6

0.8

1.2

0 20406080100

Torque [Nm/kg]

Movement Pattern [%]

Ankle

Knee

Waist

(b) Case1

0.2

0.4

0.6

0.8

1.2

0 20406080100

Torque [Nm/kg]

Movement Pattern [%]

Ankle

Knee

Waist

0.2

0.4

0.6

0.8

1.2

0 20406080100

Torque [Nm/kg]

Movement Pattern [%]

Ankle

Knee

Waist

(d) Case3

Figure 15: Traction output of each joint (Subject A) during

standing motion.

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Subject A Subject B Subject C Subject D

Traction [Nm/kg]

Case1

Case2

Case3

Figure 16: The maximum traction output in each subject.

-200

-100

100

200

300

0 20406080100

Position of COG [mm]

Movement Pattern [%]

Case1

Case2

Case3

Reference

Figure 17: The position of COG (Subject A) during

standing motion.

Development of a Standing Assistance Walker for a Patient with Low Level of Care

151

100

150

200

250

Subject A Subject B Subject C Subject D

Position of COG [mm]

Case1

Case2

Case3

Figure 18: The position of COG at 60[%] movement

pattern in each subject.

5 CONCLUSIONS

This paper proposes a novel low cost robotic walker

with standing assistance. Proposed robot focuses on

domestic use for elderly people who is low level of

care and need nursing in their day-to-day lives. For

the robot to be used widely and easily in daily life,

it is important to ensure safety and provide an

inexpensive manufacturing cost.

For realizing two opposed requirements, this

paper proposes the novel mechanism design and the

assistance procedure which leads the patient safety

and stability. Proposed mechanical design uses a gas

spring which helps the lifting linear actuator with

minimum cost and developed robot assists the

patient with wheel actuators on a powered walker

for stabilizing its user as well as for lifting up the

user. Furthermore, proposed assistance procedure

leads the patient to suitable posture by the force

guidance and voice instruction. For realizing it, we

investigate what factor is useful for leading the

patient by preliminary experiment.

The developed prototype has enough assistance

performance through experiments with elderly and

handicapped subjects. Thus, our study succeeds to

develop a safety and low cost robot which has

enough standing assistance performance for the

patient who is low level of care.

For our future work, we plan to develop the

wheel control algorithm for walking assistance.

ACKNOWLEDGEMENTS

The authors would like to thank T. Kumeda, a

physical therapist, and all the staff members at

Yokohama Rehabilitation Center for providing

technical assistance in the experiments.

This work is supported in part by the Matching

Planner Program (MP28116808328) from Japan

Science and Technology Agency (JST), Grant-in-

Aid for Scientific Research C (16K01580) from

Japan Society for the Promotion of Science (JSPS)

and the Robotic Devices for Nursing Care Project

from Japan Agency for Medical Research and

Development, AMED.

REFERENCES

N. B. Alexander, A. B. Schultz and D. N. Warwick, 1991.

Rising From a Chair: Effects of Age and Functional

Ability on Performance Biomechanics. In J. of

Geometry: MEDICAL SCIENCES, Vol.46, No.3,

M91-98.

M. A. Hughes, M. L. Schenkman, 1996. Chair rise

strategy in the functionally impaired elderly. In J. of

Rehabilitation Research and Development, Vol.33,

No.4, pp.409-412.

Cabinet Office, Government of Japan, 2016. KOUREISHA

HAKUSHO (The whitepaper on the aged society),

ISBN: 4865790519, pp.25, (in Japanese).

K. Nagai, I. Nakanishi and H. Hanabusa, 2003. Assistance

of self-transfer of patients using a power-assisting

device. In Proc. of the IEEE Int. Conf. on Robotics

and Automation, pp.4008-4015.

A. Funakubo, H. Tanishiro and Y. Fukui, 2001. Power

Assist System for Transfer Aid. In J. of the Society of

Instrument and Control Engineers, Vol.40, No.5,

pp.391-395.

M. Hirvensalo, T. Rantanen and E. Heikkinen, 2000.

Mobility difficulties and physical activity as predictors

of morality and loss of independence in the

community-living older population. In J. of the

American Geriatric Society, Vol.48, pp.493-498.

B. J. Munro, J. R. Steele, G. M. Bashford and M, Ryan,

1998. A kinematic and kinetic analysis of the sit-to

stand transfer using an ejector chair: implications for

elderly rheumatoid arthritic patients. In J. of

Biomechanics, pp.263-271.

D. Chugo, Y. Morita, Y. Sakaida, S. Yokota, H.

Kobayashi, H. Hashimoto and K. Takase, 2012.

Standing Assistance Control using a Physical Strength

of a Patient with Load Estimation. In Proc. of 21st

IEEE Int. Symp. on Robot and Human Interactive

Communication, pp.234-239.

D. Chugo, T. Yamada, S. Muramatsu, S. Yokota, and H.

Hashimoto, 2015. Assistive Robot for Standing with

Physical Activity Estimation based on Muscle

Arrangements of Human Legs. In Proc. of 12th Int.

Conf. on Informatics in Control, Automation and

Robotics, pp.35-43.

E. Maki, P. J. Holliday and A. K. Topper, 1991. Fear of

falling and postural performance in the elderly. In J. of

Gerontology, Vol.46, No.4, pp. 123-131.

K. Kamiya, 2005. Development and evaluation of life

support technology in nursing. In Proc. of 7th RACE

Symp., Research into Intelligent Artifacts for the

Generalization of Engineering, pp.116-121.

D. Chugo, E. Okada, K. Kawabata, H. Kaetsu, H. Asama,

N. Miyake, and K. Kosuge, 2006. Force Assistance

eference

ICINCO 2017 - 14th International Conference on Informatics in Control, Automation and Robotics

152

Control for Standing-Up Motion. In Proc. of the First

IEEE/RAS-EMBS Int. Conf. on Biomedical Robotics

and Biomechatronics, DOI: 10.1109/BIOROB.2006.

1639073.

S. Nuzik, R. Lamb, A. Vansant and S. Hirt, 1986. Sit-to-

Stand Movement Pattern, A kinematic Study. In

Physical Therapy, Vol.66, No.11, pp.1708-1713.

H. Okada, M. Ae, N. Fujii and Y. Morioka, 1996. Body

Segment Inertia Properties of Japanese Elderly. In

Biomechanisms, No.13, pp.125-139.

D. Chugo, S. Aburatani, T. Masushige, S. Muramatsu, S.

Yokota and H. Hashimoto, 2015. A Hand Movement

which shows the Intention of a Robotic Guide for Safe

Walking, In Proc. of the 24th Int. Symp. on Industrial

Electronics, pp.940-945.

Y. Hirata, A. Hara, and K. Kosuge, 2007. Motion Control

of Passive Intelligent Walker Using Servo Brakes. In

IEEE Trans. on Robotics, pp.981-990.

K. Omori, Y. Yamazaki, H. Yokoyama, U. Aoki, M.

Kasahara, K. Hiraki, 2001. The relationship between

strength in the lower extremity and the ability to stand

up from a chair in elderly inpatients. In Sogo

Rehabilitation. Vol.30, No.2, pp.167-171.

N. M. Fisher, D. R. Pendergast and E. C. Calkins, 1990.

Maximal Isometric Torque of Knee Extension as a

Function of Muscle Length in Subjects of Advancing

Age. In Arch. Phys. Med. Rehabil., Vol.71, No.10,

pp.729-734.

Development of a Standing Assistance Walker for a Patient with Low Level of Care

153