Body Massage Robot Technology: A Review

Jignesh Nakrani
2015H141040H
ME G511 (Mechanism and Robotics)
ME Design Engineering
BITS Pilani, Hyderabad Campus

Abstract
The interest in a healthy and happy life for people has increased of late as the quality of life in
old age has become a issue. Moreover, the interest in health has led industrial companies to
concentrate on massage technology. Body massage robot technology is one of the biggest issues
in the robotics field, and many companies and institutes have sought to develop a massage robot
technology. The massage robot must have certain abilities that would allow it to help a person
mentally and physically. There is no common interface module, however, for body massage
robot platforms, and few robots that have healthcare abilities exist. Therefore, developing a
common interface technology and applying the technology to robot platforms can pave the way
for the development of a technology that could improve the physical and mental health of
humans. This paper introduces the authors research on a massage for different parts of body and
massage service robot platform. First, different body parts including head, feet, back etc. and oral
rehabilitation massage robot area. In conclusion, the intellectual body massage robot is expected
to promote a new field and to pave the way for a new product.

1. Introduction
Robots are needed for tasks and
operations that must be performed with high
speed and accuracy in various industrial
fields. In recent years, however, special
attention has been paid to the need for robots
that can assist human beings with their
health and welfare needs. Robots which can
perform these specialized tasks require
complex mechanisms, and the ability to
realize complex motions like those
performed by human beings. [3]

Therapeutic massage is considered

an effective physiological treatment for
human body. Many health benefits can be
achieved through therapeutic massages, such
as improved blood circulation, lowered
blood pressure, reduced heart rate, relief
from fatigue, relaxed muscle tone, etc. In
medical applications, such as clinical patient
care, therapeutic massages are usually
applied for the reduction of pain, stress,
tension and anxiety of patients [2].
Previously, welfare robots such as a
walk support robot, a food tray carrier, and a
meal assistance robot have been developed.

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

In addition, there have been a number of

studies on multi-fingered hand robots. A
multi-fingered hand can perform grasping
and manipulating motions to handle objects,
and can imitate the movements of a human
hand. [3]
The development of robot hands is
toward an anthropomorphic design with
multi-finger shape and the utilization of
force/torque sensors and tactile sensor array
to mimic the perception of human hands.
Currently, most multi-finger robot hands are
mainly
developed
for
dexterous
manipulation tasks and as prostheses. In
addition to these applications, the use of
multi-finger robot hand as a massage means
is a feasible application for human health
and welfare. Some research efforts on the
massage issue were presented in literature.
With adequate designs, the multi-finger
robot hand can perform kinds of massage
motions like a massage specialist does, such
as kneading, rubbing, tapping and
acupressure. [2]
Recently, the research in the field of
humanoid robots attracts a lot of attentions
from many researchers in which most of the
studies are focused on the multi-fingered
robot hand. A gripper method and an
impedance control are the outstanding
representative
methods
among
the
researches of humanoid robots. In addition,
in the field of rehabilitation robot, welfare
and human supporting robot have also been
reported. Most of the researches are focused
on theoretical problem, but there is only a
few researchers studied about the
application of humanoid robot such as multifingered robot hand. [4]

Examples include such robots

include Toyohashi University of Technology
(TUT), reported feed forward-type and
Neural Networks massage motion control by
off-line learning in the TUT Hand robot.
This research described how a two fingered
hand was applied, but that both the force and
position control were insufficient, because a
feedback controller was not included due to
the lack of a force sensor. Therefore, the
massage motion of this hand was too
limited. Because of this, researchers decided
to design a multi-fingered hand such as a
four-fingered hand. However, it took a great
deal of time to solve the inverse problem
presented by Neural Networks. [3]

2. Types of Massage Robots

2.1 ChairBot
The robotic massage chair that uses the
bio information of its human user for health
monitoring is called ChairBot. It estimates
the users body states and analyzes the
massage regions of the body using an
intelligent
decision-making
system.
ChairBot significantly helps a user
overcome fatigue by providing him or her
with massage services.

Figure (1) The description of ChairBot

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

ChairBot utilizes a variety of

technologies: the intellectual massage force
feedback mechanism; the massage control
technique, which applies the robot
mechanism and senses pulses for diagnosis;
an active seat for the users convenience;
and a sensor and interface technology for
healthcare robots (Figure 1). [1]
2.1.1 The Intellectual Massage Force
Feedback Mechanism
A block diagram shown in Figure 2
explains the massage force feedback system.
According to the users health state and
basic information, ChairBot performs its
function with the proper user position and
massage strength. It looks for the proper Nth
actuator and an axis controller and could
work efficiently based on the results of the
health management program (HMP). [1]

Figure (3) Block diagram of the massage control

technique

2.1.3. Active Seat Technology for Users

convenience
The aim of active seat technology is
to give its user an efficient massage. A
sensor was utilized to recognize the
somatotype shown in Figure 4. The sensor is
made up of a probe tip with an acceleration
sensor and a load cell. The important
massage points for ChairBot are the neck,
shoulder, and lumbar region. The hands are
massaged repetitively from up downwards
and from left to right, as real hands would.
First, ChairBot scans the whole body
of the user. Second, the seat automatically
adjusts to the proper size of the users
somatotype and then fastens in place. Lastly,
the stiffness of the back area of the body
could be graphically shown by analyzing the
digitized data regarding the spot on the body
that is suitable for acupuncture. [1]

Figure (2) Block diagram of the intellectual massage

force feedback mechanism

2.1.2 Massage Control Technique

The results obtained from the
intellectual controller present the users
massage pattern or taste. Lower-level
modules are made up of the local controller,
the position/velocity/pressure sensor, and
the biofeedback sensor. The feedback loop
remains firmly in place when the lower-level
modules firmly connect to the sequent work.

Figure (4) The block diagram of the active seat

technology

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

2.2 Pushing and rubbing massage Multifingered Robot Hand

In recent years, a lot of research on
multi-fingered robot hand has been studied.
A multi-fingered robot hand can perform the
grasping and the manipulating motion to
handle objects, and can imitate the
movement of a human hand. However, these
researches on multi-fingered robot hand
have not given the concrete applications.
Then, the authors have reported the research
on human support and healthcare application
of robotics and also multi-fingered robot
hand. [2]
In authors group, massage motion
control for human shoulder by teachingplayback control was developed. Feedback
controllers were corporate with feed forward
linearization compensation of robots
position and force. Based on this concept,
authors reported that position control was
used to control the robot fingertip position,
until it touches shoulder. After touching,
control system was switched from position
control to force one. However, the lateral
position is gradually shifted while doing
force control for the vertical direction of
human skin. Furthermore, in massage
control, exact control for both of position
and force is not necessary for the
disturbance such as finger touches the
human bone, for example. Then, robot
finger is hoped to have compliance like a
human finger. [4]
Therefore, in this paper, hybrid
impedance control comprised of positionbased and force-based impedance control is
proposed to establish the suitable massage

control system. By using impedance control

technique to control the massage motion, the
lateral position of the robot fingertip is
controlled by position-based impedance
control, while the fingertip force for the
vertical direction of human skin is controlled
by force-based impedance control. Hence,
the massage point in the lateral can be fixed
by position-based impedance control, while
achieving force control of vertical direction
by force based impedance controller. [2]
In this paper, two kinds of massage
motion control are studied, where one is the
pushing massage and another is the rubbing
massage as shown in Figure 5. In the
authors previous research, the pushing
massage motion control has only been done.
The rubbing massage motion control is the
first study all over the world in massage
field using maltifingered robot hand. [3]

(a) Pushing
(b) Rubbing
Figure (5) Patterns of massage motion

The schematic diagram of massage

system based on the multi-fingered
humanoid type robot hand is shown in
Figure 6. The robot hand has 4 fingers with
13 joints. The 1st finger has 4 joints and the
2nd to 4th fingers have 3 joints and they are
arranged like the human hands. [3]

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

Figure (6) Robot structure and massage system

construction

The actuator for the robot hand is a

small AC servomotor. The servomotor has
an integrated harmonic gear (1/80) and
encoder, and directly drives each joint. In
this paper, the robot hand is controlled using
the velocity mode of the servo driver.
Therefore, the input torque of the robot
hand is changed to the voltage by v = K.
Here, v[V] is the voltage supplied to the
motor of robot hand. The constant K is the
coefficient which changes the torque to the
voltage, and was obtained by system
identification. The calculation method of
joint drive torque is explained later. [8]
The fingertip force of robot can be
measured by fingertip type of 6-axis force
sensors. For the measurement of a massage
therapists fingertip force, at first, the sheet
type distribution pressure sensor, as shown
in the lower left corner of Figure 5, is put on
the hand of the massage therapist. Then, the
massage therapists fingertip force is
measured when the massage therapist
actually massages a humans body.
However, this distributed pressure sensor
can measure only the pressure force but can
not measure the direction of pressure force.
So, only one direction of force is considered
in this research. [8]

Figure (7) Robot structure and massage system

construction

2.2.1 Impedance control system

Consider a finger movements
performed by an expert massage therapist
almost consisted of pushing. Pushing
massage is done strongly by fingers and the
fingers are placed on the position of
massage. In actual massage by therapists,
the position of the fingertips does scarcely
change very much, while the force of the
fingertip is depended on the condition of
massage position. But in fact, the massage
object is a human body, so the massage
position moves gradually during the
massage by robots force control. By this
reason, it is difficult to massage by a
fingertip force control with keeping at the
same position by robot. [8]
To solve this problem for both of
force and position control simultaneously, a
hybrid impedance and force control is
proposed for the massage control system by
multi-fingered robot hand. By using this
control, the impedance control is used to fix
the position on the human skin against the
disturbance while executing the force
control on the human skin. Therefore, this
impedance control is suitable to control the
movement of massage. A diagram of

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

impedance control of robot hand is shown in

Figure 8. [8]

Figure (9) Conceptual diagram of impedance control

2.2.1.1 Position-based Impedance Control

Figure (8) Conceptual diagram of impedance control

The massage object is a human body,

so the massage position is often shifted
gradually during the massage by robots
force control. Thus, it is difficult to carry out
the force control with keeping at the same
position by finger-tip force control. [3]
To solve this problem for both of
force and position control simultaneously,
the hybrid impedance control is proposed for
the massage control system by multifingered robot hand. In this control method,
the position and force control are realized by
the hybrid position and force impedance
control. Furthermore, when the finger-tip
touches the obstacle like a bone, the
fingertip slip softly against the bone in order
not to give the damage to the bone like a soft
spring-damper. A conceptual diagram of
impedance control of robot hand is shown in
figure 9. [3]

How to make the fingertip position

of a robot hand follow a target position
using impedance control is described. The
mathematical formula of the desired
impedance of massage control system with
robot hand is shown in the below: [3]

Where x, and are respectively the

fingertip position, velocity and acceleration.
And, xd, d and d are the reference position,
velocity trajectory and acceleration until
robot finger touches the object. Then, Md,
Dd, Kd and f are the desired mass coefficient,
damper coefficient, spring coefficient and
the action contact force on the fingertip from
the environment, respectively.
A general dynamic equation of robot
manipulator is shown in the below: [3]

Where is the joint angle vector, M() is an

inertia matrix, (, ) is the Corriolis and
centrifugal term, is the joint drive torque, J
is a Jacobian matrix.

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

Thus, a position-based impedance

control input from Eq.(2) becomes

2.2.1.2 Force-based Impedance Control

It is possible to also make the fingertip force follow the target force, realizing the
desired impedance characteristics for the
machine.
The desired impedance of massage
force control system is given not by Eq. (1),
but by Eq.(4), and in stationary state, the
finger-tip force applied to objects agrees
with target force. [3]

the vertical direction (x-axis) and giving a

position reference to the horizontal direction
(y, z-axis), the rubbing massage motion,
one-type pattern of massage, can be also
realized.
The block diagram of impedance
control of the massage system used in this
paper is shown in Fig.4. Here, fx() is the
notations to transform the variables of and
into the position x and also transform the
velocity via a Jacobian matrix of robot
kinematics.
Reference position, reference force,
desired mass matrix coefficient, damper
coefficient and spring coefficient is given as
follows. [3]

Thus, a force-based impedance control input

becomes

2.2.1.3 Hybrid Impedance Control

Hybrid control comprised of
position-based impedance and force-based
impedance is described. Position-based
impedance control is used for the horizontal
direction (y, z axis) of human skin muscle,
and force-based impedance control is used
for the vertical direction (x-axis) of human
skin.
By performing the position-based
impedance control horizontally, a massage
point does not shift during a massage.
Furthermore, by giving a force reference to

Figure (10) Hybrid impedance control of massage

system

2.2.2 Experimental procedure

For estimating the stiffness of human
skin muscle, at first, fingertip massage force
of the therapist in one direction are read by
the sheet type sensor at the real time and it is

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

the input to a robot in order to estimate the

parameter of the various stiffness (softness
and hardness) of the object as shown in
Figure 11. The left hand side in Figure 11 is
the measurement point of soft one. On the
other hand, the measurement point of hard
one is shown in the right side of Figure 11.

(a) Soft Point

(b) Hard Point
Figure (11) Measurement point of human body

Reference input force was given

from the data of the expert fingertip force
executed by the therapist using the sheet
sensor with real time measurement and it is
shown in Figure 12 in the case of Figure
11(a). The amplitude of the given input
force was about 10 [N]. The estimation time
was carried out after 1 [s] from the start of
massage. The estimation parameters of
human skin muscle in case of the soft point
and the hard point are shown in Table 1. [4]

Table (1) Estimated parameter of human skin muscle

Figure 12 shows the comparison of

the position output response between the
estimated values calculated from the model
of Eq. (11) and the measured position of
robot fingertip in real experiments with the
same input force. Simulation results almost
agree with experimental results, and model
validity was demonstrated. [4]

Figure (12) Estimation and experimental results of

skin model

2.3 Feet Massage

Reflexology is a kind of assisting
physical care through massaging or applying
pressure to parts of the feet where will
reflect an image of the human body in order
to improve general health of the human.
Based on this principle, we propose a new
application of 7-DOF (degree of freedom)
redundant manipulator to do the massaging
work for human feet with the tactile sensor
equipped to the end-effector. To facilitate
the flexible and dexterous manipulation of
redundant
manipulator,
the
hybrid
impedance control is adopted and extended
to include the null space motion, not only to
generate a desired motion of the endeffector, but also manage the contact force
between the end-effector and the human
feet. In order to improve the safety and
comfortability, the reactive forces of the
tactile sensor cells are monitored and
recorded. [14]

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

Reflexology, having a very long

history in China, is pressure massage to the
feet or hands in order to stimulate the reflex
points and bring about a balance of the eight
bodily systems in order for the body to work
together in harmony and unison and thus
creating a feeling of well-being and
optimum health. The human body is
completely reflected in the feet in a threedimensional form, and a point on the foot
which maps to an area on the body is called
a reflex as shown in Figure 13. The benefits
of reflexology will be the reduction of stress,
relaxation, pain reduction, amelioration of
symptoms for health concerns, rejuvenation
of tired feet, improvement in blood flow,
etc.

experiments are made and several subjects

are discussed later. [14]
2.3.1 The analysis of reflex properties
As the reflexology theory indicates,
practitioners believe that the foot is divided
into a number of reflex zones corresponding
to many parts of the body, and the practice
of massaging or applying pressure to parts
of the feet will produce a beneficial effect on
corresponding parts of the body, such as
lung, shoulder, pancreas, kidneys and so on.
Pressure applied to the feet generates a
signal through the peripheral nervous
system. From there it enters the central
nervous system where it is processed in
various parts of the brain. It is then relayed
to the internal organs to allocate the
necessary adjustments in fuel and oxygen.
Finally a response is generated that is sent
on to the motoric system. This message is
fed forward to adjust the bodys tone or
overall tension level. [14]

Figure (13) The system setup of massaging based on

reflexology

In this
topic,
a redundant
manipulator with tactile sensor equipped
end-effector is applied to massage the
human feet as shown in Figure 13. Then the
reflex properties are analyzed. After that the
system modeling is discussed with the
introduction of tactile sensor and modeling
of human skin. After that the impedance
control is extended to control both the main
task and null space motion. Simulations and

Figure (14) System modeling and human skin model

The
massaging
gestures
of
practitioners always involve the pressing,
kneading, rubbing or tapping. In this topic,

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

the action of pressing will be performed by a

robotic arm with a series of forces exerted.
Massaging the feet reflexology by
the robotic arm, will not only save the labor
to do repeated work, but also guarantee the
steady and standard practice, which will
certainly be helpful to improve the quality of
massaging work.
2.3.2 Modeling of the system and skin
It is very difficult to determine skin
properties and to model its behavior, since
human skin is a non-homogeneous,
anisotropic, non-linear viscoelastic material
whose properties also vary with age and
individual, which is a very complex
biological structure to study. Therefore in
this paper, the variances of skin properties
are not considered and we consider only
different kinds of skin surfaces and different
reflex points in the experiments. [14]
2.3.2.1 System modeling
The massaging action needs both the
position control and the orientation control,
therefore a lower-DOF manipulator is not
sufficient.
The
7-DOF
redundant
manipulator is applied to massage the
human feet as the setup of system shown in
Figure 14. The tactile sensor is installed at
the tip of the endeffector in order to get the
pressure of contact points. The tactile sensor
is chosen as the module type of DSA 9210
with the shell-like shape, having a size of
50mm25mm, which provides a spatial
resolution of 3.4mm with 70 sensor cells.
The working principle of the tactile sensors
depends on an interface effect between

metal electrodes and a conductive polymer

covering the sensing electrodes, and the
resistance between a common electrode and
a sensor cell electrode is a function of the
applied load changing with time. The output
voltage of a resistive sensor cell in the
electrical circuit can be detected on realtime, and the pressure on it will be obtained
from the transducer characteristics. Then the
forces exerted on the sensor cell can be
achieved easily by transforming the sensor
pressure. [14]
2.3.2.2 Three contact types
In general, based on the shape of
human feet, three types of contact can be
divided considering different surfaces as
convex, flat and concave surfaces as shown
in Figure 15(a), Figure 15(c) and Figure
15(e) respectively. The experiments are
made on three different parts of feet to get
the performances of tactile sensor. The
corresponding pressures measured on the
contacted tactile sensor cells are shown in
Figure 15(b), Figure 15(d) and Figure 15(f)
respectively.
Comparing among the
experiment results, it is easily found that the
reactive forces of the sensor cells are
different although the same force at the end
of the end-effector is exerted. The biggest
reactive pressures of the contacted sensor
cells appear in the convex type as shown in
Figure 15(b), but the concave type has
relatively the smallest pressures of contacted
sensor cells in three contact types as shown
in Figure 15(f), since it has more contact
areas than others. As expected, the flat type
arrives at a compromise. [14]

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

10

2.3.2.3 Derivation of human skin elasticity

Skin can be divided into two separate
layers, the epidermis and dermis, and the
studies show that the dermis is primarily
made of collagen, elastic and reticulin fibers,

Figure (15) Three contact models of human skin

surface

which all contribute to the mechanical

behavior of skin, though the epidermis is
stiffer than the dermis. In a soft
environment, the human feet skin is
modeled as the mass-spring-damper system
in the control scheme and the contact model
between the end-effector with the tactile
sensor equipped and the human feet skin is
built up as shown in Figure 14, where Mm is
the inertia of the end-effector including the
tactile sensor, Kk and Dk are skin elasticity
and damping coefficients while Km and Dm
are stiffness and dumping parameters of
impedance control.

In the simplified model, the reaction

force Fe can be expressed as
(11)
Where Fe is the external force in the
dynamic model of the massage motion and it
can be obtained easily from the impedance
control discussed in the next section. Xi is an
initial position where Fe becomes zero.
Based on reflexology, the feet
massage topic represent a way of massaging
human feet by a 7-DOF redundant
manipulator with tactile sensor installed at
the end-effector. Since the human skin is
very complicated and difficult to determine
its properties, it is simply modeled as the
mass-spring-damper system. Impedance
control is extended to manage both the task
motion and the null space of the redundant
manipulator. Several key subjects are
discussed, such as the contact types, human
skin elasticity and so on. Simulations and
real experiments are made respectively to
track the desired massaging trajectory and
the results are compared while several
factors influencing the experimental
performances are considered and analyzed
carefully. [14]
2.4 Oral Rehabilitation Robot
The expected Japan's future elderly
society will undoubtedly increase the needs
of medical and dental treatment and
rehabilitation. Due to the increase need of
medical care for patients and elderly
persons, more trained experts and financial
resources are required for providing services
effectively. This causes a big problem for

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

11

the both government and hospitals.

Therefore, it becomes necessary the
introduction of advanced tools designed for
providing treatment to patients and elderly
persons. Such concept was based within the
framework
of
cooperation
between
engineers and medical workers. This
concept was based fundamentally in the
following two points: [7]
1. In the physiotherapy, treatment
procedures by means of robotic
systems enable to evaluate the
efficacy of the treatment objectively.
In contrast, it is difficult for
conventional manual treatments to
compare between the subjective
strength
of
stimulation
and
effectiveness.
Because
the
conventional manual manipulation is
based on the experience and intuition
of experts.
2. Develop
a
new
physiatrist
manipulation technique that is taking
advantages of robot system. Robotic
physiatric
manipulations
bring
solution to cope with the increased
demands of therapy by expert
physiatrician.
Based upon their experience on
developing dental robotics, they proposed
the development of a new rehabilitation
robot designed to provide massage of the
maxillofacial region.
The
estimated
prevalence of oral health problems such as
dry-mouth, temporomandibular joint (TMJ)
disorders and swallowing disorders in Japan
is over ten million persons in all. Massage
therapy to specifically target organ, such as

masticatory muscles and salivary glands, is

effective to reduce the symptoms from these
conditions. Therefore, they set the purpose
of present study to develop a robot system
that
provides massage therapy to
maxillofacial region. In this topic, the design
concept of the Waseda-Asahi OralRehabilitation Robot No. 1 (WAO-1), which
is composed by two 6-degree of freedom
(DOF) arms and plungers attached to the
end-effectors in order to provide effective
massage to patients. Furthermore, the details
of the control system and safety features of
WAO-1 are introduced. [9]
2.4.1 Maxillofacial massage therapy for dry
mouth
The disease of dry mouth, also
known as Xerostomia, is characterized by an
abnormal dryness of the mouth due to the
reduced capability of the patient to produce
saliva. Such condition can be provoked by
several reasons, such as diabetes mellitus,
kidney diseases, stress, drugs, etc. The main
symptoms are dehydration, pain in the oral
cavity, and the abnormalities on the
sensation of tasting. In severe cases, patients
complain about difficulties to swallow and
even to speaking. Nowadays, the
pharmacotherapy and physiotherapy are the
main procedures of treatment to patients for
stimulating the production of saliva. In
addition, dental treatment and physiotherapy
are also provided. [9]
For better understanding of the
maxillofacial massage, we mention about
the human anatomy. The saliva is produced
mainly from three sorts of glands:
sublingual, submandibular and parotid. In
general, the maxillofacial tissue massage is

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

12

provided to the parotid gland due to its great

contribution on the production of saliva
(Figure 16). The therapy treatment consists
of providing massage to the parotid gland
and its excretory ducts in order to stimulate
the salivary flow. The massage includes the
both petrissage and effleurage manipulation
techniques. The petrissage manipulation is
useful to stimulate the specifically target
organ directly. The effleurage manipulation
provides indirect stimulation by rubbing the
skin surface. [10]

Figure (16) Maxillofacial tissue massage consists on

providing stimuli around the parotid gland, musculus
temporalis and masseter

2.

3.

Mechanism: In order to perform

massage on both Zone A and Zone
B, the WAO-1 mechanism should
consist of two-arm manipulators
that are utilized to control the
position/orientation of plunger
devices (attached to the endeffector). Of course, it is desirable
that the driving mechanism of the
arms is located out from the
patient's visual field.
Applied Force: Using a head model
that we will describe later, we
measured the dynamic force on
patient's head during the massage of
maxillofacial region by skilled
physiatrician. Thus we defined a
maximum force of 4 [kgfl to
massage around patient face.
Therefore, WAO-1 designed to
apply
the
same
force
to
physiatricians. [9]

2.4.2 System description

2.4.2.1 Design Concept
In order to provide effective massage to
patients, the development of WAO-1 is
based
on
the
following
design
specifications:
1. Working area: As we described
previously, the target organ of the
massage therapy for dry mouth is
parotid gland. Similarly, temporal
and masseter muscle are deemed as
the target of massage for TMJ
disorders. Therefore, the working
area of WAO-1 should be designed
to include above anatomical
structures, as shown in Figure 17.

Figure(17). Musculus temporalis is defined as zone A

and parotid gland and masseter are defined as zone B.

2.4.2.2 System Description

The mechanism of WAO-1 consists of
two main parts: robot arms and the plunger
device. Robot arms consist of two 6-DOF
manipulators used to control the movement
of the plunger device attached to the endeffector of each manipulator. The plunger
device is the only part of the robot having
direct contact with the patient face (Figure

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

13

18a). In the following paragraphs, a more

detail description of each of the parts of
WAO- 1 will be provided:
1. Robot arms. Each of the arms of WAO1 is composed by a 6-DOF manipulator
(Figure 18b). Each arm consists of two
translational DOF on the base and four
rotational DOF at the each joint of the
arm. Due to this DOF arrangement, the
robot arms do not obstruct the patient's
field of view while providing massage.
Each translational DOF consists of a ball
screws and a DC servo motors which
assures high positioning accuracy. Each
rotational DOF consists of a harmonic
drive gears and a DC servo motors with
no backlash.

2. Plunger. A plunger which applies

force to the facial tissues is located at
the end-effector of each arm. We
have designed different kinds of
plungers by changing their shape,
material, and number of passive
degrees of freedom. Such plungers
are exchanged easily depending on
the needs of the kind of massage
(Figure 19).

Figure (19) several types of plungers

(a)

(b)
Figure (18) System configuration of WAO-l: a) WAO-l is
composed by two 6-DOF arms and plungers. The torque
limiters were attached in the arm middle part as safety
device; b) DOF configuration of one arm

3. Safety device. The WAO- 1 has been

designed to applied force to facial
tissues of patients. Therefore, we
require considering safety features
that are needed to avoid as much as
possible any possible risk to the
patient. For that purpose, we have
proposed the safety system shown in
Figure 20. As it can be observed; at
first software limiters and watch dog
timer are implemented into the
control software of the robot. In
particular, rotational speed and
motion positioning are limited within
safety range of operation. The
second safety feature has been
implemented by attaching limit
sensors at each DOF as well as fuses
to limit the current supplied to each
motor. The third one is implemented
through torque limiters. The torque

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

14

limiter has been placed in the middle

part of each arm. In the case that the
plunger load Fpr exceeds the
threshold Fth; even if the robot
system presents an unexpected
failure meanwhile applying force on
the patient's face, the arm will bend
and the load will be also released to
avoid any injury to the patient
(Figure 21). Finally, the fourth and
fifth safety features enables to the
doctor and the patient to stop the
operation of the robot at any time in
case of emergency. [7]

the torque limiter, b) when Fpr> Fth is found, the

applied force is disabled.

2.5 Head Massage

In this study they focus not on the
development of control algorithms, but on
how a massage given by simple robot is
perceived. This is even more important since
for certain types of massages that are being
perceived as more pleasurable when
performed by another actor. It is well
documented, for example, that you cannot
tickle yourself. Furukawa, Kajimoto and
Tachi developed the Kusuguri system to
overcome the problem of remote tickling by
using smart phones and vibrators. Their
study provides a first clue on robotic tickling
devices. [5]

Figure (20) Outline of safety system. WAO-1 is

equipped with different kind of safety features.
The safety devices are implemented by software,
electrical and mechanical means.

Figure (22) The Happy Head Trip massage device

(a)

(b)

Figure (21) Torque Limiter principles: a) when Fpr<

Fth is found, the applied force is transmitted rough

The massage device in this study

overcomes this practical problem. The
movements of a simple robot are not
necessarily well controlled and the head
massage device functions as spring between
the robots hand and the scalp of the
participants. Even if the movements of the

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

15

robot might be slightly rough, it does not

cause any discomfort for the users. A scalp
massage using the head massage device is
therefore a good starting point for testing the
effects of a simple robot giving a
massage.[6]
Tactile interaction with a robot might
also have an influence on the attitude users
have towards robots. Negative attitudes
toward robots can decrease perceptions of
anthropomorphism and closeness of the
relationship between the human and robot.
We therefore investigated if receiving a
massage from a robot might change the
attitudes of users towards robots.[5]
Based on the arguments above we
define the main research questions of this
study as: [13]
1) Do humans perceive a head massage
received from a human masseur as
more pleasurable than a head
massage performed by themselves?
2) Do humans perceive a head massage
received from a simple robot
differently in terms of pleasure than
a head massage done by themselves?
3) Do humans change their attitude
towards robots after they had a
physical experience with them?
2.5.1 Method
We performed a within subjects
experiment in which the independent
variable masseur was either the subject
him/herself (self) or the experimenter
(human) or the Nao robot (robot). A within
participants structure of the experiment was

the preferred choice to compensate for

individual differences in the appreciation of
head massages. Some participants might
have a more sensitive scalp than others. The
assignment of the participants to one of the
six possible sequences of conditions (shr,
srh, hsr, hrs, rsh, rhs) for self (s), human (h)
and robot (r) was counterbalanced.
2.5.1.1 Measurements
To measure the pleasure of the
massage experience they used the Massage
as Pleasant subscale of the Attitudes
Toward Massage (ATOM) scale. We only
had to change the tense of the questions,
since the original ATOM scale asked about
a general attitude, while we had to ask about
a specific prior experience. This scale
consisted of five questions on a five point
likert type scales in which 1 denoted
strongly disagree and 5 denoted strongly
agree. [5]
1) I liked to be massaged
2) Receiving massage is relaxing
3) Receiving the massage improved my
mood
4) Receiving the massage made me
nervous (reversed item)
5) I liked to be touched by the masseur
We took the average of the five items to
calculate the dependent variable pleasure. A
high value on the ATOM scale therefore
indicates a high level of pleasure. In
addition, we measured the participants view
on robots with the Negative Attitude
Towards Robots Scale (NARS) [16]. The
NARS scale consists of 14 items in three sub
scales that each are being rated on a five
point likert type scales in which 1 denoted

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

16

strongly disagree and 5 denoted strongly

agree. A high value on any of the three
NARS sub scales indicates a high negative
attitude towards robots. We also asked the
participants about their prior experience
with robots. In addition we video recorded
the participants to be able to analyze their
facial expressions during the massages. The
video camera was placed slightly aside in
front of the participants so that it could
record the facial expressions of the
participants. All participants were fully
aware of the presence of the video camera
and its recording function. They were to
specifically told that their facial expression
would be analyzed, but rather that the
overall session would be recorded.
2.5.1.2 Participants
They had 18 participants at the age
between 18 and 49 (mean 25.6) of which 7
were women and 11 were men. Most were
associated to the University of Canterbury,
New Zealand. Twelve participants never
interacted with a robot before, five reported
they had interacted 1-10 times, and one
reported that he interacted more than 10
times with a robot. The participants received
cookies as a reward for their effort. [6]

the NAO robot to the box and the box to the

table. The robot could therefore not lose
balance and fall onto the participants.
The robot held the massage device in
its left hand. The experimenter was seated
behind the participants to control the robot.
This prevented that the participants might
have been under the impression that the
experimenter is controlling the robot. The
experimenter, in fact, only started the robot
action sequence and supervised the motion.
The movements of the robot action sequence
was designed using the Choreograph
software prior to executing the experiment.
The experimenter would have aborted any
action that might have made the participants
feel uncomfortable. Fortunately, this was
never necessary. [6]
For the human condition, the
masseur stood behind the participants as
shown in Figure 22(a). For the self
condition, the participants were given the
massage device. The massage devices were
washed with soap and water after every
participant.
The
questionnaires
were
administered using an iPad. We used the
Qualtrics service to design, present and
collect the questionnaires. The participant
could remain in the chair throughout the
whole experiment. [6]

2.5.1.3 Setup
2.5.1.4 Process
The experiment took place in a 3x5
meter room. The participants were seated on
a chair. The robot was standing behind them
on a box that was placed on a table (see
figure 22(b)). This was the only position in
which the robot was able to reach the scalp
of the user. For security reasons we strapped

The experimenter welcomed the

participants and asked them to sit on a chair.
Once the participants were seated, the
experimenter provided an iPad that
contained the introduction and consent form.
When the participants agreed to the consent

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

17

form, the participants handed the iPad back

to the experimenter. The experimenter then
explained to the participant the order of the
conditions they will experience, and that
after each condition, they will be asked to
fill out a short survey. Once this has been
explained, the experimenter passed the iPad
back to the participants, and asked them to
fill out a demographic survey and the NARS
survey. The NARS survey consists of three
sub scales: Negative attitude toward
situations of interaction with robots (e.g. I
would feel nervous operating a robot in front
of other people), Negative attitude toward
social influence of robots (e.g. Something
bad might happen if robots developed into
living beings) and Negative attitude toward
emotions in interaction with robots (e.g. I
feel comforted being with robots that have
emotions). After the participants finished
filling out the two surveys, they handed the
iPad back to the experimenter.
The experimenter then provided to
the participants the head massage device so
that the participants may familiarize
themselves with the device. Once
familiarized, the participants handed the
head massager back to the experimenter.
The experimenter then instructed the
participants on which of the three conditions
the participants would be exposed to first.
For example, in the masseur, self,
robot ordering, the experimenter instructed
the participant that the masseur will give a
head massage to the participants for around
45 seconds. A stop watch was used to time

(a) Human

(b) Robot

(c) Self
Figure (22) The Masseurs Conditions

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

18

the session. This duration was selected to

avoid any strain on the arm. If the
participants had to hold up the arm much
longer, it might have caused some
discomfort. The masseur asked if the
participants are ready. If the participants
respond in the affirmative, the masseur then
began the massage. Once the massage was
completed, the experimenter provided the
iPad back to the participants and asked them
to fill out the first pleasure questionnaire.
After the question had been completed, the
experimenter took the iPad away from the
participants and instructed the participants
that it is now time for the self massage. [6]
2.5.2 Discussion of Head Massage
It is not surprising that the robot was
not able to give a head massage that was as
good as the one received from a human
masseur. The humans ability to sense and
act is largely unmatched. The superiority of
humans is not limited to this study. Would a
human control condition be included in most
HRI studies then we would most of the time
have to admit that, despite all our efforts in
developing robots, humans are still the
better butlers, companions, and conversation
partners. But we shall not despair over this
obvious
disadvantage.
The
robotic
development is still at its beginning and we
shall not dismiss technologies in their early
stages. Even Apples first iPad, the Apple
Newton, was hardly usable and it failed in
the market.

characteristics of each participant. They

were different in height, skull size, hair
length, and hair strength. Moreover, some
participants appeared to have a much more
sensitive scalp than others. [6, 11]

3. Conclusion
The development of the body
massage robot has the advantage of opening
up a new market through the development of
the machinery and equipment related to
health and advanced robotic technology.
Products from various combined fields are
manufactured, such as health support
services that provide entertainment, health
devices, and sport raining systems. The body
massage robot is an interdisciplinary area
that
includes
biotechnology,
nanotechnology, information technology and
robot technology. Besides, the intellectual
massage robot is expected to promote a new
business field.

The main reason for the robots

limited massage skills are its lack of a visual
or tactile feedback loop. The robot was not
able to adjust its massage to the
Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

19

References:
1.
2.

3.

4.
5.

6.

7.

8.

9.

10.

11.
12.

13.

14.

Sungil Yi, Dongju Moon, Yongju Yang, Kangeun Kim Healthcare Robot Technology Development The
International Federation of Automatic Control Seoul, Korea, July 6-11, 2008
Ren C. Luo and Chih C. Chang Electromyographic Evaluation of Therapeutic Massage Effect Using
Multi-finger Robot Hand 2011 IEEE International Conference on Robotics and Automation Shanghai
International Conference Center May 9-13, 2011, Shanghai, China
Panya Minyong, Takanori Miyoshi and Kazuhiko Terashima Expert Massage Motion Control by Multifingered Robot Hand Pmeedings of the 2003 IEEE/RSI InU. Conference on Intelligent Robots and
Systems Las Vegas. Nevada ' Oclaber 2003
Panya Minyong, Keisuke Mouri, Hideo Kitagawa, Takanori Miyoshi and Kazuhiko Terashima Hybrid
Impedance and Force Control for Massage System by Using Humanoid Multi-fingered Robot Hand
Ryan Walker and Christoph Bartneck The Pleasure of Receiving A Head Massage From A Robot The
22nd IEEE International Symposium on Robot and Human Interactive Communication Gyeongju, Korea,
August 26-29, 2013
Keisuke Mouri, Kazuhiko Terashima, Panya Minyong, Hideo Kitagawa and Takanori Miyoshi
Identification and Hybrid Impedance Control of Human Skin Muscle by Multi-fingered Robot Hand
IEEE/RSJ International Conference on Intelligent Robots and Systems San Diego, CA, USA, Oct 29 - Nov
2, 2007
Jorge Solis, Yuichi Obokawa, Hiroyuki Ishii, Hiroki Koga, Atsuo Takanishi and Akitoshi Katsumata
Development of Oral Rehabilitation Robot WAO-1R Designed to Provide Various Massage Techniques
2009 IEEE 11th International Conference on Rehabilitation Robotics Kyoto International Conference
Center, Japan, June 23-26, 2009
Ren C. Luo, Chih C. Chang, Vi-Wen Pemg Impedance Control on a Multi-fingered Robot Hand Based on
Analyzed Electromyographic Information for Massage Applications IEEE International Symposium on
Industrial Electronics (ISlE 2009) Seoul Olympic Parktel, Seoul, Korea July 5-8, 2009
Hiroki Koga, Yuichi Usuda, Masao Matsuno , Yu Ogura, Hiroyuki Ishii, Jorge Solis, Atsuo Takanishi,
Akitoshi Katsumata Development of Oral Rehabilitation Robot for Massage Therapy 6th International
Special Topic Conference on ITAB, 2007, Tokyo
Yuichi Obokawa, Jorge Solis, Hiroyuki Ishii, Hiroki Koga, Atsuo Takanishi, and Akitoshi Katsumata
Clinical Massage Therapy with the Oral-Rehabilitation Robot in Patients with Temporomandibular Joint
Disorders 9th International Conference on Information Technology and Applications in Biomedicine,
ITAB 2009, Larnaca, Cyprus, 5-7 November 2009
Masao Kume, Yoshitosi Morita, Yutaka Yamauchi, Hideaki Aoki, Makoto Yamada, Kazuyoshi Tsukamoto
Development of a Mechanotherapy Unit for Examining the Possibility of an Intelligent Massage Robot
Chul-goo Kang, Bong-ju Lee, Ik-xu Son, Ho-yeon Kim Design of a Percussive Massage Robot Tapping
Human Backs 16th IEEE International Conference on Robot & Human Interactive Communication August
26 - 29, 2007 / Jeju, Korea
Takeshi Ando, Maki Takeda, Tomomi Maruyama, Yuto Susuki, Toshinori Hirose, Soichiro Fujioka,
Osamu Mizuno, Kenji Yamada, Yuko Ohno and Honda Yukio Biosignal-based Relaxation Evaluation of
Head-care Robot 35th Annual International Conference of the IEEE EMBS Osaka, Japan, 3 - 7 July, 2013
Jingguo Wang and Yangmin Li Massaging Human Feet by a Redundant Manipulator equipped with a
Tactile Sensor 2010 IEEE/ASME International Conference on Advanced Intelligent Mechatronics
Montral, Canada, July 6-9, 2010

Body Massage Robot Technology: A Review | Jignesh Nakrani (2015H141040H)

20