micromachines

Article
Assessment of Stickiness with Pressure Distribution
Sensor Using Offset Magnetic Force
Takayuki Kameoka 1, * , Akifumi Takahashi 1 , Vibol Yem 1 , Hiroyuki Kajimoto 1 ,
Kohei Matsumori 2 , Naoki Saito 2 and Naomi Arakawa 2
1 Department of informatics, The University of Electro-Communications, Tokyo 182-8585, Japan;
a.takahashi@kaji-lab.jp (A.T.); yem@kaji-lab.jp (V.Y.); kajimoto@kaji-lab.jp (H.K.)
2 Shiseido Global Innovation Center, Kanagawa 220-0011, Japan; kohei.matsumori@shiseido.com (K.M.);
naoki.saito@shiseido.com (N.S.); naomi.arakawa1@shiseido.com (N.A.)
* Correspondence: kameoka@kaji-lab.jp; Tel.: +81-42-443-5445




Received: 29 August 2019; Accepted: 27 September 2019; Published: 27 September 2019


Abstract: The quantification of stickiness experienced upon touching a sticky or adhesive substance
has attracted intense research attention, particularly for application to haptics, virtual reality, and
human–computer interactions. Here, we develop and evaluate a device that quantifies the feeling
of stickiness experienced upon touching an adhesive substance. Keeping in mind that a typical
pressure distribution sensor can only measure a pressing force, but not a tensile force, in our setup,
we apply an offset pressure to a pressure distribution sensor and measure the tensile force generated
by an adhesive substance as the difference from the offset pressure. We propose a method of using a
magnetic force to generate the offset pressure and develop a measuring device using a magnet that
attracts magnetic pin arrays and pin magnets; the feasibility of the method is verified with a first
prototype. We develop a second prototype that overcomes the noise problems of the first, arising from
the misalignment of the pins owing to the bending of the magnetic force lines at the sensor edges.
We also obtain measurement results for actual samples and standard viscosity liquids. Our findings
indicate the feasibility of our setup as a suitable device for measuring stickiness.

Keywords: haptics; measurement techniques; stickiness; sticky feeling

1. Introduction
Haptic perception has begun garnering increasing attention over the past few years, and in this
regard, several studies have examined the representation of the human skin sensation, particularly
in the fields of virtual reality and human–computer interactions. In reproducing a realistic feeling,
it has been found effective to measure changes in the skin condition, such as the skin deformation
distribution and contact area, in real-world situations and to reproduce this information. For example,
Levesque et al. [1] measured the horizontal displacement of the skin of a finger tracing a glass surface
in detail to capture information on how the skin deforms when in contact with an irregular shape.
Bicchi et al. [2] captured changes in the skin contact area for a finger touching a flexible object. Such
measurements are closely related to the technique used in tactile presentation [3]. The measurement of
the horizontal displacement of skin is related to the development of devices that present horizontal
displacement [4,5] and the measurement of the skin contact area has led to the development of devices
that can represent a feeling of flexing/flexibility via changing the contact area [6,7].
Against this backdrop, here, we focus on the distribution of skin deformation corresponding to
the feeling of “stickiness”. In this study, stickiness is defined as the feeling experienced when touching
an adhesive material such as glue, Nattō (which is a traditional Japanese food made from soybeans),
or honey. The feeling of sticky sensation on the surface is also expressed as a frictional resistance [8].

Micromachines 2019, 10, 652; doi:10.3390/mi10100652 www.mdpi.com/journal/micromachines

Micromachines 2019, 10, 652 2 of 17

Chen et al. [9]. investigated the correspondence between the measurement of physical properties
of texture surface and subjective evaluation of touch sensation, and also mentioned the stickiness.
However, we address stickiness that is related to the motion of the finger along surface normal, that is,
the sensation experienced after releasing a finger that has been pressed against a sticky material. Here,
we note that stickiness affects our impression of daily consumables such as lotion, creams, and so on.
Moreover, stickiness is known as one of the factors responsible for the feeling of “wetness” in fabric
perception [10,11]. Stickiness is often used as a general aversion sensation [12] and is a quality that is
attributed to a wide range of materials and products.
Though Liu et al. measured adhesive force by the MEMS device [13], the development of
the system corresponding to the adhesive force measurement on the skin is necessary in order to
evaluate the sense of human. It has been speculated that both proprioceptive and cutaneous sensations
contribute to stickiness; here, however, we mainly focus on cutaneous sensation. In this context,
Yamaoka et al. [14] derived the relationship between the contact area of an adhesive surface and
the temporal change in the pressing force, and found that there is large hysteresis in the contact
area. The authors further created a stickiness display based on this finding. However, because their
observations were limited to the change in the contact area, the detailed force distribution during the
period of stickiness was not clearly elucidated. Such detailed physical properties are often required to
be measured in the field of food [15,16]. For example, Dan et al. [17]. measured the bite force applied
to raw and cooked apples using a pressure distribution sensor sheet.
In previous reports, we have described the basic principle of a system that measures the force
distribution between adhesive substances and finger skin [18,19]. We used a pressure distribution
sensor sheet to measure the adhesive force in the form of pressure distribution. Here, it must be
noted that common pressure distribution sensors can measure only positive (compressive) pressure,
but not negative pressure. Therefore, we devised a method of inserting a pin matrix between the
skin and the pressure distribution sensor to apply a “preload” using the weight of the pins. With this
configuration, the adhesive force can be observed as a decrease in the offset force when the finger is
raised. In our previous studies, the weight of the pin was set to 0.8 g, and in the case of a highly sticky
specimen, the pin could float because of the stronger adhesive force. Therefore, it was necessary to
apply a stronger preload to the sensor to perform more stable measurements. The other issue in our
previous studies involved the sensor sheet. Pressure-sensitive rubber sensors are prone to undesired
current pathways and large hysteresis, both of which make it difficult to realize accurate sensing using
prototype systems.
In this paper, we present and evaluate a stickiness measurement device with a large measurement
range. When compared with our previous approaches, here, we used a load cell substrate with
independent sensing points and applied a more powerful preload using a magnetic force.

2. Measuring System

2.1. Pressure Distribution Sensor Using Load Cell

To measure a distributed adhesive force, we developed a pressure distribution sensor using
load cells. Figure 1 shows the load cell, schematic, and photo of sensor substrate. In our study,
we used HSFPAR003A load cells (Alps, Inc. Tokyo, Japan) to measure the pressure; this load cell
allows the measurement of forces up to 3.5 N. Sensing points are located at 2.54 mm intervals, which
correspond to the two-point discrimination threshold of the human fingertip [20]; this interval is thus
sufficient for measurement. One-unit board has 16 load cells. One load cell is selected and amplified
by analog multiplexer (ADG726BCPZ, Analog Devices, Norwood, MA, USA) and differential amplifier
(AD623ARMZ, Analog Devices, Norwood, MA, USA).
 Micromachines 2019, 10, 652 3 of 17
Micromachines 2019, 10, x FOR PEER REVIEW 3 of 17

(a) (b)

(c)

Figure
Figure1.1. (a)
(a)Load
Loadcellcellused
usedfor
forpressure
pressuresensing.
sensing.(b,c)
(b,c)Schematic
Schematicand
andphoto
photoofofsensor
sensorsubstrate
substrate
comprising
comprising4-by-4
4-by-4load
loadcells.
cells.

2.2.Offset
2.2. OffsetPressure
PressureGenerated
GeneratedbybyMagnetic
MagneticForce
Force
Figure2 2shows
Figure showsthe theoverall
overallmeasurement
measurementsystem. system.The Thedevice
deviceconsists
consistsofofananacrylic
acrylicpinpininsertion
insertion
plate,anan8 8××1212magnetic
plate, magneticpin pinarray,
array,three
threebase
basemagnets
magnets(50 (50mmmm××5050mm mm××1010mm), mm),six sixload
loadcellcell
substrates (each mounted with 4 × 4 load cells), and an acrylic
substrates (each mounted with 4 × 4 load cells), and an acrylic fixed pedestal. fixed pedestal.
InInthis
thisstudy,
study,to to apply
apply an an offset
offset preload
preload to tothe
theload
loadcell,
cell,aamagnet
magnetwas wasinstalled
installedunder
under thethe
load
loadcell
substrate
cell substrate andanda pin matrix
a pin made
matrix of pin
made of magnets
pin magnetswas inserted between
was inserted the point
between of skinofcontact
the point and the
skin contact
load cell substrate (Figure 2). Each pin of the magnet pin array was aligned
and the load cell substrate (Figure 2). Each pin of the magnet pin array was aligned to correspond to correspond to a single
to
a load
singlecell.
loadWith
cell.this
With configuration, because
this configuration, the magnet
because underunder
the magnet the substrate and the
the substrate pin
and magnet
the matrix
pin magnet
attractattract
matrix each other,
eachan offsetan
other, pressure (i.e., preload)
offset pressure (i.e., can be applied
preload) can beto the load cell.
applied to theThree
loadmagnets were
cell. Three
stacked to strengthen the magnetic field (which was about 430 mT at
magnets were stacked to strengthen the magnetic field (which was about 430 mT at the center). Thethe center). The magnetic field
of adjacent
magnetic fieldpins
of might
adjacent interfere with each
pins might other.
interfere withHowever, because
each other. the polarities
However, becauseofthe thepolarities
pin magnets of
are all set in the same direction, and the pin magnets are all at the same
the pin magnets are all set in the same direction, and the pin magnets are all at the same height and height and approximately
horizontal, thehorizontal,
approximately force generated by the
the force interference
generated is the
by the repulsive force
interference is thein repulsive
the tangential
forcedirection,
in the
which, indirection,
tangential principle,which,
has no in effect on the has
principle, normal force on
no effect measurement.
the normal forceThe load cells and magnet
measurement. The load pin
cells and magnet pin array were both positioned at 2.54 mm intervals. The pin magnet was 2 mm inin
array were both positioned at 2.54 mm intervals. The pin magnet was 2 mm in diameter and 10 mm
height, and
diameter andthe10 mmmagnetic forceand
in height, wasthe275magnetic
mT. The force
acrylicwasplate
275andmT.the
Thepedestal were and
acrylic plate made the with a laser
pedestal
cutter. The acrylic plate was chosen because it is easy to cut using the laser
were made with a laser cutter. The acrylic plate was chosen because it is easy to cut using the laser cutter. The cut surface
becomes
cutter. Theslightly conical,
cut surface and the
becomes contact
slightly area with
conical, andthethepin becomes
contact areasmaller,
with thereducing the friction.
pin becomes smaller,
reducingWe the
used six load cell substrates mounted on the base board, constituting 96 measurement points.
friction.
The voltage signal from a single load cell is selected by the analog multiplexer, amplified by the
differential amplifier, and then measured by an AD converter (MCP3208, Microchip Technology
Inc., Chandler, AZ, USA). All operation is conducted by a micro-controller (mbed LPC1768, NXP
Semiconductors N.V., Eindhoven, Netherland), and the data are sent to PC via USB serial port.
Measurements of all points were conducted 60 times per second, and simple moving averages were
calculated for noise removal (window size was 16).

(a) (b)
 approximately horizontal, the force generated by the interference is the repulsive force in the
tangential direction, which, in principle, has no effect on the normal force measurement. The load
cells and magnet pin array were both positioned at 2.54 mm intervals. The pin magnet was 2 mm in
diameter and 10 mm in height, and the magnetic force was 275 mT. The acrylic plate and the pedestal
were made with a laser cutter. The acrylic plate was chosen because it is easy to cut using the laser
Micromachines 2019, 10, 652 4 of 17
cutter. The cut surface becomes slightly conical, and the contact area with the pin becomes smaller,
reducing the friction.

Micromachines 2019, 10, x FOR PEER(a)

REVIEW (b) 4 of 17

(c) (d)

Figure
Figure2.2.Measurement
Measurement system
systembased
basedon on
magnet
magnet(a) structure, (b) oblique
(a) structure, view,view,
(b) oblique (c) base
(c) magnet, and
base magnet,
(d) circuit board of sensing device. The attractive force between the base magnet and the
and (d) circuit board of sensing device. The attractive force between the base magnet and the pin pin magnets
works
magnetsas works
a preload to the load
as a preload to thecell
loadatcell
theatsensing point,
the sensing enabling
point, measurement
enabling measurement of oftensile
tensileforce
force
distribution
distributionofofadhesive
adhesivematerial.
material.Six
Sixload
loadcell
cellsubstrates
substratesare
aremounted
mountedon onthe
thebase
baseboard,
board,constituting
constituting
96
96measurement
measurementpoints.
points.

We
Weused six load
calibrated cell
the substrates
load cell by mounted
adding a on the base
known board,
weight onconstituting 96 measurement
load cell. Figure 3 shows the points.
result
The voltage
of single signal
unit, from high
showing a single
linearity (R >
load cell 2 is 0.99)
selected
andby thethe
loadanalog multiplexer,
cell value was 10.5amplified
per 1 g. by
That theis,
differential
the measured amplifier,
weightand then gf
is 0.095 measured
(=0.94 mN) by an AD
per converter
value 1 of the(MCP3208,
load cell. Microchip
It also showsTechnology Inc,
a large offset
Chandler,
value (i.e., AZ, USA). value
the output All operation
was aboutis700 conducted
out of 4095byof athemicro-controller
resolution of MCP (mbed
3208 LPC1768, NXP
without weight)
Semiconductors
owing to magnetic N.V., Eindhoven,
force, Netherland),
which corresponds and the686
to around data
mNare sent to
preload, PC via
which USB serial
is sufficient forport.
daily
Measurements
tangible materials of all points
such were and
as foods conducted
cosmetics.60 times per4 shows
Figure second,the and simple moving
magnetic averages were
field distribution of the
calculated
magnet under for noise removal (window
the substrate, measured size
at was 16).
the surface of the base magnet at 6.25 mm intervals using
TM-801We calibrated
tesla meter the load cell byInc.,
(KANETEC, adding a known
Nagano, weight
Japan). on loadthe
Although cell.magnetic
Figure 3 force
showsofthetheresult
magnetof
single
underunit, showingvaries
the substrate high linearity
depending (R2on
> 0.99) and the loadposition,
the measurement cell value wewasnote 10.5
thatper 1 g. Thatpreload
a sufficient is, the
measured
ranging from weight686isto0.095
1323 gfmN(= can
0.94be
mN) per value
exerted at any1 cells.
of theInload cell.words,
other It alsothis
shows a large
system offset
cannot value
measure
(i.e., the output
adhesive value with
substances was about 700 force
adhesive out ofof4095
more ofthan
the resolution
686 mN atofa MCP single3208
pin. without
While eachweight) owing
pin receives
to magnetic
the differentforce, whichforce,
magnetic corresponds
this valueto is
around
treated 686asmN preload,
an offset andwhich is sufficient
the pressure changefor amount
daily tangible
can be
materials
measuredsuch as a as foods value
relative and cosmetics.
regardlessFigure
of the 4initial
shows the magnetic
offset value. field distribution of the magnet
under the substrate, measured at the surface of the base magnet at 6.25 mm intervals using TM-801
tesla meter（KANETEC, Inc. Nagano, Japan). Although the magnetic force of the magnet under the
substrate varies depending on the measurement position, we note that a sufficient preload ranging
from 686 to 1323 mN can be exerted at any cells. In other words, this system cannot measure adhesive
substances with adhesive force of more than 686 mN at a single pin. While each pin receives the
different magnetic force, this value is treated as an offset and the pressure change amount can be
measured as a relative value regardless of the initial offset value.
Micromachines 2019, 10, x FOR PEER REVIEW 5 of 17

Micromachines 2019, 10, 652 5 of 17

Micromachines 2019, 10, x FOR PEER REVIEW 5 of 17

Figure 3. Load cell calibration. We measured load cell value by applying offset pressure by the
magnetic Load
FigureFigure
3. force cell calibration.
and
3. Load arbitrary We measured
weight.
cell calibration. loadload
We measured cell value by applying
cell value offset
by applying pressure
offset by the
pressure magnetic
by the
force and arbitrary weight.
magnetic force and arbitrary weight.

Figure 4. Magnetic force map. The magnetic force was measured at intervals of 6.25 mm along the
Figurelength and widthforce
4. Magnetic directions
map.immediately at the
The magnetic surface
force wasof measured
the magnet.at intervals of 6.25 mm along the
Figure 4. Magnetic force map. The magnetic force was measured at intervals of 6.25 mm along the
length and width directions immediately at the surface of the magnet.
2.3. Preliminary
length and widthExperiment
directions immediately at the surface of the magnet.
2.3. Preliminary
In our Experiment
stickiness measurement experiments, the adhesive material of interest was applied to a
2.3. Preliminary
participant’s Experiment
fingertip, who then pressed the fingertip onto the surface of the pin array. When the to a
In our stickiness measurement experiments, the adhesive material of interest was applied
finger pressure
In our stickiness reached 5 N
measurement (summed force
experiments, over all pins), the
the adhesive participant was asked to release the to a
participant’s fingertip, who then pressed the fingertip onto thematerial
surfaceof ofinterest
the pin was
array.applied
When the
fingertip along the vertical direction. The lifting process was completed in around 1 s. In the study,
participant’s
finger pressure fingertip,
reachedwho 5 Nthen pressed
(summed theover
force fingertip ontothe
all pins), theparticipant
surface of thewaspin array.
asked When the
to release the
we used Nattō stirred for 100 times by chopsticks as an adhesive material and baby powder (Johnson
finger pressure
fingertip along reached
& Johnson, the 5 Ndirection.
Inc.vertical
(summed
Tokyo, Japan)
force
The overprocess
lifting
as a non-adhesive
all pins),Natto
material.was
thecompleted
participant inwas
is a fermented
askeds.toInrelease
around
food in 1Japan,
the
the study,
and
fingertip
we used whenalong
Natt theit vertical
ō stirred
mixed, direction.
for 100
becomes times by The
sticky. lifting as
chopsticks process was completed
an adhesive material andin around 1 s. In (Johnson
baby powder the study,&
we used Nattō
Johnson, stirredJapan)
Inc. Tokyo, for 100astimes by chopsticks
a non-adhesive as an adhesive
material. Natto is amaterial
fermented andfood
babyinpowder
Japan, and(Johnson
when
& Johnson,
mixed, Inc. Tokyo,
it becomes Japan) as a non-adhesive material. Natto is a fermented food in Japan, and
sticky.
when mixed, it becomes sticky.
2.4. Results and Discussion
Figures 5 and 6 show the measurement results for Nattō and baby powder, respectively. Although
we acquired 2D distributed data, representation by 3D graph was not easy to grasp and we chose
 Micromachines 2019, 10, x FOR PEER REVIEW 6 of 17

Micromachines 2019, 10, 652 6 of 17

2.4. Results and Discussion
Figures 5,6 show the measurement results for Nattō and baby powder, respectively. Although
to show the 2D2Dview,
we acquired longitudinal
distributed section of the
data, representation bycenter.
3D graph Aswas thenot
measurement
easy to grasppoints
and weare 8 by
chose to 12,
theshow the 2Daxis
horizontal view, oflongitudinal section of 1the
the graphs becomes to center.
12. The Asvertical
the measurement pointsforce
axis represents are 8(mN),
by 12, with
the a
horizontal
positive valueaxis of the graphs
meaning tensile becomes
force (i.e.,1 negative
to 12. Thepressure).
vertical axis
Inrepresents
the case offorce (mN),
Natto, with abegins
pressing positive
from
0.00value
s andmeaning tensile force begins
finger detachment (i.e., negative pressure).
after 2.67 In detached
s, totally the case ofatNatto,
4.01 s.pressing begins
In the case of from
baby 0.00 s
powder,
and finger
pressing beginsdetachment
from 0.00 s,begins
fingerafter 2.67 s, totally
detachment beginsdetached
at 1.67 s,atand4.01 s. In detached
totally the case of
at baby
2.67 s.powder,
We asked
thepressing begins
participant from 0.00
to release s, finger
his finger detachment
within 1 s, andbegins at 1.67confirmed
we visually s, and totally
thatdetached
the fingerathad
2.67separated
s. We
asked the participant
around this time. to release his finger within 1 s, and we visually confirmed that the finger had
separated around this time.

Figure 5. Change in pressure distribution for Nattō (2D view, longitudinal section of the center).
The vertical axis represents the force (in milli Newton). The horizontal axis represents the location
of the sensing point. The four-digit numbers in each graph show the frame numbers at the time of
measurement. Measurements were taken in 0.0165 s/frame (60 fps).
Micromachines 2019, 10, x FOR PEER REVIEW 7 of 17

Figure 5. Change in pressure distribution for Nattō (2D view, longitudinal section of the center). The
vertical axis represents the force (in milli Newton). The horizontal axis represents the location of the
sensing 2019,
Micromachines point.
10, The
652 four-digit numbers in each graph show the frame numbers at the time of7 of 17
measurement. Measurements were taken in 0.0165 s/frame (60 fps).

Figure 6. Change
Figure Change ininpressure
pressuredistribution
distributionforfor
baby powder
baby (2D (2D
powder view,view,
longitudinal sectionsection
longitudinal of the center).
of the
The vertical
center). axis represents
The vertical the force
axis represents the (in milli
force Newton).
(in milli The The
Newton). horizontal axisaxis
horizontal shows thethe
shows location of
location
thethe
of sensing point.
sensing point.The
Thefour-digit
four-digitnumbers
numbersin ineach
each graph
graph show
show the
the frame
frame numbers at the time time of
of
measurement. Measurements
measurement. Measurements were were taken
taken in
in 0.0165
0.0165 s/frame
s/frame (60 fps).

Upon comparing
Upon comparing the the two
two figures,
figures, wewe note
notethat
thatpeak
peaktensile
tensileforce
forceisisstronger
strongerininthe
thecase
caseofofNattō.
Nattō.
From 2.67 s, it can be observed that the tensile force is generated from the periphery
From 2.67 s, it can be observed that the tensile force is generated from the periphery of the surface of the surface
being pressed
being pressed and and gradually
gradually concentrated
concentrated at at the
the center.
center. InIn both
both cases,
cases, after
after the
the finger
finger was
was totally
totally
detached, we still observe remaining tensile force, which is considered
detached, we still observe remaining tensile force, which is considered as noise. as noise.
One possible
One possible explanation
explanation of of this
this offset
offset noise
noise is
is that
that the
the pin
pin magnets
magnets werewere aligned
aligned along
along the
the
magnetic lines
magnetic lines ofof the
the base
base magnet
magnet and and the
the lines
lines were
were not
not strictly
strictly vertical,
vertical, which
which gavegave rise
rise to
to the
the
interference between the pin and the pin insertion plate, which may have generated
interference between the pin and the pin insertion plate, which may have generated friction and thus friction and thus
noise (Figure
noise (Figure 7).7). This
This noise
noise can
can be
be regarded
regarded as as aa hysteresis
hysteresis ofof the
the measuring
measuring device.
device.
 Micromachines 2019, 10, x FOR PEER REVIEW 8 of 17
Micromachines 2019,
Micromachines
Micromachines 10,10,
2019,
2019, 652
10,xxFOR
FORPEER
PEERREVIEW
REVIEW 8 8of
8 of of17
17 17

Figure 7. Magnetic lines of force in the setup. The magnetic field lines of the base magnet have a
Figure 7. Magnetic
Figure lines of force in the setup. The magnetic field lines of the base magnet have a tangential
Figure 7.7.component,
tangential Magnetic
Magneticlines ofofforce
generating
lines ininthe
thesetup.
forcefrictional forceThe
setup. Themagnetic
between field
the pin
magnetic lines
linesof
magnet
field andthe
of base
the
the pin
basemagnet
insertion
magnet have a a
acrylic
have
component,
tangential generating
component, frictional
generating force between
frictional thebetween
force pin magnet
the and
pin the pin
magnet andinsertion
the pin acrylic plate.
insertion acrylic
plate.
tangential component, generating frictional force between the pin magnet and the pin insertion acrylic
plate.
3. Improvement
plate. of Measuring System
3. Improvement of Measuring System
3. To solve the problem
Improvement of the System
of Measuring previous prototype—in which the pin magnets were aligned along
3. Improvement of Measuring System
the baseToTo solve
magnet’sthe magnetic lines ofprevious
problem of the force and prototype—in
did not standwhich the pin
vertically, magnets
thus were causing
potentially aligned along friction
solve the problem of the previous prototype—in which the pin magnets were aligned along
the
andthe To
base solve
magnet’s
noise—we the
next problem
magnetic
devised of
lines
a theof previous
force
one-row and prototype—in
did not
pin-array stand
measuring which
vertically,the
device. pin
thus magnets
potentially
This devicewere aligned
causing
is based along
friction
on the
base magnet’s magnetic lines of force and did not stand vertically, thus potentially causing friction
the
and
principle base magnet’s
noise—we next
that whennext magnetic
devised
an infinitely lines
a of
one-rowforce and did not
pin-array stand
measuring vertically,
device. thus potentially
This device is causing
based friction
on the
and noise—we devised along rectangular
one-row magnet
pin-array is used,device.
measuring the magnetic forceisline
This device at itson
based center
the is
and noise—we
principle
vertical that when next
an devised
infinitelya longone-row pin-array
rectangular measuring
magnet isthe
used, device.
the This
magnetic device
force is based
line at on the
its center
principle that when an infinitely long rectangular magnet is used, the magnetic force line at its center on
(Figure 8). Consequently, in the improved device, pin magnets are installed vertically
principle that when an infinitely long in rectangular magnet
theisisvertical (Figure
centerline of the8).
vertical (Figure
Consequently,
rectangular
8). Consequently, magnet, the
in the
improved
which reduces
improved theisinterference
device,
device,
used,
the pin the
the pin magnets
magnetic
magnets
with areforce
the
are pinline
installed
installed
atvertically
insertionits center
vertically plate
onis the
vertical (Figureof8).
centerline Consequently,
the rectangular magnet, in the improved device,the
which reduces theinterference
pin magnetswith are installed vertically
the pin insertion
(Figure
on the9).centerline
Althoughofonly the one row can magnet,
rectangular be measured
whichusing this the
reduces configuration
interferencewhen with the
the object in contact
pin insertion
on the
plate centerline
(Figure of the rectangular
9). Although only one row magnet,
can be which reduces
measured usingthe thisinterference
configurationwith the the
when pin object
insertion
in
with plate
the(Figure
adhesive9). Although
substanceonly one row can be
is semicircular, themeasured usingofthis
distribution theconfiguration
adhesive force when the object inand
is concentric
plate (Figure
contact with the9).adhesive
Although only oneisrow
substance can be measured
semicircular, using this
the distribution ofconfiguration whenisthe
the adhesive force object in
concentric
contact withatthe
measurement one adhesive substance
row is thus is semicircular,
sufficient. As we usethe distribution
a human of theoradhesive
fingertip artificialforce
human is concentric
finger here,
contact
and with the adhesive
measurement at one row substance is semicircular,
is thus sufficient. As wethe
usedistribution
a human fingertip of the adhesive force
or artificial is concentric
human finger
and measurement at one row is thus sufficient. As we use a human
which can be considered semicircular, we consider a line measurement to be sufficient. fingertip or artificial human finger
and measurement
here, at one row is
which can be considered thus sufficient.
semicircular, we As we usea aline
consider human fingertip or
measurement toartificial human finger
be sufficient.
here, which can be considered semicircular, we consider a line measurement to be sufficient.
here, which can be considered semicircular, we consider a line measurement to be sufficient.

(a)
(a) (b)
(b)
(a) (b)
Figure
Figure 8. Magnetic field lines of square (a) and rectangular
rectangular magnets (b). (b). When an
aninfinitely
infinitelylong
long
Figure 8. 8. Magneticfield
Magnetic fieldlines
linesofofsquare (a) and rectangular magnets
magnets (b).When When an infinitely long
rectangular magnet
magnetis
Figure 8. Magnetic
rectangular used,
isfield the
the magnetic
used,lines of squareforce
magnetic (a) and its center
center is
line rectangular
at its is vertical.(b). When an infinitely long
magnets
vertical.
rectangular magnet is used, the magnetic force line at its center is vertical.
rectangular magnet is used, the magnetic force line at its center is vertical.

Figure 9. Pin magnet positioned at center and peripheral regions of rectangular magnet. The photographs
indicate that the pin can remain vertical because of vertical magnetic lines.
Micromachines 2019, 10, x FOR PEER REVIEW 9 of 17

Figure 9. Pin magnet positioned at center and peripheral regions of rectangular magnet. The
Micromachines 2019, 10, 652 9 of 17
photographs indicate that the pin can remain vertical because of vertical magnetic lines.

3.1. Measuring
Measuring System
System Using
Using Single-Axis
Single-Axis Robot
To apply
apply the theoffset
offsetpreload
preloadtoto a load cell,cell,
a load we installed a rectangular
we installed magnet
a rectangular such that
magnet suchthe center
that the
line
centerofline
the ofrectangular
the rectangularmagnet overlapped
magnet overlapped withwith
the the
sensor
sensor portion
portionofofthe
thesingle-row
single-row load-cell
substrate. As As aa result,
result, the
the pin
pin magnets
magnets werewere positioned
positioned along
along thethe vertical
verticallines
linesof ofthe
themagnetic
magneticforce.
force.
In addition,
addition, two rows of pin magnets were added so as to sandwich the measuring pins, although
actual measurement
measurementwas wasperformed
performedonly onlyforfor
thethe
pinpin magnets
magnets of the
of the center
center row.row.
TheseThese additional
additional rows
rows of pins push the central pins from both sides with magnetic repulsive force,
of pins push the central pins from both sides with magnetic repulsive force, making the central pins making the central
pins perpendicular
perpendicular to thetobase
thesurface
base surface
and thusandminimizing
thus minimizing
frictionfriction
between between
the pin the pin insertion
insertion plate andplate
the
and
pins.the
Thepins. Thedevice
actual actual is
device
shownis shown
in Figure in Figure
10. The10. The device
device consists consists of an acrylic
of an acrylic pin insertion
pin insertion plate,
plate, a 3magnet
a 3 × 16 × 16 magnet pin array,
pin array, two two
100 mm100 mm× 20× mm
20 mm × 7×mm
7 mm magnets,and
magnets, andaa11×× 16
16 load cell substrate.
substrate.
Two basemagnets
Two base magnetswere werestacked
stackedto to strengthen
strengthen thethe magnetic
magnetic field,field,
which which was about
was about 410 mT 410atmT at the
the center.
center. The load cell and its surrounding circuit components are the same as
The load cell and its surrounding circuit components are the same as in the previous setup, while the in the previous setup,
while thesubstrate
load cell load cellwas substrate was redesigned
redesigned to achieve 1to×achieve
16 load 1cell
× 16 load cell configuration.
configuration. Each pin of Each pin of the
the magnet pin
magnet
array was pinaligned
array was aligned to to
to correspond correspond
one load cell,to one
andload cell, and
a preload wasa applied
preload to wastheapplied to the load
load cells.
cells.The load cells were arranged at intervals of 2.54 mm, and the magnet pin array was arranged in
The load
the same cells Each
manner. were arranged
pin magnet at intervals
was 2 mm ofin
2.54 mm, and
diameter and the10magnet
mm inpin arrayand
height, wasthe
arranged
magnetic in
the same
force was manner. Each pin magnet
275 mT, corresponding to awas 2 mmofin
preload diameter
about 980 mN.and 10 mm in height, and the magnetic
forceForwascontact
275 mT, corresponding
with the adhesive to a preloadwe
substance, ofused
about 980 mN.
a hemispherical artificial human skin gel (with a
For contact
diameter of 5 cm,with the adhesive
manufactured bysubstance,
BEAULAXwe used a hemispherical
Corporation, Saitama, Japan)artificial
withhuman skinequivalent
elasticity gel (with
ato diameter of 5 cm, manufactured by BEAULAX Corporation, Saitama,
that of human skin as the contactor along with a single-axis robot (T4L manufactured by YAMAHA, Japan) with elasticity
equivalent to thattoofdepress
Shizuoka, Japan) human theskinartificial
as the contactor
“finger”. along with a single-axis robot (T4L manufactured
by YAMAHA, Shizuoka, Japan) to depress the artificial “finger”.

(a) (b)

Figure
Figure 10.
10. Single-column
Single-column pin-array
pin-array measuring
measuring device
device (a)
(a) overall
overall view,
view, (b)
(b) magnified
magnified view
view of
of the
the
measurement part.
measurement part.

Experiment 11
3.2. Experiment
Adhesive substances were uniformly applied to the upper part of the pin array array beforehand,
beforehand, and
the contactor was pressed against the upper surface of the pin array placed on the load cell. We We placed
placed
1 mm thick adhesive materials on each pin, with the procedure depicted in Figure 11. The temporal
pressure distribution was
change in the pressure was measured
measured when
when pulling
pulling apart
apart the
the contactor
contactor from
from the
thesurface.
surface.
The single-axis
single-axis robot
robotwas
wasused
usedfor
forpressing
pressingandandseparating
separatingthe
the “finger”, and the pushing distance
“finger”, and the pushing distance of
of the
the contactor
contactor waswas set2 to
set to mm 2 mm vertically
vertically downward
downward fromfrom the in
the state state in contact
contact with
with the pinthe pin while
array, array,
while
the the pulling-off
pulling-off distance
distance was setwas
to 2set
mm to vertically
2 mm vertically
upwardupward from
from the theinstate
state in contact
contact with
with the pinthe pin
array.
array. The moving speed of the contactor was 1 mm/s. Honey, toothpaste, shaving gel, and shampoo
Micromachines 2019, 10, x FOR PEER REVIEW 10 of 17

Micromachines
The moving2019, 10, of
speed 652the contactor was 1 mm/s. Honey, toothpaste, shaving gel, and shampoo10 of 17
were
prepared as adhesive substances. These adhesive substances are familiar to us, and we can find
difference
were preparedby touching them.substances.
as adhesive If this system canadhesive
These measuresubstances
the difference in adhesive
are familiar force
to us, andofwethese
can
substances, there is a possibility that the system can grasp the difference in the adhesive feeling
find difference by touching them. If this system can measure the difference in adhesive force of these felt
by human skin.
substances, The
there is asticky substance
possibility attached
that the systemtocan
the grasp
contactor and pin arrays
the difference in thewas removed
adhesive eachfelt
feeling timeby
before the next trial.
human skin. The sticky substance attached to the contactor and pin arrays was removed each time
before the next trial.

Figure 11. Procedure of pasting adhesive material on each pin. O 1 Place the insertion plate over the
pin magnet,
Figure O2 coat with
11. Procedure an adhesive
of pasting sample,
adhesive O
3 remove
material excess
on each  Place the
pin. adhesive samples, O
andplate
insertion 4 remove the
over the
pin magnet,  coat with an adhesive sample,  remove excess adhesive samples, and  remove the
insertion plate.
insertion plate.
3.3. Results and Discussion
3.3. Results and Discussionresults in the case of honey are shown in Figure 12. The horizontal axis
The measurement
represents the pin location, whereas the vertical axis represents the force. A positive value indicates
The measurement results in the case of honey are shown in Figure 12. The horizontal axis
a tensile force (i.e., negative pressure). From the figure, we note that pushing starts at 0 s, whereas
represents the pin location, whereas the vertical axis represents the force. A positive value indicates
pulling begins at around 3.00 s. A tensile (i.e., adhesive) force is observed at the edge of the contact
a tensile force (i.e., negative pressure). From the figure, we note that pushing starts at 0 s, whereas
surface. It can also be confirmed that the adhesive force transfers to the center as the contact surface
pulling begins at around 3.00 s. A tensile (i.e., adhesive) force is observed at the edge of the contact
area changes with the motion of the contactor. Overall, the noise was reduced from the previous
surface. It can also be confirmed that the adhesive force transfers to the center as the contact surface
prototype, and we now can clearly observe tensile force behavior.
area changes with the motion of the contactor. Overall, the noise was reduced from the previous
On the other hand, we still have some issues. When comparing the pressure distribution after
prototype, and we now can clearly observe tensile force behavior.
measurement (5.00 s) and before measurement (0 s), we observed a residual force in the positive
On the other hand, we still have some issues. When comparing the pressure distribution after
direction after the measurement. The reason for this hysteresis phenomenon is unclear (we have
measurement (5.00 s) and before measurement (0 s), we observed a residual force in the positive
confirmed that the load cell itself does not have observable hysteresis), but we presume that friction
direction after the measurement. The reason for this hysteresis phenomenon is unclear (we have
between the contact pin and the plate still existed. Further, a tensile force can be observed to the
confirmed that the load cell itself does not have observable hysteresis), but we presume that friction
leftmost section of the graph in the interval from 1.33 s in Figure 12, which should not have contacted
between the contact pin and the plate still existed. Further, a tensile force can be observed to the
the contactor and should be regarded as noise. As the rightmost and leftmost sensing points are not
leftmost section of the graph in the interval from 1.33 s in Figure 12, which should not have contacted
surrounded by other pin magnets, they experience a magnetic force from the neighboring pin magnets
the contactor and should be regarded as noise. As the rightmost and leftmost sensing points are not
to generate a repulsive force in the left and right directions and interfere with the pin insertion plate.
surrounded by other pin magnets, they experience a magnetic force from the neighboring pin
Figures 13–15 show the measurement results for toothpaste, shaving gel, and shampoo, respectively.
magnets to generate a repulsive force in the left and right directions and interfere with the pin
As in the case of honey, we were able to measure the change in adhesion. When the results were
insertion plate.
compared for each adhesive sample, it was found that in the case of toothpaste, the adhesive force
Figures 13,14 and 15 show the measurement results for toothpaste, shaving gel, and shampoo,
was the strongest and that the viscosity was high and the duration of the adhesive force was long,
respectively. As in the case of honey, we were able to measure the change in adhesion. When the
together with the fact that the adhesive force remained up to 5.0 s. The shaving gel and shampoo had
results were compared for each adhesive sample, it was found that in the case of toothpaste, the
weak adhesion.
adhesive force was the strongest and that the viscosity was high and the duration of the adhesive
force was long, together with the fact that the adhesive force remained up to 5.0 s. The shaving gel
and shampoo had weak adhesion.
Micromachines 2019, 10, 652 11 of 17
Micromachines 2019, 10, x FOR PEER REVIEW 11 of 17

Figure 12.
Figure 12. Change
Change in
in pressure
pressure distribution
distribution for
for honey.
honey. The
The vertical
vertical axis
axis represents
represents the
the force
force (in
(in milli
milli
newton). The horizontal plane represents the location of the sensing point.
newton). The horizontal plane represents the location of the sensing point.
Micromachines 2019, 10, 652 12 of 17
Micromachines 2019, 10, x FOR PEER REVIEW 12 of 17

13. Change
Figure 13. Changeininpressure
pressuredistribution forfor
distribution toothpaste. TheThe
toothpaste. vertical axisaxis
vertical represents the force
represents (in milli
the force (in
newton). The horizontal
milli newton). planeplane
The horizontal represents the location
represents of theof
the location sensing point.point.
the sensing
Micromachines 2019, 10, 652 13 of 17
Micromachines 2019, 10, x FOR PEER REVIEW 13 of 17

Figure
Figure 14.
14. Change
Changeininpressure
pressuredistribution
distributionfor
forshaving
shavinggel. TheThe
gel. vertical axisaxis
vertical represents the the
represents force (in
force
milli newton). The horizontal plane represents the location of the sensing point.
(in milli newton). The horizontal plane represents the location of the sensing point.
Micromachines 2019, 10, 652 14 of 17
Micromachines 2019, 10, x FOR PEER REVIEW 14 of 17

Figure 15. Change in pressure distribution for shampoo. The vertical axis represents the force (in milli
Figure 15. Change in pressure distribution for shampoo. The vertical axis represents the force (in milli
newton). The horizontal plane represents the location of the sensing point.
newton). The horizontal plane represents the location of the sensing point.

3.4. Experiment
Experiment 22
Subsequently,
Subsequently, ininorder
ordertotoverify
verifywhether
whetherthe
the difference
difference in in adhesive
adhesive force
force could
could be measured
be measured by
by this
this system,
system, we carried
we carried out measurements
out measurements using ausing
standarda standard
viscosityviscosity
liquid. Inliquid. In this experiment,
this experiment, kinematic
kinematic
viscosities viscosities (Centi-Stokes
(Centi-Stokes Visco Liquid,Visco Liquid,Inc.,
ASONE, ASONE,
Tokyo,Inc., Tokyo,
Japan) Japan)
of 500, 1000,of3000,
500, 1000,
5000, 3000, 5000,
and 10,000
and 10,000 (cSt) were measured. The moving speed and force of the contact object in
(cSt) were measured. The moving speed and force of the contact object in the measurement were set the measurement
were set to
to values values to
identical identical to the corresponding
the corresponding ones in experiment
ones in experiment 1. The peak1.adhesive
The peak adhesive
force duringforce
one
during one measurement
measurement was acquired was
10 acquired 10 times consecutively
times consecutively for each
for each sample, andsample,
the averageand value
the average value
was used as
was used as the measurement data
the measurement data of each sample. of each sample.

Results and
3.5. Results and Discussion
Discussion

Figure 16 shows the peak value of the adhesive force for each sample as acquired from the
As can
measured data. As can be
be observed
observed from
from the
the figure,
figure, there
there is an
an obvious
obvious correlation
correlation between the peak
values of the adhesive force and kinematic viscosity, and the difference in the adhesive force can be
suitably measured by means of our measuring device. The peak value of the adhesive force becomes
constant when the kinematic viscosity exceeds 5000 cSt.
Micromachines 2019, 10, 652 15 of 17

values of the adhesive force and kinematic viscosity, and the difference in the adhesive force can be
suitably measured
Micromachines 2019, 10, xby means
FOR PEER of our measuring device. The peak value of the adhesive force becomes
REVIEW 15 of 17
constant when the kinematic viscosity exceeds 5000 cSt.
The purpose
The purpose of of fabricating
fabricating the
the second
second prototype
prototype was was toto solve
solve the
the issues
issues of
of the
the first
first prototype,
prototype,
specifically
specifically the large observed noise possibly owing to friction between the pins and the plate
friction between the pins and the plate caused caused
by inclined
by inclined magnetic
magnetic lines.
lines. Our
Our results
results suggest
suggest that
that this
this noise
noise isis reduced
reduced with
with the
the new
new setup,
setup, and
and the
the
tensile force
tensile force distribution
distribution can be clearly observed. This This in
in turn
turn implies
implies that
that our
our speculation
speculation of of the
the cause
cause
of the
of the noise
noise was
was correct.

Figure 16. Plot of peak values of adhesive force (mN) and kinematic viscosity (cSt). The blue line
Figure 16.the
indicates Plot of peak
average values
peak value,ofwhereas
adhesive
theforce
gray(mN)
pointsand kinematic
indicate viscosity
the row (cSt).
data (for The blue
standard line
viscosity
indicates
liquids ofthe average
500, peak 5000,
1000, 3000, value,and
whereas
10,000the gray points indicate the row data (for standard viscosity
cSt).
liquids of 500, 1000, 3000, 5000, and 10,000 cSt).
4. Conclusions
4. Conclusion
In our study, we developed a measuring device for quantifying stickiness. A pressure distribution
sensorIn was
our used
study, we developed
to observe a measuring
the temporal change indevice for quantifying
the pressure distribution stickiness.
upon applyingA pressure
finger
distribution
pressure to asensor was usedmaterial.
given adhesive to observe Here,thewetemporal
note thatchange
a typicalin pressure
the pressure distribution
distribution sensor upon
can
applying
measure afinger
pressing pressure
force, butto not
a given
a tensile adhesive material.
force, and thus weHere,proposedwe note that a typical
and implemented pressure
a method of
distribution
measuring the sensor canforce
tensile measure a pressing
by applying force,pressure
an offset but not in a tensile
advance force,
to theand thusand
sensor we proposed
measuringand the
implemented a method
resulting difference uponoffinger
measuring
pressuretheapplication.
tensile force by applying an offset pressure in advance to
the sensor and measuring
Subsequently, the resulting
we developed difference
a pressure upon finger
distribution sensorpressure
boardapplication.
with built-in load cells and
Subsequently,
independent sensingwe developed
points for highlya pressure
accuratedistribution
measurements sensor
and board with built-in
implemented loadapplication
a preload cells and
independent
method usingsensinga magnetic points
force.forWehighly accurate measurements
next compared measurementsacquired and implemented
using a fingertipa preload
coated
application
with Nattō as method using and
an adhesive a magnetic
a fingertip force. We with
coated next baby
comparedpowder. measurements acquired using
The results confirmed a
that the
fingertip coated with Nattō as an adhesive and a fingertip coated with baby
tensile force generated by the adhesive substance was initially at the edge of the contact part, but moved powder. The results
confirmed that the
with the change tensile force
in contact generated
pressure, by the adhesive
and eventually converged substance was initially at the edge of the
to the center.
contact part, butowing
However, moved to with the changeofinthe
the inclination contact pressure,
magnetic lines and eventually
of force, there wasconverged to thebetween
interference center.
However,
the pin magnetowing
and the topin
the insertion
inclination of the
plate. magneticwe
Therefore, lines of force,athere
proposed linearwas interference
arrangement tobetween
simplify
the
the pin magnet
system and the apin
and provide insertion
stable plate.
vertical Therefore,
preload. we proposed
We subsequently a linear several
measured arrangement to adhesive
kinds of simplify
the system using
substances and provide a stable
a single-axis robotvertical preload.
that pressed We subsequently
an artificial finger ontomeasured
the deviceseveral
coated kinds
with theof
adhesive substances
sticky substance using aThe
of interest. single-axis robot thatthat
results indicated pressed
the newan setup
artificial
wasfinger
able toonto the device
measure coated
the adhesive
with
forcethe sticky substance
distribution of interest.We
more accurately. The resultsthat
believe indicated
our device that can
the benew setup
used was ablequantify
to suitably to measure the
the adhesive force distribution
stickiness of adhesive substances. more accurately. We believe that our device can be used to suitably
quantify
The the stickiness
current obvious of limitation
adhesive substances.
is that we did not fully eliminate noise. Our method also has some
innateThe current obvious
drawbacks, such as limitation
the fact that is itthat we did
cannot not fully
measure eliminate
adhesive noise. Ourthe
force exceeding method also has
offset pressure,
some innate drawbacks,
and excessively such as
low viscosity the fact that
substances suchitascannot
watermeasure
cannot be adhesive
treated force
because exceeding
they fallthefromoffset
the
pressure,
magnet pins.and Still,
excessively lowthat
we believe viscosity substances such
the measurement as water
of adhesive cannot
force be treated
distribution because
should they fall
contribute to
from the magnet pins. Still, we believe that the measurement of adhesive force distribution should
contribute to the study of stickiness, and comparing our measured data with a human’s subjective
tactile feeling is our next step.
Author Contributions: Conceptualization, T.K. and H.K.; methodology, T.K. and H.K.; software, T.K. and A.T.;
validation, K.M., N.S., and N.A.; formal analysis, T.K., A.T., and V.Y.; investigation, T.K.; resources, H.K., K.M.,
Micromachines 2019, 10, 652 16 of 17

the study of stickiness, and comparing our measured data with a human’s subjective tactile feeling is
our next step.

Author Contributions: Conceptualization, T.K. and H.K.; methodology, T.K. and H.K.; software, T.K. and A.T.;
validation, K.M., N.S., and N.A.; formal analysis, T.K., A.T., and V.Y.; investigation, T.K.; resources, H.K., K.M.,
N.S., and N.A.; data curation, T.K.; writing—original draft preparation, T.K.; writing—review and editing, T.K.
and H.K.; visualization, T.K.; supervision, H.K.; project administration, H.K.; funding acquisition, H.K.
Funding: This research was funded by JSPS Grant-in-Aid for Scientific Research, grant number 15H05923 (New
academic field research "Multiple quality sensing").
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Hayward, V.; Levesque, V. Experimental evidence of lateral skin strain during tactile exploration.
In Proceedings of the Eurohaptics, Dublin, Ireland, 6–9 July 2003.
2. Bicchi, A.; Scilingo, E.P.; and De Rossi, D. Haptic discrimination of softness in teleoperation: the role of the
contact area spread rate. IEEE Trans. Rob. Autom. 2000, 16, 496–504. [CrossRef]
3. Kimura, F.; Yamamoto, A.; Higuchi, T. Development of a contact width sensor for tactile tele-presentation
of softness. In Proceedings of the 18th IEEE International Symposium on Robot and Human Interactive
Communication, Toyama, Japan, 27 September–2 October 2009; pp. 34–39.
4. Levesque, V.; Jerome, P.; Hayward, V. Braille Display by Lateral Skin Deformation with the STReSS2 Tactile
Transducer. In Proceedings of the Second Joint EuroHaptics Conference and Symposium on Haptic Interfaces
for Virtual Environment and Teleoperator Systems (WHC’07), sukaba, Japan, 22–24 March 2007; pp. 115–120.
5. Schorr, S.B.; Okamura, A.M. Three-Dimensional Skin Deformation as Force Substitution: Wearable Device
Design and Performance During Haptic Exploration of Virtual Environments. IEEE Trans. Haptic 2017,
10, 418–430. [CrossRef] [PubMed]
6. Fujita, K.; Ikeda, Y. Remote haptic sharing of elastic soft objects. Proceedings of The First Joint Eurohaptics
Conference and Symposium on Haptic Interface for Virtual Environment and Teleoperator Systems, Pisa,
Italy, 18–20 March 2005.
7. Endo, T.; Kusakabe, A.; Kazuma, Y.; Kawasaki, H. Haptic Interface for Displaying Softness at Multiple
Fingers: Combining a Side-Faced-Type Multifingered Haptic Interface Robot and Improved Softness-Display
Devices. IEEE/ASME Trans. Mechatron. 2016, 21, 2343–2351. [CrossRef]
8. Bau, O.; Poupyrev, I.; Isrer, A.; Harrison, C. TeslaTouch: Electrovibration for Touch Surfaces. In Proceedings
of the 23nd annual ACM symposium on User interface software and technology, New York, NY, USA, 3–6
October 2010; pp. 283–292.
9. Chen, X.; Shao, F.; Barnes, C.; Childs, T.; Henson, B. Exploring Relationships between Touch Perception and
Surface Physical Properties. Int. J. Des. 2009, 3, 67–76.
10. Tiest, W.M.B.; Kosters, N.D.; Kappers, A.M.L.; Daanen, H.A.M. Haptic perception of wetness. Acta Psychol
2012, 141, 159–163. [CrossRef] [PubMed]
11. Filingeri, D.; Fournet, D.; Hodder, S.; Havenith, G. Why wet feels wet? A neurophysiological model of
human cutaneous wetness sensitivity. J. Neurophysiol. 2014, 112, 1457–1469. [CrossRef] [PubMed]
12. Fisher, T.H. What we touch, touches us: Materials, affects, and affordances. Des. Issue. 2004, 20, 20–31.
[CrossRef]
13. Liu, H.; Bhushan, B. Adhesion and friction studies of microelectromechanical systems/nanoelectromechanical
systems materials using a novel microtriboapparatus. J. Vac. Sci. Technol. A 2003, 21, 1528–1538. [CrossRef]
14. Yamaoka, M.; Yamamoto, A.; Higuchi, T. Basic Analysis of Stickiness Sensation for Tactile Displays.
In Computer Science; Springer: Berlin, Heidelberg, Germany, 2008; pp. 427–436.
15. Angioloni, A.; Collar, C. Bread crumb quality assessment: a plural physical approach. Eur. Food Res. Technol.
2009, 229, 21–30. [CrossRef]
16. Liu, Z.; Scanlon, M.G. Modelling Indentation of Bread Crumb by Finite Element Analysis. Biosystems Eng.
2003, 85, 477–484. [CrossRef]
17. Dan, H.; Watanabe, H.; Dan, I.; Kohyama, K. Effects of textural changes in cooked apples on the human bite,
and instrumental tests. J. Texture Stud. 2003, 34, 499–514. [CrossRef]
Micromachines 2019, 10, 652 17 of 17

18. Kameoka, T.; Takahashi, A.; Yem, V.; Hiroyuki, K. Measurement of Stickiness with Pressure Distribution
Sensor. In Haptic Interaction; Springer: Singapore, 2016; pp. 315–319.
19. Kameoka, T.; Takahashi, A.; Vibol, Y.; Kajimoto, H. Quantification of stickiness using a pressure distribution
sensor. In Proceedings of the 2017 IEEE World Haptics Conference (WHC), Munich, Germany, 6–9 June 2017;
pp. 617–622.
20. Sidney, W. Intensive and Extensive Aspects of Tactile Sensitivity as a Function of Body Part, Sex and Laterality;
Thomas International Publishing: New York, NY, USA, 1968; pp. 195–222.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).