-
Research ArticleHighly Selective Biomimetic Flexible
TactileSensor for Neuroprosthetics
Yue Li,1,2 Zhiguang Cao ,1 Tie Li ,1,2 Fuqin Sun,1 Yuanyuan
Bai,1 Qifeng Lu,1
Shuqi Wang ,1 Xianqing Yang,1 Manzhao Hao ,3 Ning Lan ,3 and
Ting Zhang 1,2,4
1i-Lab, Key Laboratory of Multifunctional Nanomaterials and
Smart Systems, Suzhou Institute of Nano-Tech andNano-Bionics
(SINANO), Chinese Academy of Sciences (CAS), 398 Ruoshui Road,
Suzhou 215123, China2School of Nano-Tech and Nano-Bionics,
University of Science and Technology of China, 96 Jinzhai Road,
Hefei,Anhui 230026, China3Laboratory of Neurorehabilitation
Engineering, School of Biomedical Engineering and Institute of
Medical Robotics, Shanghai JiaoTong University, 1954 Huashan Road,
Shanghai 20030, China4Center for Excellence in Brain Science and
Intelligence Technology, Chinese Academy of Sciences, Shanghai
200031, China
Correspondence should be addressed to Tie Li;
[email protected] and Ting Zhang; [email protected]
Received 12 May 2020; Accepted 2 August 2020; Published 24
August 2020
Copyright © 2020 Yue Li et al. Exclusive Licensee Science and
Technology Review Publishing House. Distributed under a
CreativeCommons Attribution License (CC BY 4.0).
Biomimetic flexible tactile sensors endow prosthetics with the
ability tomanipulate objects, similar to human hands. However, it
is stilla great challenge to selectively respond to static and
sliding friction forces, which is crucial tactile information
relevant to theperception of weight and slippage during grasps.
Here, inspired by the structure of fingerprints and the selective
response ofRuffini endings to friction forces, we developed a
biomimetic flexible capacitive sensor to selectively detect static
and slidingfriction forces. The sensor is designed as a novel
plane-parallel capacitor, in which silver nanowire–3D
polydimethylsiloxane(PDMS) electrodes are placed in a spiral
configuration and set perpendicular to the substrate. Silver
nanowires are uniformlydistributed on the surfaces of 3D
polydimethylsiloxane microcolumns, and silicon rubber (Ecoflex®)
acts as the dielectric material.The capacitance of the sensor
remains nearly constant under different applied normal forces but
increases with the static frictionforce and decreases when sliding
occurs. Furthermore, aiming at the slippage perception of
neuroprosthetics, a custom-designedsignal encoding circuit was
designed to transform the capacitance signal into a bionic pulsed
signal modulated by the appliedsliding friction force. Test results
demonstrate the great potential of the novel biomimetic flexible
sensors with directional anddynamic sensitivity of haptic force for
smart neuroprosthetics.
1. Introduction
Recently, flexible bionic sensors have attracted notableresearch
interest and have been envisioned as key technolo-gies for the
applications of neuroprosthetics [1–4], robotics[5–8], and
human-machine interactions [9–11]. Especiallyfor neuroprosthetic
systems, flexible sensors assembled onprosthetic hands provide
front-end sensory signals for subse-quent signal encoding,
transmission, and neural interfacing,resulting in the regeneration
of bionic tactile information[5, 12]. However, the abandonment rate
of artificial limbs ishigh, and the applications of robotic hands
are still not pop-ular at present. One of the reasons is that these
hands are notdexterous enough to manipulate objects in practice and
are
limited by unknown information such as slippage and
weightperception. Sliding friction force is a critical criterion
govern-ing whether hands can grab objects stably (Figure 1(a)),
andstatic friction force is crucial for people to estimate the
weightof an object in hand (Figure 1(b)). Therefore, to improve
theintelligent and manipulative levels of prosthetic and
robotichands, the detection of static and sliding friction forces
isnecessary for dexterous in-hand manipulation.
In the past several years, there have been many studies
onflexible force sensors with great performance [13, 14], someof
which have already surpassed the sensitivities of humanbeings [15,
16]. However, the detection of sliding and staticfriction forces
has been overlooked. Most of the reportedflexible force sensors are
sensitive to yet not selective of
AAASResearchVolume 2020, Article ID 8910692, 11
pageshttps://doi.org/10.34133/2020/8910692
https://orcid.org/0000-0002-6899-7818https://orcid.org/0000-0002-0851-2508https://orcid.org/0000-0003-4071-1598https://orcid.org/0000-0001-8744-7128https://orcid.org/0000-0001-6061-5419https://orcid.org/0000-0001-5008-2081https://doi.org/10.34133/2020/8910692
-
multitype force stimuli. To address the requirement of
aselective response to multitype forces, until recently, a
fewtechnologies, such as capacitive sensor arrays [5] and
tribo-electric nanogenerator (TENG) arrays [17], have beendesigned
to detect the shear force. Clementine et al. proposedcapacitive
sensor arrays based on the 3D hill structure, witheach hill
corresponding to 25 capacitor pixels on top of andaround the hill.
Due to the anisotropic deformation of the25 sensor pixels under
multidirectional pressure, the sensorarray can measure and
discriminate both normal and shearforces. With the following
control program for robot arms,capacitive sensor arrays provide
sensing feedback for con-trolling a robot arm in various tasks [5].
Ren et al. designeda triboelectric nanogenerator array with a
four-partitionedelectrode structure. The four electrodes of the
sensors reactnonuniformly under a shear force, resulting in the
ability todetect both normal and shear forces. However, these
sensorarrays still cannot distinguish between the specific static
fric-tion force and sliding friction force. In addition, the
forma-tion of the as-assembled array also increases the
complexity
of signal processing for practical applications. A specific
sen-sor that can detect and differentiate the normal force,
staticfriction force, and sliding friction force, to the best of
ourknowledge, has not been developed.
To regenerate the bionic tactile perception of
“smart”neuroprosthetics, a sensor design can learn from the
tactileperception mechanism of fingers (Figure 1(c)). Fingers
havesuperior sensitivity and multiple tactile sensation
abilitiescompared with other parts of the body, benefitting from
thecombined effects of fingerprint epidermal morphology andfour
types of mechanoreceptors interred in the dermis [18].The
morphology of fingerprints is an uneven spiral, whichis attributed
to the perception of texture and sensitivity[19]. Four types of
mechanoreceptors (Meissner corpuscles,Merkel cells, Ruffini
endings, and Pacinian corpuscles) areable to efficiently convert
various mechanical stimuli intophysiological spike signals (Figure
1(c)), and then, the actionpotential signals representing
information are transmitted tothe somatosensory cortex by the nerve
bundles in submillise-conds [20]. Mechanoreceptors distributed in
different regions
Fingerprint
mg
+
Mechanoreceptor
+
Slippage Weight(b)(a)
(c)
Perception
Sensor
Artificial limb
Human limb
Action potentials
Bionic
pulsed
signal
s
Fsliding Fstatic
mg
A/D analog
Figure 1: Schematic illustration of the biomimetic flexible
friction force sensor for dexterous neuroprosthetics. (a, b) The
significance of staticand sliding friction forces in daily life.
The perception of slippage allows people to grasp the target
objects more stably by adjusting thegripping strength without
visual aid. The perception of the static friction force is closely
related to estimating the weight of an object. Withthe perception
of slippage and weight, prosthetics become “smarter” and are able
to achieve some complex tasks without visual aid. (c)The bionic
mechanism of neuroprosthetics to regenerate the perception of
touch. For human touch, two key factors are fingerprints andfour
types of mechanoreceptors, which selectively respond to different
types of forces. We designed a fingerprint-structured
flexiblecapacitive sensor with custom-designed signal encoding
circuit to mimic Ruffini ending functions. The output bionic pulsed
signals can betransferred to the nerve tissue.
2 Research
-
of the skin on the hand have selective sensitivity to
differenttypes of forces. Particularly, Ruffini endings are located
inthe dermis, which has directional preferences [21–26].
Benefit-ing from the best response ability to skin stretch [22],
Ruffiniendings are sensitive to the shear forces containing the
staticand sliding friction forces during object manipulation
[23].The larger the shear force is, the higher the frequency
ofresponse spike signal will be [26].
Inspired by the responsive function of Ruffini endings
tofriction forces, we design and fabricate unique fingerprint-like
flexible capacitive sensors with selective sensitivity tononnormal
forces (static and sliding friction forces). Thecapacitance of the
sensor remains constant when normal forceis applied, increases when
static friction force is applied, anddecreases when sliding occurs.
Furthermore, we demonstratethat the specific response of the
sensors can be used for slidingdetection and object weight
recognition in robotic hands. Inaddition, the circuit for encoding
the biomimetic output iscustom-designed to resolve the signal
incompatibility betweenthe flexible sensors and the nervous system,
which is useful inthe transfer of sensing signals from
neuroprosthetics to ampu-tees via appropriate neural interfaces
[27, 28].
2. Results
2.1. Design and Characterization of Flexible Friction
ForceSensors. Typical capacitance sensors are usually designedwith
a plane-parallel capacitor structure, containing an inter-mediate
dielectric layer and two electrode plates: one on thetop and one on
the bottom. The classic equation (1) of theplane-parallel capacitor
is as follows:
C = εrε0Sd
, ð1Þ
where C is the capacitance, εr is the relative permittivity,
ε0is the permittivity of free space, S is the effective
overlap-ping area between the two capacitance plates, and d is
thevertical distance between the two plates. According to
thisstructure design, for flexible capacitive sensors, flexible
thinfilm electrodes at the top will deform elastically when
anexternal force is applied, which will lead to a decreased
ver-tical distance (d) and increased capacitance regardlesswhether
the direction of the force is perpendicular or paral-lel to the
sensor. Thus, it is difficult for these traditionalflexible
capacitive sensors to discriminate the different typesof force.
Inspired by the morphology of the fingerprint, wepropose a novel
flexible capacitor, the capacitance platesof which are spiral and
perpendicular to the substrate.The spiral is centrosymmetric, which
ensures the similarsensitivity to shear force from any direction in
plane(Figure S5). The sensor consists of silver nanowire (AgNW)–3D
polydimethylsiloxane (PDMS) electrodes andsilicon rubber (Ecoflex®)
dielectrics. The fabricationprocess is shown in Figure 2(a). To
balance the capacitanceand size of the sensors, spiral electrodes
with differentheight-width ratios were designed with a fixed
width(15μm), a fixed spacing (50μm) between two adjacent
electrodes, and different heights of 15μm, 25μm, and 35μm.First,
a silicon wafer mold with spiral grooves was fabricatedby plasma
etching. Second, Ag nanowires (30nm diameter,20μm length) were
spray-coated onto the silicon wafermold. Then, the Ag nanowires on
the top layer of the Sisubstrate were removed by scraping with the
inclined planeof a syringe, and subsequently, the PDMS mixture
wascoated onto the substrate and peeled off after beingcompletely
cured for 3 hours at 80°C. Through thereplication process, Ag
nanowires were embedded into thesurface of the 3D PDMS microcolumn
(Figure 2(b) i).Figure 2(b) ii is a crossview scanning electron
microscopy(SEM) image of PDMS demolded from the silicon
mold,showing column-type electrodes and Ag nanowires that
arehighlighted in green. In particular, the top-view SEM
image(Figure 2(b) iii) with the analysis of the energy
dispersivesystem (EDS) (Figure 2(b) iv) demonstrates that no
Agnanowires existed between the microcolumns, which
ensuredinsulation between the two electrodes. Finally, silicon
rubber(Ecoflex®) was employed as the dielectric to fill the
groovesbetween the microcolumns (Figure 2(c)). The
Ecoflex-AgNW-PDMS sandwich structure prevents the shedding
andoxidation of Ag nanowires. Microcolumn electrodes couldstill be
easily deformed under shear force due to the lowerYoung modulus
(0.13MPa) of Ecoflex [29] compared to thatof PDMS (3MPa) [30]. The
resulting sensors were flexible(Figure 2(d) i) and had the proper
size (2 cm × 2 cm) forprosthetic applications (Figure 2(d) ii).
According toEquation (1), the theoretical capacitance was
approximately11.27pF, which is in accordance with the
measuredcapacitance (12.41pF) of the as-assembled bionic
sensor,proving the validity of this capacitive microstructure
design(Figure 2(d) iii).
2.2. Response and Sensing Mechanism of the Flexible
FrictionForce Sensors. To investigate the capacitive response to
theapplied normal force, static friction force, and sliding
frictionforce, the testing apparatuses were set up, consisting of
aforce gauge and a computer-controlled moving stage(Figures 3(a)
i–3(c) i). The motion direction of the forcegauge has two degrees
of freedom: parallel and perpendicularrelative to the tested
sensors. For the normal force, pressurewas applied to the sensor
through the vertical movement ofthe force gauge (Figure 3(a) i).
For the sliding friction force(Figure 3(b) i), sliding was applied
through the horizontalmovement of the force gauge at a constant
speed. The forcegauge probe gradually contacts the sensor from one
sideand leaves it from the other side. According to the classic
fric-tion law,
Fsliding = μFN, ð2Þ
where Fsliding is the sliding friction force, μ is the
coefficient ofsliding friction, and FN is the normal force. In this
research, μwas fixed at 0.65 [31]. As FN changes, Fsliding
changes.
For the static friction force, the sensor was fixed on anoblique
plane of 45° (Figure 3(c) i). Therefore, the normalforce and the
static friction force applied to the sensor were
3Research
-
Silicon mold Si@Ag NWs Si@regional Ag NWs PDMS@Ag NWs PDMS@Ag
NWs@Ecoflex
Spray coating Scrape
Coating PDMSand peel off
Spin coating Ecoflex
(a)
100 𝜇m
3 𝜇m 50 𝜇m
A
B
Energy (keV)0 1 2 3 4 5
Inte
nsity
(a.u
.)
Ag
A
B
(iii)
(ii)(i)
(iv)
(b)
100 𝜇m 100 𝜇m
(i) (ii)
(c)
Figure 2: Continued.
4 Research
-
equal during the vertical movement of the force gauge.According
to the force analysis in
Ftotal =ffiffiffi
2p ∗
Fstatic, ð3Þ
Ftotal is the value displayed on the force gauge, and Fstatic
isthe applied static friction force.
The sensitivity of a capacitive sensor is defined as
ΔCC0
= C − C0C0
, ð4Þ
where C and C0 are the measured capacitance and the
initialcapacitance before applying force, respectively. Figures
3(a)ii–3(c) ii show the real-time response curves under the
normal, static friction, and sliding friction forces(height :
width = 7 : 3), respectively. As shown in Figure 3(a)ii, the
capacitance remained nearly constant when an 11.2N(119.1kPa) normal
force was applied to the sensor. Duringthe sliding of the force
gauge across the sensor(Fsliding = 3:8N ð67:6 kPaÞ), the
capacitance decreased initiallyand then returned to its original
value when the force gaugeleft the sensor (Figure 3(b) ii). In
contrast, when an 11.2N(119.1kPa) static friction force was
applied, the capacitanceincreased immediately (Figure 3(c) ii).
These results revealthat the type of force can be readily
distinguished accordingto the capacitive signal change due to the
selective response.
Moreover, the height-width ratio is a key parameter ofthe
sensitivity. Structures were designed with the same elec-trode
distance and three different aspect ratios of 3 : 3, 5 : 3,
(i)
(ii)
(iii)
0 1 2 3 4 5
0
5
10
15
Time (s)
C =d
= 11.27 pF
Capa
cita
nce (
pF)
𝜀r𝜀0S
(d)
Figure 2: Fabrication and characterization of flexible shear
force sensors. (a) A schematic illustration of the sensor
fabrication process. (b)Scanning electron microscopy (SEM) images
of the PDMS@Ag NWs. (i) SEM image of Ag NWs on PDMS. (ii) Side-view
SEM imageshowing that Ag NWs were uniformly distributed on the top
and sidewalls of the spiral microcolumns. (iii) Top-view SEM image.
(iv)Energy spectra of the spiral column and the substrate. No Ag
NWs existed between the microcolumns. (c) SEM images of
PDMS@AgNWs@Ecoflex. (i) Top-view SEM image. (ii) Side-view SEM
image showing that Ecoflex filled the air gap evenly. (d) (i)
Photographshowing the flexibility of the sensor. (ii) Photograph of
a fabricated sensor. (iii) The actually measured capacitance of the
fabricatedsensor, which is in accordance with the theoretical
capacitance of approximately 11.27 pF.
5Research
-
0 1 2 3 4 5 6–30
–20
–10
0
10
20
30
Time (s)
Force gauge
Sensor
0 5 10 15 20–30
–20
–10
0
10
20
30
Normal force (N)
𝛥C
/C0 (
%)
(ii)
(iii)
(i) Normal force
(iv)A
B
0
1
2
3
4
5
6
7
810–6
Direction of motion
𝛥C
/C0 (
%)
h:w = 3:3h:w = 5:3h:w = 7:3
(a)
0 3 6 9 12–40
–30
–20
–10
0
10
Time (s)
(iii)
(ii)
(i)
(iv)
0 2 4 6 8 10–80
–60
–40
–20
0
20
𝛥C
/C0 (
%)
Sliding friction force (N)
Sliding friction force
Direction of motion
Force gauge
Sensor
A
B
h:w = 3:3h:w = 5:3h:w = 7:3
0
2
4
6
8
10
12
14
10–6
𝛥C
/C0 (
%)
(b)
0 1 2 3 4 5
0
2
4
6
8
10
Time (s)
0 5 10 15 20–20
–10
0
10
20
30
40
Static friction force (N)𝛥C
/C0 (
%)
𝛥C
/C0 (%
)
(iii)
(ii)
(i)
y = A2 + (A1-A2)/(1 + exp (x-x0)/dx))
Static friction force
Direction of motion
Force gauge
Sensor
45°
Ftotal
FstaticFN
(iv) A
B
h:w = 3:3h:w = 5:3h:w = 7:3
0
1
2
3
4
5
6
7
8
(c)
Figure 3: Response ability of the flexible shear force sensors
and its corresponding mechanism. (a) Normal force: (i) experimental
setup. (ii)The response curve for an 11.2N (199.1 kPa) normal
force. The capacitance was kept basically constant. (iii) The
different height-width ratioresponses of the sensor to the 0-20N
(0-355 kPa) normal force. (iv) Finite element analysis (FEA)
showing that the structure remains stableunder 14 kPa of pressure.
(A) The deformation of the film containing electrodes and
dielectric. (B) The deformation of electrodes specifically.(b)
Sliding friction force: (i) experimental setup, in which the force
gauge moves parallel to the sensor. (ii) The response curve for a
3.8N(67.6 kPa) sliding friction force. The capacitance decreased
initially and returned to its original value when the force gauge
left the sensor.(iii) The different height-width ratio responses of
the sensor to the 0-10N (0-178 kPa) sliding friction force. The
capacitance variationsfollow the fitting formula ΔC/C0 = A ∗
e−Fsliding/B + C. (iv) Finite element analysis showing that the
electrodes were laterally bent when a7 kPa sliding friction force
and 14 kPa pressure were applied. (A) The deformation of the film
containing electrodes and dielectric. (B) Thedeformation of
electrodes specifically. (c) Static friction force: (i)
experimental setup, in which the force gauge moves vertically and
thesensor is fixed on the oblique plane of 45°. (ii) The response
curve for an 11.2N (199.1 kPa) static friction force. The
capacitance increasedimmediately. (iii) The different height-width
ratio responses of the sensor to the 0-20N (0-355 kPa) static
friction force. The capacitancevariations follow the fitting
formula ΔC/C0 = a − b/ð1 + eðFstatic−cÞFstatic−c/dÞ. (iv) Finite
element analysis (FEA) showing that the electrodeswere laterally
stretched when a 14 kPa static friction force and 14 kPa pressure
were applied. (A) The deformation of the film containingelectrodes
and dielectric. (B) The deformation of electrodes specifically.
6 Research
-
and 7 : 3. As shown in Figure 3(a) iii, the capacitanceremained
nearly constant when applying the normal force(0-20N) (0-355 kPa)
regardless of the change in the height-width ratio. The capacitance
decreased with increasingsliding friction force (Figure 3(b) iii)
following the fitting for-mula ΔC/C0 = A ∗ e−Fsliding/B + C. A, B,
and C are constantterms, as shown in Table 1. This demonstrates
that the higherthe height-width ratio is, the higher the
sensitivity of the sen-sors to the sliding friction force (Figure
3(b) iii). In contrast,the capacitance increased with increasing
static friction force(Figure 3(c) iii). The capacitance variations
followed thefitting formula ΔC/C0 = a − b/ð1 +
eðFstatic−cÞFstatic−c/dÞ. a, b, c,and d are constant terms, as
shown in Table 1. The lowerthe height-width ratio is, the higher
the sensitivity of thesensors to the static friction force (Figure
3(c) iii).
To elucidate the underlying mechanism, deformations ofthe
electrodes were investigated by performing finite elementanalysis
(FEA). Different from interior voids (i.e., a foamstructure), the
bulk structure with the Ecoflex-filled micro-grooves exhibited low
structural compressibility, resultingin insensitivity to the normal
force, as exemplified by theFEA results in Figure S1. Figures 3(a)
iv–3(c) iv show thedeformations of the electrodes via FEA when the
normalforce, sliding friction force, and static friction force
wereapplied. The forces applied through the top silica glass andthe
bottom of the sensors were fixed during the analysis.When only the
normal force (11.4 Pa) was applied, noobvious deformation in the
structure of the electrodes wasobserved (Figure 3(a) iv), leading
to a constant capacitance.Under the sliding friction force (7 kPa)
and normal force(14 kPa), the electrodes are laterally bent but not
stretched(Figure 3(b) iv). Although d remains unchanged, a
decreasein S caused by the lateral bending of the electrodes leads
tothe decreased capacitance during sliding. Conversely, whenthe
static friction force (14 kPa) and normal force (14 kPa)are
applied, there is no relative displacement at the contactsurface,
and the electrodes are laterally stretched (Figure 3(c)iv). The
increased capacitance is caused by the increased Sand the decreased
d. The detailed analyzation of changes inS and d is listed in
supplementary materials (Figure S2).
2.3. Slippage Detection and Weight Perception of
FlexibleFriction Force Sensors. Benefiting from the ability of
flexiblebiomimetic sensors to selectively respond to static and
slidingfriction forces, we set up the corresponding prosthetic
dem-onstration scenarios (Figure 4). First, to demonstrate the
ability to detect the weight, a robotic hand mounted with
aRuffini-ending-inspired sensor was set to grasp a plasticbottle
with a constant force (Figure 4(a)). Then, water wasgradually
poured into the bottle by a graduated cylinder toincrease the
static friction force. The static friction forcewas the only
variable in this process. As shown inFigure 4(b), the capacitance
of the sensor increased as the12mL of water was poured in real time
(movie S1), and thecapacitance remained at the steady state after
the pouringof water was stopped. This action was repeated three
timesto demonstrate the ability of weight perception.
Then, the slippage detection ability was demonstrated byusing a
glove equipped with a Ruffini-ending-inspired sen-sor, which was
put on the hand to perform the sliding action(Figure 4(c)). During
sliding, the friction force was exertedon the sensor. The sliding
action was repeated three times.As movie S2 shows, during the
sliding process, the normalforce immediately increased to 0.5N in
the early stage andthen decreased gradually. The corresponding
capacitancesubstantially decreased by 40% once slippage occurred
andthen returned to the original value immediately when theslippage
stopped (Figure 4(d)). These results demonstratethe slippage
detection ability of flexible friction force sensors.
Finally, the insensitivity to the normal force is demon-strated
in Figure 4(e) and movie S3, where standard weightsof 20 g, 50 g,
and 100 g were put on the Ruffini-ending-inspired sensor, in that
order. The corresponding capacitiveresponses are shown in Figure
4(f). When the weights wereput down or taken away, two spikes were
generated due tothe static friction force caused by the nonvertical
motion.Overall, the capacitance that changes under these
differentweights were less than 1%, showing that the sensor is
insen-sitive to the normal force.
2.4. The Bionic Behavior of Flexible Friction Force Sensors.The
ability to distinguish slippage is still lacking for
neuro-prosthetics, limiting dexterous manipulation. With the
per-ception of slippage, humans can naturally estimate whetheran
object is grasped without visual aids, which is meaningfulfor daily
life. The expression of tactile information is anaction potential
based on biologically driven models in thecentral nervous system,
the frequency of which conveys tac-tile sensation to the
somatosensory cortex. In addition, thefrequency of spikes increases
with increasing applied force[26, 32, 33]. However, for
neuroprosthetics, slippage detec-tion at the sensor level is not
sufficient due to the absenceof proprioception. To regenerate the
slippage perception ofneuroprosthetics (Figure 5(a)), a signal
encoding circuit wasdesigned for bionic stimulus response signals,
which ishopeful to transmit analog signals from the sensors into
thenerve tissue.
To achieve the above expected functions, a signal encod-ing
circuit was assembled specifically for this Ruffini-ending-inspired
sensor, as shown in Figure 5(b). First, the recordedcapacitance
signal was transformed into a sine signal throughaWien bridge
oscillation circuit, which can be easily adjustedwithin a wide
frequency range. The oscillating circuit is com-posed of a
capacitance sensor (C), a matching capacitor (C1),and two matching
resistors (R1 and R2), forming an RC
Table 1: The fitting formula for the sensitivities to the
sliding andstatic friction forces.
Height-width ratioΔC/C0 = A ∗e−Fsliding
B + CΔC/C0 = a − b1 + e Fstatic−cð Þ/d
A B C a b c d
3 : 3 15.0 5.31 -14.87 20.48 21.76 6.36 2.12
5 : 3 73.9 21.09 -74.71 13.05 13.44 5.77 1.45
7 : 3 52.9 4.08 -51.15 7.89 8.56 3.89 2.10
7Research
-
series-parallel network. When the positive and negative
feed-back of the operational amplifier circuit is in equilibrium,
theoscillation can continue. At this time, the output
waveformfrequency follows formula
f = 12
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiR1R2CC1p ,
ð5Þ
where f is the frequency of the sine signal and R1 and R2 arethe
matching resistances with known resistance values(R1 = R2 = R). C1
is the matching capacitor with a fixedcapacitance, and C is the
capacitance of the sensor. There-fore, the relation between the
oscillation frequency and thecapacitance of the sensor can be
simplified as shown in
f = 1a
ffiffiffiffi
Cp , ð6Þ
where a is a constant.
Next, the high-frequency sine signal was transformed intothe
same frequency square wave through buffer and signalconversion
circuits. To fit the vibration frequency range ofmechanoreceptors
(0~400Hz), the square wave was processedby frequency division [20].
Last but not least, the dividedsquared wave was transformed into a
bionic spike, whichmakes it possible to connect the sensors to the
nervous system.
According to the signal encoding circuit, the frequency ofthe
resulting bionic pulsed signal was modulated by theapplied sliding
friction force. Figure 5(c) illustrates the signalresponses under
sliding friction loadings of 0, 2, and 4N. Thespike frequency
increased as the capacitance decreased due toslippage. As shown in
Figure 5(d), the fitting formula off = 63:21ex/6:79 − 20:02 was
obtained with loading in therange of 0 to ~4N. Moreover, a glove
equipped with thedesigned sensor and the circuit was put on a hand
to performthe sliding action (movie S4), which proved the
feasibility ofthe bionic signal encoding circuit. The designed Wien
bridgeoscillator circuit is jam-proof and easy to realize and can
be
(a)
(c)
0 20 40 60 80–2
02468
10
𝛥C
/C0 (
%)
Time (s)
0 10 20 30 40 50–20
–10
0
10
20
30
Time (s)
𝛥C
/C0 (
%)
–80
–60
–40
–20
0
20
0 3 6 9 12 15 18
𝛥C
/C0 (
%)
Time (s)
(e)
20 g
50 g 100 g
20 g 50 g 100 g
(i) (ii) (iii)
(i) (ii)
(iii)(iv)
(i) (ii)
(iii) (iv)
0 g
(b)
(d) (f)
Figure 4: Demonstrations of the Ruffini-ending-inspired flexible
shear force sensor with the ability to selectively respond to
static and slidingfriction forces. (a, b) The perception of weight
related to the static friction force. The robotic hand equipped
with the sensor grasped a bottlewith a constant force. The
capacitance increased as the water volume increased, which
corresponded to the static friction force. (c, d) Theperception of
slippage related to the sliding friction force. A glove equipped
with the sensor was put on to perform the sliding action.
Thecapacitance decreased immediately as soon as sliding began. (e,
f) The characteristic nonsensitivity to the normal force. Weights
of 20 g,50 g, and 100 g were put on this Ruffini-ending-inspired
sensor, in that order.
8 Research
-
easily adjusted within a wide frequency range. Through
thedesigned signal encoding circuit, the frequency range and
var-iation trend of the sensors are able to achieve similar results
tothe human response to force stimuli and thus could
endowneuroprosthetics with the ability of sliding perception.
3. Conclusion
In this work, we developed a Ruffini-ending-inspired flexi-ble
sensor with friction force selectivity. The sensor cannot only
measure shear forces with no response to the nor-mal force but also
further discriminate the static frictionforce and sliding friction
force according to the variation
tendency of the capacitance. Through FEA simulations,the
mechanism of the selective response capacitor wasconcluded to be
the deformation models of the verticaldouble helix architecture of
the capacitance plates. Further-more, we designed conversion
circuits to code the sensor’sanalog signals into bionic stimulus
signals, which may betransmitted to either the central nervous
system (CNS, thebrain and spinal cord) or the peripheral nervous
system(PNS, the muscle or peripheral nerve electrical activity).The
novel biomimetic flexible sensor presented in this workis
meaningful and important not only for neuroprostheticsbut also for
human-machine fusion, such as wearablerobots and exosuits.
(a)
(b)
(c) (d)
Slippage
Sensor
RC oscillationcircuit Buffer circuit
Sine wave tosquare wave
Frequencydivision
Square wave conversion topulse signal
0.00 0.05 0.10 0.15 0.20 0.25–0.05
0.00
0.05
0.10
0.15
0.20
Volta
ge (V
)
Time (s)1
40
50
60
70
80
90
100
Freq
uenc
y (H
z)
Sliding friction force (N)0 2 3 4
F = 0 Nfre = 44 Hz
F = 2 Nfre = 66 Hz
F = 4 Nfre = 93 Hz
Figure 5: The bionic signal encoding of the
Ruffini-ending-inspired flexible sensor. (a) Illustration of
prosthesis sliding perception. TheRuffini-ending-inspired sensor
with processing circuits is applied to a prosthetic hand of a woman
with a lower-arm amputation. (b)Schematic diagram showing a circuit
that converts analog signals recorded from the sensor into
low-frequency pulse signals (nerve-likesignals). (c) The pulse
shape through the processing circuit. The pulse frequency changed
with the applied sliding friction forces. (d) Thepulse frequency
responses in the 0 to 4N range of the sliding friction force.
9Research
-
4. Materials and Methods
4.1. Device Fabrication
4.1.1. Fabrication of the Silicon Wafer with
Fingerprint-likeMicrogrooves. Silicon molds with spiral
microgrooves ofdifferent depth-width ratios were fabricated by
traditionallithography and dry etching processes.
4.1.2. Fabrication of Flexible Friction Force Sensors.A
replica-tion method was employed to prepare a
spiral-column-typecapacitive sensor. (i) Ag nanowires (30 nm
diameter, 20μmlength) were dispersed in ethanol, and the Ag
nanowireconcentration was 1mg/mL. The silicon mold with
micro-grooves was soaked in trimethylchlorosilane for
approxi-mately 30 minutes, which was used as the mold-releaseagent.
(ii) Then, silver-coated copper wires as electrodes werefixed onto
the groove end of the silicon mold. Ag nanowiresolution (2mL) was
sprayed onto the silicon mold and elec-trodes. Next, to form two
plates of a capacitor, Ag nanowiresat the sidewalls and bottom of
the microgrooves wereretained, but Ag nanowires on the top layer of
the siliconwafer were scraped off. (iii) The PDMS prepolymer and
itscuring agent (Sylgard-184, Dow Corning) were stirred for20
minutes with a ratio of 7 : 1 (w/w). Poured PDMS wascured at 80°C
for 3 hours. With the aid of a mold-releaseagent, the Ag NW-PDMS
composite thin film was easilypeeled off from the silicon wafer
without any damage. (iv)Last, because Ag NWs fall off easily during
large deforma-tion, Ecoflex with a low Young modulus was chosen to
fillin the gaps. Part A and part B of the platinum cure
siliconerubber compound were diluted in n-hexane at a ratio of1 : 1
: 4 by weight. After stirring the diluted Ecoflex, the dis-persion
liquid was spin-coated onto the Ag NW-PDMS com-posite thin film at
different revolutions per minute accordingto the height of the
microcolumns. After spin-coating, thesensor was cured in a vacuum
oven at 70°C for 30 minutes.
4.1.3. Device Characterization. Capacitance measurementswere
taken using the Agilent B1500A semiconductor deviceanalyzer.
Capacitances were measured at a 1MHz frequencywith a 250mVAC
signal. Three kinds of forces were appliedby a customized
apparatus, which contains a z-axis electricmoving stage with a
force gauge and an x-axis electricmoving stage (Beijing Optical
Century Instrument Co., Ltd.,SC100 series stepper motor
controllers). In addition, the bionicspikes that occurred through
the conversion circuits wererecorded by an oscilloscope (Tektronix
DPO5034B). DC volt-age was provided by the voltage meter (RIGOL,
DP832).
Conflicts of Interest
The authors declare no competing financial interest.
Authors’ Contributions
Y.L. and Z.C. contributed equally to this work. T.L. and T.Z.are
cocorresponding authors. Y.L., T.L., and T.Z. designedthe
experiments and developed the theory. Y.L. and T.L.performed the
experiment including sensor fabrication, char-
acterization, and performance test. Z.C. designed the
signalencoding circuit. Y.L. and Y.B. performed finite
elementanalysis. All coauthors discussed the results. Y.L., Z.C.,
T.L.,N.L., and T.Z. all contribute to writing the manuscript. YueLi
and Zhiguang Cao contributed equally to this work.
Acknowledgments
The authors acknowledge the funding support from theNational Key
R&D Program of China (2017YFA0701101,2018YFB1304700), the
National Natural Science Foundationof China (51702354, 61801473),
the Youth PromotionAssociation of Chinese Academy of Sciences
(2020320),and the Foundation Research Project of Jiangsu
Province(SBK2020021442). We specially thank the scientific
supportfrom the Shanghai KESHEN Prostheses Co., Ltd., and theKey
Laboratory of Multifunctional Nanomaterials and SmartSystems,
Chinese Academy of Sciences.
Supplementary Materials
Supplementary 1. Figure S1: deformation under 200MPa ofnormal
force observed by finite element analysis. Figure S2:analysis of
changes in S and d under static (A) or sliding fric-tion forces
(B). Figure S3: changing process of the capacitancethat the static
converted to sliding friction force. Figure S4: thepulse-like
signal under static friction force. (A) The pulse fre-quency
responses in the 0 to 8.5N range of static friction force.(B) The
pulse shape through the custom-designed circuit. Thepulse frequency
decreased with the applied static friction force.Figure S5:
comparison between traditional parallel structureand spiral
structure. The spiral is centrosymmetric, whichensures the same
sensitivity to shear force from any directionin plane. Table S1:
the method for distinguishing shear forces.
Supplementary 2. Movie S1: weight perception of the
flexiblefriction force sensors. A robotic hand was set to grasp a
plas-tic bottle with a constant force. The capacitance of the
sensorincreased with the addition of water.
Supplementary 3. Movie S2: slippage detection of the
flexiblefriction force sensors. The flexible friction force sensor
wasmounted onto a wearable glove. The capacitance of the sen-sor
decreased once slippage occurred.
Supplementary 4. Movie S3: insensitivity to the normal forceof
the flexible friction force sensors. The capacitance of thesensor
was constant under different standard weights placedon the
sensor.
Supplementary 5. Movie S4: the bionic behavior of
flexiblefriction force sensors. The flexible friction force
sensormounted onto a wearable glove. The glove was used to applya
sliding friction force. The output is the frequency outputthrough
the conversion circuits. The frequency increasedonce slippage
occurred.
Supplementary 6. Movie S5: changing of the capacitance as
thestatic friction force converted to sliding friction force. A
gloveequipped with a Ruffini-ending-inspired sensor was put onhand
to perform the action. The sensor was exerting static fric-tion
force at the beginning and then converted to slide mode.
10 Research
http://downloads.spj.sciencemag.org/research/2020/8910692.f1.docxhttp://downloads.spj.sciencemag.org/research/2020/8910692.f2.mp4http://downloads.spj.sciencemag.org/research/2020/8910692.f3.mp4http://downloads.spj.sciencemag.org/research/2020/8910692.f4.mp4http://downloads.spj.sciencemag.org/research/2020/8910692.f5.mp4http://downloads.spj.sciencemag.org/research/2020/8910692.f6.mp4
-
References
[1] L. Zollo, G. di Pino, A. L. Ciancio et al., “Restoring
tactile sen-sations via neural interfaces for real-time
force-and-slippageclosed-loop control of bionic hands,” Science
Robotics, vol. 4,no. 27, article eaau9924, 2019.
[2] E. D’Anna, G. Valle, A. Mazzoni et al., “A closed-loop
handprosthesis with simultaneous intraneural tactile and
positionfeedback,” Science Robotics, vol. 4, no. 27, article
eaau8892,2019.
[3] A. P. Gerratt, H. O. Michaud, and S. P. Lacour,
“Elastomericelectronic skin for prosthetic tactile sensation,”
AdvancedFunctional Materials, vol. 25, no. 15, pp. 2287–2295,
2015.
[4] J. Kim, M. Lee, H. J. Shim et al., “Stretchable silicon
nanorib-bon electronics for skin prosthesis,” Nature
Communications,vol. 5, no. 1, article 5747, 2014.
[5] C. M. Boutry, M. Negre, M. Jorda et al., “A hierarchically
pat-terned, bioinspired e-skin able to detect the direction
ofapplied pressure for robotics,” Science Robotics, vol. 3, no.
24,article eaau6914, 2018.
[6] C. Bartolozzi, L. Natale, F. Nori, and G. Metta, “Robots
with asense of touch,” Nature Materials, vol. 15, no. 9, pp.
921–925,2016.
[7] P. Maiolino, M. Maggiali, G. Cannata, G. Metta, and L.
Natale,“A flexible and robust large scale capacitive tactile system
forrobots,” IEEE Sensors Journal, vol. 13, no. 10, pp.
3910–3917,2013.
[8] N. Wettels, V. J. Santos, R. S. Johansson, and G. E. Loeb,
“Bio-mimetic tactile sensor array,” Advanced Robotics, vol. 22, no.
8,pp. 829–849, 2012.
[9] S. K. Ameri, M. Kim, I. A. Kuang et al., “Imperceptible
electro-oculography graphene sensor system for human–robot
inter-face,” npj 2D Materials and Applications, vol. 2, no. 1, p.
19,2018.
[10] X. Yu, Z. Xie, Y. Yu et al., “Skin-integrated wireless
hapticinterfaces for virtual and augmented reality,” Nature,vol.
575, no. 7783, pp. 473–479, 2019.
[11] E. Roh, B. U. Hwang, D. Kim, B. Y. Kim, and N. E.
Lee,“Stretchable, transparent, ultrasensitive, and patchable
strainsensor for human_machine interfaces comprising a nanohy-brid
of carbon nanotubes and conductive elastomers,” ACSNano, vol. 9,
no. 6, pp. 6252–6261, 2015.
[12] L. E. Osborn, A. Dragomir, J. L. Betthauser et al.,
“Prosthesiswith neuromorphic multilayered e-dermis perceives touch
andpain,” Science Robotics, vol. 3, no. 19, article eaat3818,
2018.
[13] Y. Wu, Y. Liu, Y. Zhou et al., “A skin-inspired tactile
sensor forsmart prosthetics,” Science Robotics, vol. 3, no. 22,
articleeaat0429, 2018.
[14] H. U. Chung, B. H. Kim, J. Y. Lee et al., “Binodal,
wireless epider-mal electronic systems with in-sensor analytics for
neonatalintensive care,” Science, vol. 363, no. 6430, article
eaau0780, 2019.
[15] S. Wang, J. Xu, W. Wang et al., “Skin electronics from
scalablefabrication of an intrinsically stretchable transistor
array,”Nature, vol. 555, no. 7694, pp. 83–88, 2018.
[16] N. Bai, L. Wang, Q. Wang et al., “Graded
intrafillablearchitecture-based iontronic pressure sensor with
ultra-broad-range high sensitivity,” Nature Communications,vol. 11,
no. 1, p. 209, 2020.
[17] D. Kang, P. V. Pikhitsa, Y. W. Choi et al.,
“Ultrasensitivemechanical crack-based sensor inspired by the spider
sensorysystem,” Nature, vol. 516, no. 7530, pp. 222–226, 2014.
[18] Z. Ren, J. Nie, J. Shao et al., “Fully elastic and
metal-free tactilesensors for detecting both normal and tangential
forces basedon triboelectric nanogenerators,” Advanced Functional
Mate-rials, vol. 28, no. 31, article 1802989, 2018.
[19] C. Schwarz, “The slip hypothesis: tactile perception and
itsneuronal bases,” Trends in Neurosciences, vol. 39, no. 7,pp.
449–462, 2016.
[20] J. Scheibert, S. Leurent, A. Prevost, and G. Debregeas,
“Therole of fingerprints in the coding of tactile information
probedwith a biomimetic sensor,” Science, vol. 323, no. 5920,pp.
1503–1506, 2009.
[21] R. S. Johansson and J. R. Flanagan, “Coding and use of
tactilesignals from the fingertips in object manipulation
tasks,”Nature Reviews Neuroscience, vol. 10, no. 5, pp. 345–359,
2009.
[22] M. Knibestöl, “Stimulus-response functions of slowly
adaptingmechanoreceptors in the human glabrous skin area,” The
Jour-nal of Physiology, vol. 245, no. 1, pp. 63–80, 1975.
[23] V. E. Abraira and D. D. Ginty, “The sensory neurons of
touch,”Neuron, vol. 79, no. 4, pp. 618–639, 2013.
[24] K. O. Johnson, T. Yoshioka, and F. Vega–Bermudez,
“Tactilefunctions of mechanoreceptive afferents innervating
thehand,” Journal of Clinical Neurophysiology, vol. 17, no. 6,pp.
539–558, 2000.
[25] V. G. Macefield, C. Häger-Ross, and R. S. Johansson,
“Controlof grip force during restraint of an object held between
fingerand thumb: responses of cutaneous afferents from the
digits,”Experimental Brain Research, vol. 108, no. 1, pp.
155–171,1996.
[26] M. Knibestöl and Å. B. Vallbo, “Single unit analysis of
mecha-noreceptor activity from the human glabrous skin,” Acta
Phy-siologica Scandinavica, vol. 80, no. 2, pp. 178–195, 1970.
[27] D. J. Weber, M. Hao, M. A. Urbin, C. Schoenewald, andN.
Lan, “Sensory information feedback for neural prostheses,”in
Biomedical Information Technology (Second Edition),pp. 687–715,
Academic Press, USA, 2020.
[28] M. Hao, C. H. Chou, J. Zhang et al., “Restoring
finger-specificsensory feedback for transradial amputees via
non-invasiveevoked tactile sensation,” IEEE Open Journal of
Engineeringin Medicine and Biology, vol. 1, pp. 98–107, 2020.
[29] Y.-L. Park, C. Majidi, R. Kramer, P. Bérard, and R. J.
Wood,“Hyperelastic pressure sensing with a liquid-embedded
elasto-mer,” Journal of Micromechanics and Microengineering,vol.
20, no. 12, article 125029, 2010.
[30] Z. Wang, A. A. Volinsky, and N. D. Gallant,
“Crosslinkingeffect on polydimethylsiloxane elastic modulus
measured bycustom-built compression instrument,” Journal of
AppliedPolymer Science, vol. 131, no. 22, article 41050, 2014.
[31] M. Zouari, M. Kharrat, and M. Dammak, “Wear and
frictionanalysis of polyester coatings with solid lubricant,”
Surfaceand Coating Technology, vol. 204, no. 16-17, pp.
2593–2599,2010.
[32] E. Ribot-Ciscar, J. P. Vedel, and J. P. Roll, “Vibration
sensitiv-ity of slowly and rapidly adapting cutaneous
mechanorecep-tors in the human foot and leg,” Neuroscience
Letters,vol. 104, no. 1-2, pp. 130–135, 1989.
[33] J. P. Vedel and J. P. Roll, “Response to pressure and
vibrationof slowly adapting cutaneous mechanoreceptors in the
humanfoot,” Neuroscience Letters, vol. 34, no. 3, pp. 289–294,
1982.
11Research
Highly Selective Biomimetic Flexible Tactile Sensor for
Neuroprosthetics1. Introduction2. Results2.1. Design and
Characterization of Flexible Friction Force Sensors2.2. Response
and Sensing Mechanism of the Flexible Friction Force Sensors2.3.
Slippage Detection and Weight Perception of Flexible Friction Force
Sensors2.4. The Bionic Behavior of Flexible Friction Force
Sensors
3. Conclusion4. Materials and Methods4.1. Device
Fabrication4.1.1. Fabrication of the Silicon Wafer with
Fingerprint-like Microgrooves4.1.2. Fabrication of Flexible
Friction Force Sensors4.1.3. Device Characterization
Conflicts of InterestAuthors’
ContributionsAcknowledgmentsSupplementary Materials