Lightweight soft neuroprosthetic hand

Guoying Gu1,2†*, Ningbin Zhang1†, Haipeng Xu1, Shaoting Lin3, Yang Yu1, Guohong Chai1, Lisen
Ge1, Houle Yang1, Qiwen Shao1, Xinjun Sheng1,2, Xiangyang Zhu1,2†*, and Xuanhe Zhao3,4†*
1
Robotics Institute, School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240,
China; 2State Key Laboratory of Mechanical System and Vibration, Shanghai Jiao Tong University,
Shanghai 200240, China; 3Department of Mechanical Engineering, Massachusetts Institute of Technology,
Cambridge, MA 02139, USA; 4Department of Civil and Environmental Engineering, Massachusetts
Institute of Technology, Cambridge, MA 02139, USA
†
These authors contributed equally to the current work.
*Authors to whom correspondence should be addressed: [email protected], [email protected],
[email protected]

Abstract

Mainly composed of electrical motors and sophisticated mechanical components, existing

neuroprosthetic hands1,2 are typically heavy (>400 g) and expensive (>USD 10,000), and they
lack the compliance and tactile feedback of human hands. These limitations hamper
neuroprosthetic hands’ innovation and broad utility for amputees3-5. Here we report the
design, fabrication and applications of a lightweight (292 g) and potentially low-cost
(component cost below USD 500) soft neuroprosthetic hand with simultaneous myoelectric
control and tactile feedback. The soft neuroprosthetic hand consists of five soft fingers and a
palm to give six active degrees of freedom under pneumatic actuation, four electromyography
sensors that measure the surface electromyogram signals to control the hand to deliver four
common grasp types, and five hydrogel-elastomer capacitive sensors on the fingertips that
measure the touch pressure and elicit electrical stimulation on the skin of the residual limb.
The soft finger is made of a fiber-reinforced elastomeric structure embedded with rigid
segments to mimic the soft-joint/rigid-bone anatomy of the human finger. We use a set of
standardized tests6 to compare the speed and dexterity of the soft neuroprosthetic hand and
a conventional rigid neuroprosthetic hand7 on two transradial amputees. The soft
neuroprosthetic hand gives overall superior performances to the rigid hand. We further
demonstrate that one transradial amputee wearing the soft neuroprosthetic hand can regain
the versatile hand functions with primitive touch sensation and real-time closed-loop control
in daily activities such as handling tools, eating, shaking hands, petting animals, and
recognizing touch pressure. This work not only represents a new paradigm for designing soft

1
neuroprosthetic devices but also opens an avenue to widespread applications of lightweight,
low-cost, and compliant hand replacements for amputees.

Introduction

There are over five million upper-limb amputees worldwide, and the number increases by a

substantial margin each year. Losing a hand is generally catastrophic, seriously limiting a person’s

ability in daily activities1. Although artificial prostheses are available, the most widely used

prostheses are still cosmetic devices or functional hook-like grippers. While a few anthropomorphic

neuroprosthetic hands1,2 (such as i-Limb Hand, Michelangelo Hand, Bebionic Hand, and Vincent

Hand) have been commercialized, they all rely on electrical motors and sophisticated mechanical

components. The high weights (>400 g) and high prices (>USD 10,000) of these neuroprosthetic

hands severely limit their broad utility for amputees3,4. It is also desirable for neuroprosthetic hands

to have the compliance and tactile feedback of human hands1,2,5. For instance, the Pisa/IIT

SoftHand8-10 composed of electrical motors and tendon-driven mechanisms with compliant joints

or skins has been developed as a myoelectric prosthesis with high compliance and 520 g of weight.

In addition, several recent studies11-18 have been reported to investigate the invasive or noninvasive

sensory feedback for upper-limb amputees, but they are only implemented in rigid neuroprosthetic

hands. The emerging field of soft robotics19-21 that combines the compliance of human skin and

muscle with simple design and fabrication of lightweight elastomeric components may open a

promising avenue for future neuroprosthetic hands. Although many bioinspired soft robotic systems

such as locomotive robots22, finger-like actuators23-26, robotic exoskeletons27-29 and hands30-34 have

been developed, to the best of our knowledge, none of them has been demonstrated as functional

neuroprosthetic hands on upper-limb amputees (Extended Data Table 1).

2
 Here, we report the design, fabrication and applications of a lightweight (292 g) and potentially

low-cost (component cost below USD 500) soft neuroprosthetic hand with simultaneous

myoelectric control and tactile feedback for transradial amputees (Fig. 1A and Extended Data

Tables 2, 3). The soft neuroprosthetic hand consists of five soft fingers and a palm to give six active

degrees of freedom (DoFs), four electromyography sensors that measure the surface

electromyogram (EMG) signals of residual forearm muscles to control the hand to deliver four

common grasp types, and five hydrogel-elastomer capacitive sensors on the fingertips that measure

touch pressure and elicit electrical stimulation on the skin of the residual limb. To evaluate the

function of the soft neuroprosthetic hand, two transradial amputees have carried out a set of

standardized tests6 (including the Box and Blocks Test, all seven tasks in the Jebsen-Taylor Hand

Function Test, and nine selected tasks in the Southampton Hand Assessment Procedure). One

transradial amputee with the soft neuroprosthetic hand has further demonstrated the dexterous and

versatile hand functions with primitive tactile sensation and closed-loop control in daily activities

such as handling tools, eating, shaking hands, petting animals, and recognizing touch pressure.

Design

Hand. Each finger of the soft neuroprosthetic hand is based on a fiber-reinforced elastomeric

tubular structure, in which two or three rigid segments with specific lengths are embedded to mimic

the soft-joint/rigid-bone anatomy of the thumb or other fingers, respectively, of a human hand35,36

(Fig. 1B and Supplementary Figs. 1, 2). The thumb-palm connection in the soft neuroprosthetic

hand is based on a fiber-reinforced elastomeric hollow pad with a strain-limiting layer

(Supplementary Fig. 3). The soft neuroprosthetic hand is designed to possess six active DoFs (Fig.

1C): each soft finger can be pneumatically actuated to provide one flexion DoF and the thumb-palm

3
connection under pneumatic actuation further gives the thumb another circumduction DoF. In

addition, the inherent compliance of the soft fingers can introduce many passive DoFs to the soft

hand for dexterous and adaptive grasps even on fragile and soft objects (Fig. 1D), mimicking those

passive compliance of human hands8.

We choose the pneumatically-actuated soft fingers owing to their advantages of low cost, light

weight, and scalable fabrication21. As the applied pneumatic pressure to a soft finger increases, the

bending angles of the flexible joints increase. We develop an analytical model to study the bending

angles of a soft finger’s flexible joints (see Methods section “Analytical model for the design of the

soft finger”). The model demonstrates that, with a fixed finger length, the bending angles of the

flexible joints are linearly proportional to the lengths of the flexible joints. This trend is consistent

with the experimental results (Supplementary Figs. 4, 5). We further develop a finite-element model

to quantitatively predict the bending angles of the flexible joints and the deformation of the soft

finger upon pneumatic actuation (see Methods section “Finite-element model for the design of the

soft finger”). The finite-element model is in quantitative agreement with the experimental results

(Fig. 1E, Extended Data Fig. 1 and Supplementary Video 1). For example, the finite-element model

predicts that the three flexible joints in the soft index finger under the maximum pneumatic

actuation of 120 kPa give the bending angles of 84.02o, 112.58o and 52.55o, only deviating from the

experimental results by 3.25%, 11.60% and 7.44%, respectively (Extended Data Table 4).

The five soft fingers are mounted on a three-dimensionally (3D) printed plastic palm (Imagine

8000, SOMOS Inc., Netherlands) in the shape of a human palm. Both the fingers and the palm are

covered by a soft elastomeric layer mimicking the skin on the human hand. The palm skeleton is

further connected to a customized plastic socket that fits with the residual limb of the transradial

amputee. Owing to the pneumatic actuation and modular design of the soft neuroprosthetic hand,

4
the pumps, valves, electronic boards and battery for the hand can be contained in a small bag (total

weight of 444 g) on the waist of the amputee (Extended Data Fig. 2A). The pumps and electronic

board are connected by soft tubing and electrical wires (hidden under clothes and in the socket) to

the soft fingers and sensors on the hand, respectively. This modular design can dramatically reduce

the weight of the soft neuroprosthetic hand to 292 g, much lighter than the weights of commercially-

available neuroprosthetic hands (420 g - 628 g; Extended Data Table 3) and the average human

hands (400 g)1,2. Furthermore, since the commercially-available neuroprosthetic hands integrate the

electric motors, transmission mechanisms and batteries in their palms and sockets, it is challenging

to significantly reduce their weights. We further demonstrate that the six active DoFs of the soft

neuroprosthetic hand can be independently controlled with one pump and twelve valves

(Supplementary Video 2). Notably we can also integrate the pumps, valves, electronic boards and

battery in the socket of the soft neuroprosthetic hand if required in any application (Extended Data

Fig. 2B). Such a design increases the weight of the soft neuroprosthetic hand to 604 g, which is still

lighter than or comparable to the weights of commercially-available neuroprosthetic hands such as

the i-Limb large hand and the Bebionic medium hand (Extended Data Table 3).

EMG sensors. In the socket of the hand, we implement four customized EMG sensors (Fig. 1A

and Extended Data Fig. 3), which will be mounted on the skin of the residual limb to record the

surface EMG signals from the target muscles in the residual limb. Each EMG sensor (weight of 10

g) consists of three electrodes (a reference electrode and a pair of differential electrodes) and readout

electronics (including two-level signal amplifier and filter circuits; see Supplementary Fig. 6). The

locations of the EMG sensors with respect to the target muscles have been optimized, aiming to

achieve a superior performance on decoding the amputees’ motion intention. The decoded motion

intention will be used to control the soft fingers and palm to deliver the corresponding grasp types.

5
 Touch sensors and electrical stimulators. To sense the touch pressure of the soft neuroprosthetic

hand applying on objects, we implement five soft capacitive touch sensors, each on a fingertip of

the hand (Fig. 1C). The capacitive touch sensor is composed of an ionic hydrogel-elastomer hybrid

structure37-39 that forms a capacitor for sensing the touch pressure (Supplementary Fig. 7A). An

increase in the touch pressure reduces the thickness of the elastomeric layer and, therefore, increases

the capacitance of the capacitor (Supplementary Fig. 7B). The measured relative capacitance

change is used to control an electrical stimulator (Supplementary Fig. 8), which outputs

programmable electrical pulses via a noninvasive stimulation electrode on a specific region of the

residual limb (Supplementary Fig. 9) to inform the amputee the touch pressure on the corresponding

fingertip. This is the first time that hydrogel-elastomer capacitive touch sensors have been used on

the soft neuroprosthetic hand to measure the touch pressure at the fingertips.

Control algorithm. By integrating the EMG sensors, touch sensors and electrical stimulators

with the hand, we develop a bidirectional human-machine interface for the soft neuroprosthetic

hand (Supplementary Fig. 8 for the description of experimental setup). We choose a pattern

recognition approach40 to classify the surface EMG signals into five classes corresponding to the

four common grasp types41,42 and rest type of human hands (see Supplementary Information for the

descriptions of pattern recognition algorithms). The decoded grasp types are then mapped onto the

actuation command for the corresponding soft fingers and palm. Based on the relative changes of

the measured capacitances of the capacitive touch sensors (caused by the touch pressure), we can

program the electrical stimulator to output the corresponding stimulation patterns, which inform the

amputee the touch pressure on each fingertip. Based on the informed touch pressure, the amputee

can further change the grasp type of the soft neuroprosthetic hand by varying the EMG signals,

therefore forming a closed-loop control of the hand.

6
Characterization

We use a simple and scalable method to fabricate and assemble the soft neuroprosthetic hand

(see Methods section “Fabrication and assembly of the soft neuroprosthetic hand” and Extended

Data Fig. 4). We next characterize the performance of the fabricated soft neuroprosthetic hand. To

monitor its kinematics, we develop a motion tracking system (Supplementary Fig. 10). For the 1-

DoF flexion fingers (i.e., index, middle, ring, and little fingers), we attach markers to their flexible

joints and fingertips to measure the bending angles of the joints and the fingertips (Fig. 2A). We

find that an increase of the applied pneumatic pressure increases the bending angles of the fingertip

to a maximum of 231° at 120 kPa pneumatic pressure (Fig. 2B). In addition, the measured bending

angles of the flexible joints are consistent with the finite-element model’s prediction (Fig. 2B,

Extended Data Fig. 1A and Extended Data Table 4). For the 2-DoF thumb, we attach markers to

its flexible joints, thumb tip and thumb-palm connection (Fig. 2C). The results indicate that the

thumb has a maximum flexion of 69° and a maximum palm circumduction of 28° at 80 kPa

pneumatic pressure, which also agrees well with the finite-element model’s prediction (Fig. 2D,

Extended Data Fig. 1B, Supplementary Fig. 11 and Extended Data Table 4). The maximum

bending angles of the soft fingers are comparable to those of existing rigid counterparts (Extended

Data Table 3).

To evaluate the load capacity of the soft neuroprosthetic hand, we use the hand (at 100 kPa

pneumatic pressure to the 1-DoF fingers and 80 kPa pneumatic pressure to the 2-DoF thumb) to

grasp a 55 mm-diameter cylinder while measuring the grasping force with an electronic

dynamometer (Supplementary Fig. 12). The results indicate that the soft neuroprosthetic hand gives

maximum grasping forces of 18 N and 17 N along the vertical and horizontal directions, respectively

7
(Fig. 2E, F). Therefore, the soft neuroprosthetic hand can perform most grasping tasks in daily

activities, which generally require grasping forces below 10 N for human hands43. The grasping

forces of the soft fingers can be potentially increased in future studies by using stiffer elastomers

for the fingers and applying higher pneumatic pressures.

Furthermore, we demonstrate that our soft fingers have repeatable pressure-flexion relations

with small hysteresis over 10,000 cycles of actuations (Fig. 2G), and they are resilient to be bent

by arbitrary angles, struck with a steel hammer, and run over back and forth by one wheel of a 1500

kg vehicle (Supplementary Fig. 13 and Supplementary Video 3).

Based on the usage frequencies of grasp types in daily activities41,42, we choose to recognize the

most commonly used four grasp types (i.e., Power, Precision disk, Tripod and Lateral pinch) of

amputees through the myoelectric control interface (Extended Data Fig. 5). Using the four-channel

EMG sensors (one sensor at each channel), the soft neuroprosthetic hand can decode the intended

four grasp types and rest type (Fig. 2H and Supplementary Fig. 14).

We next calibrate the capacitive touch sensor by uniformly compressing the sensor and then

measuring its capacitance. Under zero pressure, we denote the measured capacitance of the sensor

as 𝐶 . With the increase of the applied pressure, we can obtain the current capacitance C of the

touch sensor and calculate the change of the capacitance ∆𝐶 𝐶 𝐶 . The experimental results

demonstrate that, when the uniformly applied pressure on the sensor increases from 0 kPa to 55.8

kPa, the ∆𝐶/𝐶 varies from 0 to 0.85, giving a sensitivity of 0.016 kPa-1 (Supplementary Fig. 7B).

Applications on amputees

Performances of soft and rigid neuroprosthetic hands. The soft neuroprosthetic hand can

restore the versatile hand functions for amputees. A subject with transradial amputation can adapt

8
to the soft neuroprosthetic hand and master its functions after training for 15 minutes (Extended

Data Fig. 6 and Supplementary Video 4). We use a set of standardized tests6 to evaluate the

performances of the soft neuroprosthetic hand (Fig. 3A, Extended Data Fig. 7, and Supplementary

Video 5), including the Box and Blocks Test (BBT, i.e., counting the number of blocks that can be

grasped and transported per minute), all seven tasks in the Jebsen-Taylor Hand Function Test (i.e.,

J1-writing, J2-simulated page-turning, J3-lifting small common objects, J4-simulated feeding, J5-

stacking checkers, J6-lifting large light objects, and J7-lifting large heavy objects), and nine selected

tasks in the Southampton Hand Assessment Procedure (i.e., grasping nine kinds of objects including

S1-spherical light, S2-spherical heavy, S3-tripod light, S4-power light, S5-power heavy, S6-tip light,

S7-tip heavy, S8-extension light, and S9-extension heavy). As a comparison, the same subject

wearing a conventional rigid neuroprosthetic hand7 trained by the same amount of time (about 15

minutes) has performed the same set of standardized tests (Supplementary Fig. 15 and

Supplementary Video 6). Throughout the tests, the subject feels that the soft neuroprosthetic hand

is much lighter to wear than the rigid neuroprosthetic hand. Statistical analyses (Fig. 3A and

Extended Data Table 5) further demonstrate that, compared to the rigid neuroprosthetic hand, the

soft neuroprosthetic hand has significantly superior performances in 7 items (i.e., BBT, J3, J4, J7,

S1, S3 and S5), significantly inferior performances in 3 items (i.e., J5, S6 and S7), and statistically

similar performances (p > 0.05) in 7 items (i.e., J1, J2, J6, S2, S4, S8 and S9). Based on the same

set of standardized tests and statistical analyses (Extended Data Fig. 8 and Extended Data Table

6), another subject with transradial amputation demonstrates that the performances of the soft

neuroprosthetic hand are significantly superior in 5 items (i.e., J3, J5, S4, S5 and S8), significantly

inferior in 4 items (i.e., J1, S6, S7 and S9), and statistically similar in 8 items (i.e., BBT, J2, J4, J6,

J7, S1, S2 and S3) compared to those of the rigid counterpart (Supplementary Fig. 16)45-47. (Note

9
the reported results have only provided mean values of their performances, whose statistical

difference from the performances of the soft neuroprosthetic hand cannot be evaluated.)

Next, we perform comparative experiments for the same subject wearing the soft

neuroprosthetic hand and the rigid i-Limb hand to grasp fragile objects (e.g., strawberry, bread,

paper cup). The results indicate that the rigid neuroprosthetic hand damages the strawberry and

bread, and tends to crush the paper cup (Extended Data Fig. 9A and Supplementary Video 7).

Owing to the inherent compliance, the soft neuroprosthetic hand can guarantee safe interactions

with these fragile and soft objects (Extended Data Fig. 9B and Supplementary Video 7).

Furthermore, we demonstrate that one subject with the soft neuroprosthetic hand can intuitively

perform the four common grasp types to grasp different objects (Supplementary Video 8) and

handle commonly-used items in daily activities, such as food (e.g., potato chips, cakes, strawberries

and apples), commodities (e.g., clothes, bags, laptops, water glasses, bottles, tissues and dishes),

and tools (e.g., hammers and pliers). The subject also achieves safe interactions with other persons

(e.g., shaking hands), animals (e.g., petting a cat) and environments (e.g., touching a flower) (Fig.

3B and Supplementary Video 9). We further demonstrate that the subject can successfully carry out

delicate tasks to handle objects with complex shapes and different sizes and then insert them in the

corresponding slots precisely (Fig. 3C, and Supplementary Video 10). In a load test, the subject

wearing the soft neuroprosthetic hand can lift a payload of 2.3 kg (Fig. 3D and Supplementary

Video 11).

Tactile feedback and closed-loop control. Furthermore, we demonstrate that the subject with

the soft neuroprosthetic hand can restore primitive touch sensation based on the capacitive touch

sensors and electrical stimulators. When the effective pressure on the touch sensor of a fingertip

reaches a threshold (i.e., threshold ∆𝐶/𝐶 ), the electrical stimulator will be triggered to generate an

10
electrical pulse (amplitude of 4.0 mA, pulse width of 200 μs, and pulse frequency of 20 Hz) to

stimulate a specific region on the residual limb corresponding to the fingertip (Fig. 4A and

Supplementary Fig. 9). We set the threshold effective pressure to be 2.3 kPa (i.e., the threshold

∆𝐶/𝐶 = 0.1), so that the touch sensors are sufficiently sensitive to touch pressures commonly

experienced in daily activities44 yet unaffected by environmental noises and crosstalk among

sensors (Supplementary Fig. 17). In a blindfolded and acoustically-shielded interaction experiment,

we gently compress the five fingers of the soft neuroprosthetic hand in random combinations. The

subject can almost instantaneously distinguish any individual finger or multiple fingers being

compressed (Fig. 4B and Supplementary Video 12).

Next, we demonstrate the closed-loop control capability of the soft neuroprosthetic hand

enabled by integrating the myoelectric control and the tactile feedback. In a blindfolded and

acoustically-shielded experiment, the subject uses his own EMG signals to control the soft

neuroprosthetic hand to give the power grasp type. If the hand firmly grasps a bottle so that the

effective pressures on the five fingertips are above the threshold, the subject is informed by the

electrical stimulators to lift up the bottle (Fig. 4 C, D and Supplementary Video 13). In contrast, if

the hand does not grasp anything that applies pressure on the fingertips, the subject does not lift up

but relaxes the hand after a few seconds (Fig. 4 C, D and Supplementary Video 13).

Furthermore, by programing the frequencies (i.e., 5 Hz, 20 Hz or 35 Hz) of the electrical pulses

to map different ranges of ∆𝐶/𝐶 of the touch sensor on the middle finger (i.e., no stimulation when

∆𝐶/𝐶 0.1, 5 Hz when 0.1 ∆𝐶/𝐶 0.3, 20 Hz when 0.3 ∆𝐶/𝐶 0.4, and 35 Hz when

∆𝐶/𝐶 0.4), we demonstrate that, in the blindfolded and acoustically-shielded experiment, the

subject can restore the graded tactile feedback to discriminate three cylinders with different

11
diameters (i.e., 60 mm, 70 mm and 80 mm) with an accuracy of 96.25% (Extended Data Fig. 10

and Supplementary Video 14).

Conclusion

We report a lightweight and potentially low-cost soft neuroprosthetic hand for transradial

amputees to restore versatile hand functions and primitive tactile sensation. We choose a modular

design to enable efficient iteration of the design, fabrication and control, as well as rapid

replacement of the components in case of wear or damage. Compared to commercially-available

neuroprosthetic hands, our soft neuroprosthetic hand has advantages including the intrinsic

compliance, light-weight (292 g), potentially low-cost (component cost below USD 500), and

embedded soft touch sensors, while maintaining the similar active DoFs, the number of joints, and

the maximum bending angles of the joints (Extended Data Table 3). As the first demonstration of

the soft neuroprosthetic hand on transradial amputees, we employ the most commonly used EMG-

decoding algorithm and the electrotactile feedback. To further improve the performance of the soft

neuroprosthetic hand, advanced EMG-decoding algorithms48-50 and sensory feedback approaches11-

18
can be implemented in the future. Overall, this work has the potential to provide the next-

generation personalized neuroprosthetic hands that are intrinsically soft, lightweight and potentially

low-cost for upper-limb amputees, and to broaden the future applications of soft robotic systems.

12
Methods
Synthesis of the ionic hydrogel
The ionic hydrogel37-39 is a polyacrylamide (PAAm) hydrogel containing lithium chloride (LiCl).
The PAAm-LiCl hydrogel is synthesized by using the acrylamide (AAm; J&K) as the monomer,
N,N’-methylenebisacrylamide (MBAA; Molbase) as the crosslinker, LiCl monohydrate (LiClꞏH2O;
Sinopharm Chemical Reagent) as the ionic conductive medium, and 2-Ketoglutaric Acid (Adamas)
as the photoinitiator. The monomer solution is prepared by mixing AAm, LiClꞏH2O and deionized
water with a mass ratio of 9.98%: 16.16%: 73.86%. MBAA solution is dissolved into deionized
water with a mass ratio of 1.2%. Then, the monomer solution, MBAA solution, and 2-Ketoglutaric
Acid are mixed with the mass ratio of 96.67%: 1.13%: 2.20% to form the hydrogel precursor ink
for further fabrication. Note that the LiCl is added into the hydrogel, not only serving as a conductive
medium but also as a hygroscopic salt to maintain water in the hydrogel in ambient environments37-
39
.

Fabrication and assembly of the soft neuroprosthetic hand

We present a simple and scalable method to fabricate and assemble the soft neuroprosthetic hand
(Extended Data Fig. 4). For the soft fingers, we use the Dragon Skin 10 (Smooth-On Inc., USA)
silicone rubber for the inner elastomeric tubular structure, the Ecoflex 0030 (Smooth-On Inc., USA)
silicone rubber for the outer elastomeric skin, and the polyethylene thread for the fiber
reinforcement. We attach the carbon fiber-reinforced plastics lamination with the heat-shrink tubes
as the embedded rigid segments of the soft fingers. The palm skeleton is 3D printed with the
commercial photosensitive resin (Imagine 8000, SOMOS Inc., Netherlands) and the covered
elastomeric skin is made of Ecoflex 0030 silicone rubber. To tune the color of the elastomeric skin
close to the amputee’s skin color, we can add Slic Pig pigment PMS 488C (Flesh color, Smooth on
Inc., USA) into the Ecoflex 0030 silicone rubber (see Supplementary Fig. 18 for an illustration).
The socket is fabricated with the vacuum-forming thermoplastic acrylic resin and the commercial
gold-plated copper blocks are embedded in the socket as the EMG electrodes. We fabricate the
capacitive touch sensors by curing the hydrogel precursor (containing acrylamide, crosslinker,
photoinitiator, and ionic conductive medium) into a VHB elastomeric matrix. Notably, all the
mechanical components can be constructed with low-cost commercially available materials
(Extended Data Table 2). The fabrication and assembly steps of the soft neuroprosthetic hand are
detailed in the Supplementary Information.

Analytical model for the design of the soft finger

The method for efficiently predicting the bending angles of the flexible joints in the soft fingers
plays an important role in mimicking the structure and function of human fingers24. We first develop
an analytical model to analyze the pneumatic response of a flexible joint. We use a non-linear
elasticity approach to analytically model the response of the flexible joint under pneumatic actuation.
Specifically, we model the flexible joint as a hollow cuboid of isotropic incompressible neo-

13
Hookean solid with shear modulus of μ and a stiff inextensible layer beneath the cuboid. We take
the thickness, height, and width of the hollow cuboid at the undeformed state as t, H, and W,
respectively (Supplementary Fig. 4A). Due to the lateral constraint from the stiff fibers surrounding
the cuboid, we assume the cross-sections of the cuboid remain planar upon pressurization, which
has been validated by the finite-element simulation as well. Furthermore, since the arrangement of
fibers are symmetric without twisting, we express the axial stretch of the cuboid cross-section in a
linear form of lz ( y) = y [lz ( H ) -1] / H +1 , where l z ( H ) = l zH = l / L is the axial stretch at the top
surface with the elongated length of l at the top surface (Supplementary Fig. 4A). The deformation
gradient at the side surfaces reduces to
élz-1 ( y ) 0 0 ùú
ê
F = êê 0 1 0 úú (1)
ê 0 ú
0 lz ( y )úû
êë
We model the elastomeric cuboid as an incompressible neo-Hookean solid with strain energy of
W    FF T  3  / 2 . Therefore, the axial nominal stress in the cuboid is
-3
éy ù éy ù
S zz ( y ) = m ê (lzH -1) + 1ú - m ê (lzH -1) + 1ú (2)
êë H úû êë H úû
Assuming the deformation of the hollow cuboid is pure bending, we can obtain the geometrical
relation between the bending angle α and the dimensions of the flexible joint (i.e., l, L, and H)
l-L
a= (3)
H
Since the bending actuation is driven by the moment created by the internal pressure imposing on
the cap of the flexible joint Mp, the relation between the pneumatic pressure p and the bending angle
α can be found by equating Mp to the moment generated by the internal stress of the elastomeric
cuboid Mc. To calculate Mc, we take the stiff inextensible layer as the position of the neutral axis
H
M c = 2ò S zz ( y )ytdy + S zz ( H )WHt (4)
0

The moment created by the internal pressure is

H
Mp =ò pWydy (5)
0

Solving Eq. (4) and Eq. (5) yields the relation between the pneumatic pressure and the bending
angle as
2t é 2 l H 3 - 3 l H 2 + 6 l H - 1 - 3 l H - 2 - 2 ù + 2t é l H - l H - 3 ù = p
ê ( z ) ( z) ( z) ( z) ú ê( z ) ( z ) ú (6a)
3W ë û H ë û m
H
lzH = a
+1 (6b)
L
To verify the above analytical model, we further simulate the pneumatic response of the flexible
joint, which consists of a cuboid elastomeric chamber, a network of stiff fibers surrounding the

14
elastomeric chamber, and a stiff layer beneath the chamber. We model the elastomeric chamber as
a neo-Hookean solid with solid elements (C3D10H), the strain energy density of which is
 
W
2
 FF T
 3 
2
 J  1
2
(7)

where 𝜇 is the shear modulus,  is the bulk modulus, 𝐽 det F . We set 𝜇=80 kPa and  /𝜇 = 1000
to impose the nearly incompressibility of the material. The stiff layer is modeled as a skin of shell
elements with Young’s modulus of 210 GPa. The fiber is modeled as a beam element (B32H) with
Young’s modulus of 1 GPa. The fiber direction is set perpendicular to the axial direction of the
finger to constrain the lateral expansion of the flexible joints during inflation17. Static simulations
are performed by applying pressure on all internal faces of the elastomer chamber with zero
displacements at the base portion of the finger as the boundary conditions. We first compare the
axial stretch at the top surface of the flexible joint in experiments and simulations, verifying the
main deformation mode of the flexible joints is the bending motion (Supplementary Fig. 4B). We
further compare the increase of the bending angle α as a function of the applied inflation pressure p
for the flexible joints with various lengths in the finite-element simulation and the analytical model,
showing good agreement (Supplementary Fig. 4C). Despite the nonlinear large deformation of the
elastomeric cuboid, we also show that the bending angle of the joint is almost linearly proportional
to its initial length L under a constant inflation pressure p (Supplementary Fig. 4D).
We further study the pneumatic response of a soft finger consisting of an inner hollow cuboid
elastomeric tubular structure surrounded by a network of fibers, rigid segments, and a stiff layer
beneath the elastomeric tube (Supplementary Fig. 5A). Mimicking the dimensions of the distal
phalanx, middle phalanx, and proximal phalanx in human fingers17,18, we set the total length of the
finger as 85 mm and the length ratio of the top, middle, and bottom parts as 2: 3: 5 in the model
(Supplementary Fig. 5A). We show that the ratio of the bending angles of the three joints are nearly
constant due to the same pneumatic pressure applied to the three joints (Supplementary Fig. 5B).
Moreover, the ratio of the bending angles of the three joints are independent of the total length of
all flexible joints 𝐿 𝐿 𝐿 as long as the ratio of the lengths of the three joints are fixed (e.g.,
𝐿 :𝐿 :𝐿 2: 4: 1, Supplementary Fig. 5C).
In existing prosthetic hands1,2, the bending motion of a finger typically follows two
characteristics: 1) the bending angle of α is approximately half of that of α ; and 2) the bending
angle of α is approximately the same as α . Based on our analysis, we show that the central idea
to mimic the motion trajectory of human fingers is to design the length ratio of the three flexible
joints. Specifically, the length of the distal interphalangeal joint 𝐿 needs to be half of the length of
the metacarpophalangeal joint 𝐿 , while the length of the proximal interphalangeal joint 𝐿 is
preferred to be the same as the length of the metacarpophalangeal joint 𝐿 . Although we analyze
the soft fingers with three flexible joints, the conclusion also applies to the thumb with two flexible
joints, which will also be verified in the simulation with the finite-element model.

15
Finite-element model for the design of the soft finger
To further verify the experimental results of the hand with five soft fingers, we develop a finite-
element model of the soft fingers in ABAQUS to quantitatively predict the bending angles of the
flexible joints and the motion trajectories of the fingers upon inflation. All material parameters are
experimentally measured from mechanical characterizations51. The elastomeric tube is modeled as
a neo-Hookean solid (Eq. 7) with the shear modulus 𝜇=85 kPa and the ratio of bulk modulus to
shear modulus  /𝜇 = 1000 to impose the near incompressibility of the material. The silicone skins
on the outer surface are modeled as a linear elastic material with elastic modulus 0.125 MPa, and
Poisson’s ratio 0.45. The PE fibers are modeled as a linear elastic material with the Young’s modulus
2.9 GPa, and Poisson’s ratio 0.41. The heat-shrinkable tube is modeled as a linear elastic material
with the Young’s modulus 55 MPa, and Poisson’s ratio 0.4. The carbon fiber-reinforced plastics is
modeled as a linear elastic material with the Young’s modulus 210 GPa and Poisson’s ratio 0.3. The
fiberglass grid is modeled as a linear elastic material with the Young’s modulus 73 GPa, and
Poisson’s ratio 0.22.
To reduce the computational cost, we use the skin module to model the heat-shrinkable tube,
the carbon fiber-reinforced plastics, and the fiberglass grid. The Ecoflex 00-30 layer are modeled
as skins with uniform thickness as the outer surfaces of the fiber-reinforced inner elastomeric tube.
An extra skin of a stiffer rubber material is added on the surfaces of the thumb-palm connection to
model the constraint effect of the silicone layer. The elastomer of the inner tube is modeled as solid
elements C3D4H. The fibers are modeled as beam elements B31. Tie constraints are set between
the fibers and the elastomeric chambers of the finger actuators, the thumb actuator, and the thumb-
palm connection actuator. The proximal ends of the thumb-palm connection are fixed to the rigid
palm skeleton. The gravity load and a pressure load are applied to the inner surface of the tube of
the soft fingers. The finite-element model is in quantitative agreement with the experimental results
(Extended Data Fig. 1, Extended Data Table 4 and Supplementary Video 1).

Participant recruitments
All experiments were conducted in accordance with the declaration of Helsinki and approved by
the Ethics Committee of Human and Animal Experiments of Shanghai Jiao Tong University. The
transradial amputees participated in this study were recommended by Shanghai Liankang
Prosthetics and Orthotics Manufacturing Co. Ltd, Shanghai, China. The amputees did not have any
prior neuromuscular disorders, and were informed about the experimental procedure and signed the
informed consent forms (ICFs) prior to the participation.

Training process for EMG decoding

During the training process, the subject who performs the predefined grasp types (Fig. 2H) will be
notified by the vibration from a motor placed on the socket. Each grasp type is maintained for 5
seconds and a rest period is required between two grasp types to prevent possible fatigue of the
subject. In the experiments, short-time vibrations are provided at the beginning and end of each

16
grasp type to notify the subject to transit between different hand grasps, and a long-time vibration
is provided at the end of the training session. Each channel (recorded by each EMG sensor) of the
EMG signals from the subject performing a predefined grasp type is segmented into data fragments
with 200-ms windows. Thereafter, a time-domain feature set from the data fragments of each
channel for the predefined grasp type is extracted, including the mean absolute value, waveform
length, zero-crossings, and slope sign changes. The features from different channels for the same
predefined grasp type are cascaded to give a feature vector. Subsequently, the feature vectors
obtained from the training data are fed into the linear discriminant analysis (LDA)40 classifier (as
detailed in the Supplementary Information) to specify the parameters for the predefined grasp types.
The whole training session lasts about 15 min (See Extended Data Fig. 6 and Supplementary Video
4 for a demonstration of the training process).
Although the LDA-based surface EMG decoding algorithms can accurately classify the 5 class
in laboratory settings (Supplementary Fig. 14), the performance will be reduced for amputees in
realistic activities52. Therefore, in many demonstrations of the functions of the soft neuroprosthetic
hand, we choose a single grasp classifier based on testing items of the standardized tests. We have
indicated in the Video Captions either 5 class classifier or a single grasp classifier has been chosen.

Training process for tactile feedback

In the training process for tactile feedback, the stimulation current is set as bi-phasic, rectangular
current pulses. Based on our previous results53, we set the current amplitudes, pulse widths and
frequencies for all five channels as 4 mA, 200 μs and 20 Hz, respectively. During the training
process, we compress each soft finger three times in a random order to check if the subject can
discriminate the compressed finger based on the electrical stimulation. The training process lasts
about 5 min.

Data analysis and statistics.

All data are analyzed using the available built-in functions of MATLAB (R2016b, The Mathworks
Inc., USA) and SPSS (version 22, IBM Inc., USA). All data are reported as mean values with
standard deviations when indicated. In the statistical analyses of standardized tests, the factors
(independent variables) are the neuroprosthetic hand type (the soft neuroprosthetic hand and the
rigid neuroprosthetic hand) and task type (items J2-J7 or S1-S9). The dependent variable is the task
time. All the data are demonstrated normally distributed through the Kolgomorov-Smirnov test (p
> 0.05) before significant analysis. A two-way analysis of variance (ANOVA) is used to statistically
analyze the significant influences of the two factors (neuroprosthetic hand types and task types) on
the task time in the two sessions (items J2-J7 or S1-S9). According to the two-way ANOVA, there
is interaction between the two factors. Bonferroni correction is used to correct for multiple
comparisons. Two-tailed, paired-samples t-test is used to compare the difference of blocks per
minute (in the Box and Blocks Test) or words per minute (item J1) with different neuroprosthetic
hand types.

17
REFERENCES
1. Cordella, F. et al. Literature review on needs of upper limb prosthesis users. Front. Neurosci. 10 (2016).
2. Belter, J. T., Segil, J. L., Dollar, A. M. & Weir, R. F. Mechanical design and performance specifications
of anthropomorphic prosthetic hands: a review. J. Rehabil. Res. Dev. 50, 599 (2013).
3. Lewis, S., Russold, M. F., Diet, H. & Kaniusas, E. Satisfaction of prosthesis users with electrical hand
prostheses and their sugggested improvements. Biomed. Tech. 58, (2013).
4. Biddiss, W. A. & Chau, T. T. Upper limb prosthesis use and abandonment: A survey of the last 25 years.
Prosthet. Orthot. Int. 31, 236-257 (2007).
5. Farina, D. & Aszmann, O. Bionic limbs: clinical reality and academic promises. Sci. Transl. Med. 6,
257ps12 (2014).
6. Hill W. Upper limb prosthetic outcome measures (ULPOM): a working group and their findings. JPO
J. Prosthet. Orthot. 21, P69-P82 (2009).
7. Xu, K., Guo, W., Hua, L., Sheng, X. & Zhu, X. A prosthetic arm based on EMG pattern recognition. In
IEEE Int. Conf. on Robotics and Biomimetics (ROBIO). 1179-1184 (IEEE, 2016).
8. Catalano, M. G. et al. Adaptive synergies for the design and control of the Pisa/IIT SoftHand. Int. J.
Robot. Res. 33, 768–782 (2014).
9. Godfrey, S. B. et al. The SoftHand Pro: functional evaluation of a novel, flexible, and robust myoelectric
prosthesis. Plos One 13, e0205653 (2018)
10. Kanzler, C. M. et al. An Objective Functional Evaluation of Myoelectrically-Controlled Hand
Prostheses: A Pilot Study Using the Virtual Peg Insertion Test. In 2019 IEEE 16th International
Conference on Rehabilitation Robotics (ICORR). 392–397 (2019).
11. Tan, D. W. et al. A neural interface provides long-term stable natural touch perception. Sci. Transl. Med.
6, 257ra138-257ra138 (2014).
12. Raspopovic, S. et al. Restoring Natural Sensory Feedback in Real-Time Bidirectional Hand Prostheses.
Sci. Transl. Med. 6, 222ra19-222ra19 (2014).
13. Clemente, F. et al. Non-Invasive, Temporally discrete feedback of object contact and release improves
grasp control of closed-loop myoelectric transradial prostheses. IEEE Trans. Neural Syst. Rehabil. Eng.
24, 1314–1322 (2016).
14. D’Anna, E. et al. A somatotopic bidirectional hand prosthesis with transcutaneous electrical nerve
stimulation based sensory feedback. Sci. Rep. 7, 10930 (2017).
15. Valle, G. et al. Biomimetic intraneural sensory feedback enhances sensation naturalness, tactile
sensitivity, and manual dexterity in a bidirectional prosthesis. Neuron 100, 37-45 (2018).
16. Osborn, L. E. et al. Prosthesis with neuromorphic multilayered e-dermis perceives touch and pain. Sci.
Robot. 3, eaat3818 (2018).
17. Zollo, L. et al. Restoring tactile sensations via neural interfaces for real-time force-and-slippage closed-
loop control of bionic hands. Sci. Robot. 4, eaau9924 (2019).
18. D’Anna, E. et al. A closed-loop hand prosthesis with simultaneous intraneural tactile and position
feedback. Sci. Robot. 4, eaau8892 (2019).
19. Rus, D. & Tolley, M. T. Design, fabrication and control of soft robots. Nature. 521, 467–475 (2015).
20. Rich, S. I., Wood, R. J. & Majidi C. Untethered soft robotics. Nat. Electron. 1, 102 (2018).
21. Polygerinos, P. et al. Soft Robotics: review of fluid-driven intrinsically soft devices; manufacturing,
sensing, control, and applications in human-robot interaction. Adv. Eng. Mater. 19, 1700016 (2017).
22. Hu, W., Lum, G. Z., Mastrangeli, M. & Sitti, M. Small-scale soft-bodied robot with multimodal
locomotion. Nature. 554, 81–85 (2018).

18
23. Acome, E.et al. Hydraulically amplified self-healing electrostatic actuators with muscle-like
performance. Science. 359, 61–65 (2018).
24. Connolly, F., Walsh, C. J. & Bertoldi, K. Automatic design of fiber-reinforced soft actuators for
trajectory matching. Proc. Natl. Acad. Sci. 114, 51–56 (2017).
25. Cacucciolo, V., Shintake, J., Kuwajima, Y., Maeda, S., Floreano, D. & Shea H. Stretchable pumps for
soft machines. Nature 572, 516-519 (2019).
26. Farrow, N. & Correll, N. A soft pneumatic actuator that can sense grasp and touch. In IEEE Int. Conf.
on Intelligent Robots and Systems (IROS). 2317-2323 (IEEE, 2015).
27. Ferris, D. P. & Lewis, C. L. Robotic lower limb exoskeletons using proportional myoelectric control.
In Annual Int. Conf. Proc. IEEE Eng. Med. Biol. Soc. 2119-2124 (2009).
28. Polygerinos, P. et al. EMG controlled soft robotic glove for assistance during activities of daily living.
In IEEE Int. Conf. on Rehabilitation Robotics (ICORR). 55-60 (IEEE, 2015).
29. Ge, L. et al. Design, Modeling, and evaluation of fabric-based pneumatic actuators for soft wearable
assistive gloves. Soft Robot. Doi: 10.1089/soro.2019.0105 (2020).
30. Deimel, R. & Brock O. A novel type of compliant and underactuated robotic hand for dexterous grasping.
Int. J. Robot. Res. 35, 161–185 (2016).
31. Scharff et al. Towards behavior design of a 3D-printed soft robotic hand. Soft robotics: rends,
applications and challenges, Springer, 2017.
32. Zhao, H., O’Brien, K., Li, S. & Shepherd R. F. Optoelectronically innervated soft prosthetic hand via
stretchable optical waveguides. Sci. Robot., 1, 7529 (2016).
33. Zhou, J. et al. BCL-13: A 13-DOF soft robotic hand for dexterous grasping and in-hand manipulation.
IEEE Robot. Autom. Lett. 3, 3379–3386 (2018).
34. Zhou, J. et al. A Soft-robotic approach to anthropomorphic robotic hand dexterity. IEEE Access. 7,
101483-95 (2019).
35. Emerson, E. T., Krizek, T. J. & Greenwald, D. P. Anatomy, physiology, and functional restoration of
the thumb. Ann. Plast. Surg. 36, 180–191 (1996).
36. Gustus, A., Stillfried, G., Visser, J., Jörntell, H. & Smagt, P. Human hand modelling: kinematics,
dynamics, applications. Biol. Cybern. 106, 741–755 (2012).
37. Keplinger, C. et al. Stretchable, transparent, ionic conductors. Science. 341, 984–987 (2013).
38. Yang, C. & Suo, Z. Hydrogel ionotronics. Nat. Rev. Mater. 3, 125–142 (2018).
39. Gu, G. et al. Integrated soft ionotronic skin with stretchable and transparent hydrogel-elastomer ionic
sensors for hand-motion monitoring. Soft Robot. 6, 368–376 (2019).
40. Englehart, K. & Hudgins, B. A robust, real-time control scheme for multifunction myoelectric control.
IEEE Trans. Biomed. Eng. 50, 848–854 (2003).
41. Feix, T., Romero, J., Schmiedmayer, H., Dollar, A. M. & Kragic, D. The GRASP taxonomy of human
grasp types. IEEE Trans. Hum.-Mach. Syst. 46, 66–77 (2016).
42. Bullock, I. M. et al. Grasp frequency and usage in daily household and machine shop tasks. IEEE Trans.
Haptics. 6, 296–308 (2013).
43. Kyberd, P. J., Evans, M. & Winkel, S. An intelligent anthropomorphic hand, with Automatic Grasp.
Robotica. 16, 531–536 (1998).
44. Hammock, M. L., Chortos, A., Tee, B., Tok, J. & Bao Z. 25th anniversary article: The evolution of
electronic skin (e-skin): a brief history, design considerations, and recent progress. Adv. Mater. 25,
5997–6038 (2013).
45. Resnik, L. & Borgia, M. Reliability and validity of outcome measures for upper limb amputation. JPO
J. Prosthet. Orthot. 24, 192–201 (2012).

19
46. Phillips, S. L., Harris, M. S., Koss, L. & Latlief, G. Experiences and outcomes with powered partial
hand prostheses: A case series of subjects with multiple limb amputations. JPO J. Prosthet. Orthot. 24,
5 (2012).
47. Smit, G., Plettenburg, D. & Helm, F. The lightweight delft cylinder hand, the first multi-articulating
hand that meets the basic user requirements. IEEE Trans. Neural Syst. Rehabil. Eng.23, 431-440 (2015).
48. Fougner, A. et al. Control of upper limb prostheses: Terminology and proportional myoelectric control-
A review. IEEE Trans. Neural Syst. Rehabil. Eng. 20, 663-677 (2012).
49. Hahne, J. M. et al. Simultaneous control of multiple functions of bionic hand prostheses: Performance
and robustness in end users. Sci. Robot. 3, eaat3630 (2018).
50. Zhuang, K. Z. et al. Shared human-robot proportional control of a dexterous myoelectric prosthesis. Nat.
Mach. Intell. 1, 400-411 (2019).
51. Jones, D. R. H. & Ashby, M. F. Engineering Materials 1: An Introduction to Properties, Applications
and Design (Butterworth-Heinemann, Amsterdam; Boston, 4 edition. (2011).
52. Farina, D. et al. The extraction of neural information from the surface EMG for the control of upper-
limb prostheses: emerging avenues and challenges. IEEE Trans. Neural Syst. Rehabil. Eng. 22, 797-809
(2014).
53. Chai, G. et al. Characterization of evoked tactile sensation in forearm amputees with transcutaneous
electrical nerve stimulation. J. Neural Eng. 12, 066002 (2015).

Data availability
The datasets generated or analyzed during the current study are available from the corresponding
author on reasonable request.

Acknowledgments: The research was supported in part by the National Natural Science Foundation
of China (grant nos. 91948302 and 51620105002), Shanghai Jiao Tong University Scientific and
Technological Innovation Funds (grant no. 2019QYB08) and Massachusetts Institute of
Technology. We thank the subjects for the agreements to this research. We thank M. Feng, Z. Shen
and N. Ding for the participations of building the experimental setups, and Q. He for the discussions
of the model and simulation. All data needed to evaluate the conclusions are presented in the paper
or the supplementary materials.

Author Contributions
G.G., N.Z., X.Y.Z and X.H.Z. conceived the idea and designed the study. X.H.Z. and G.G. proposed
the decoupled design of the hand and the waist bag. G.G., N.Z. H.X., H.Y, Q.S., X.Y.Z. and X.H.Z.
performed experiments and analyzed the experimental data. Y.Y., G.C., X.S. and X.Y.Z. developed
the EMG sensors and electrical stimulation platform. G.G., S.L., L.G. and X.H.Z. developed the
theoretical model and performed the FEM simulation for verification. X.H.Z, G.G. and X.Y.Z.
directed the project. G.G., X.H.Z., N.Z., H.X. X.Y.Z. prepared the manuscript and all authors
provided feedback and agree with the final version of the manuscript.

Competing Financial Interests

G.G., N.Z., H.X., S.L. X.Y.Z., and X.H.Z. are inventors of a patent application that covers the
design and fabrication of the soft neuroprosthetic hand.

20
Figures and Figure Captions

Fig. 1 | Design and operation of the soft neuroprosthetic hand. A, Schematic illustration of the
soft neuroprosthetic hand mounted on a transradial amputee with a waist bag. The soft
neuroprosthetic hand consists of five soft fingers and a palm, four electromyography sensors that
measure the surface EMG signals of residual forearm muscles to control the hand, and five
hydrogel-elastomer capacitive sensors on the fingertips that measure touch pressure and elicit
electrical stimulation on the skin of the residual limb. B, Working principle of a soft finger made of
a fiber-reinforced elastomeric tubular structure, in which three rigid segments with specific lengths
are embedded to mimic the soft-joint/rigid-bone anatomy of the human finger. Our design objective
is to generate the desired bending angles (α , α , α ) of the flexible joints by tuning their lengths
(LD, LP, LM). C, Schematic illustration of the soft neuroprosthetic hand with six active DoF motions
and five soft capacitive touch sensors on fingertips. D, Photograph of a transradial amputee wearing
the soft neuroprosthetic hand to grab a cupcake. E, Simulation of the finite-element model and
experimental result of a soft finger under an applied pneumatic pressure of 120 kPa.

21
Fig. 2 | Performance characterization of the soft neuroprosthetic hand. A-D, Measurement and
results of the ranges of motions for 1-DoF flexion fingers in (A-B), and for 2-DoF opposable thumb
in (C-D). Dot markers and shaded areas represent the average and standard deviation for n = 3
measurements at each data point; dash lines represent the finite-element model’s predictions. Sim.
and Exp. are abbreviations for simulation and experiment, respectively. MCP, PIP, DIP, CMC, and
IP are abbreviations for metacarpophalangeal, proximal interphalangeal, distal interphalangeal,
carpometacarpal and interphalangeal, respectively. E-F, Load capability tests of the soft
neuroprosthetic hand by grasping a 55 mm-diameter cylinder at 100 kPa pneumatic pressure to the
1-DoF flexion fingers and 80 kPa pneumatic pressure to the 2-DoF thumb. G, Pressure-flexion
hysteresis curves for the soft finger in the 1st, 10th, 100th, 1,000th and 10,000th cycles of actuations
with the actuation frequency of 0.2 Hz. H, The four-channel EMG signals from the residual forearm
muscles to decode the grasp intention (e.g., the four grasp types of Power, Precision disk, Tripod
and Lateral pinch, and rest).

22
Fig. 3 | A transradial amputee wearing the soft neuroprosthetic hand restores the versatile
hand functions in daily activities. A, Evaluation of the soft neuroprosthetic hand with a set of
standardized tests, including the Box and Blocks Test (e.g., counting the number of blocks per
minute), all seven tasks in the Jebsen-Taylor Hand Function Test (e.g., J1: writing, J2: simulated
page-turning, J3: lifting small common objects, J4: simulated feeding, J5: stacking checkers, J6:
lifting large light objects, and J7: lifting large heavy objects), and nine selected tasks of the
Southampton Hand Assessment Procedure (e.g., grasping nine kinds of objects, such as S1:
spherical light, S2: spherical heavy, S3: tripod light, S4: power light, S5: power heavy, S6: tip light,
S7: tip heavy, S8: extension light, and S9: extension heavy). Values in panel represent the mean
and the standard deviation (n = 3). A p value less than 0.05 (i.e. p < 0.05) is considered statistically
significant. B, Photographs of grasping and manipulating commonly used items in daily activities,
such as food (e.g., cakes and strawberries), commodities (e.g., clothes, bags, water glasses, and
tissues), and tools (e.g., pliers), and safe interaction with environment (e.g., shaking a hand, petting
a cat, and touching a flower). C, Photographs of carrying out delicate tasks to handle objects with
complex shapes and different sizes and then insert them in the corresponding slots precisely. D,
Photograph of lifting about 2.3 kg payload (The presented image is one of three experimental trials).

23
Fig. 4 | A transradial amputee wearing the soft neuroprosthetic hand restores the primitive
touch sensation and the closed-loop control in blindfolded and acoustically-shielded
interaction experiments. A, Electrical stimulation on five regions (Supplementary Fig. 9) on the
residual limb of the amputee corresponding to the effective touch pressures above a threshold
measured by touch sensors on five fingertips, respectively. B, Demonstration of the touch sensation
of any individual finger or multiple fingers being compressed (the threshold ∆𝐶/𝐶 = 0.1). C,
Photographs of the amputee that grasps a bottle, senses the touch pressure and lifts it up, or grasps
nothing and does not lift up. D, Demonstration of the closed-loop control capability of the soft
neuroprosthetic hand enabled by integrating the myoelectric control and tactile feedback.

24
Extended Data Figures and Tables

Extended Data Fig. 1 | Simulation of the fine finite-element model and experimental results of
the soft neuroprosthetic hand under different applied pneumatic pressures. A, Photographs of
1-DoF flexion of the index finger (as an exemplary motion of normal fingers) under pressures from
0 kPa to 120 kPa. B, Photographs of 2-DoF flexion of the thumb under pressures from 0 kPa to 80
kPa.

25
Extended Data Fig. 2 | Schematic illustration of two kinds of soft neuroprosthetic hands and
their pneumatic control schemes. A, The pumps, valves, electronic boards and battery (the
rechargeable lithium battery with the capacity of 800 mAh and the weight of 67 g by Geshi Inc.,
China) are contained in a small bag (length: 240 mm, width: 80 mm, height: 110 mm; weight: 444
g). B, The pumps, valves, electronic boards and battery are integrated in the palm and socket.

26
Extended Data Fig. 3 | Block diagram of the myoelectric control interface in the soft
neuroprosthetic hand. The myoelectric control interface is designed for intuitive control of the
soft neuroprosthetic hand. The myoelectric control is achieved by a customized onboard
measurement and control system consisting of four-channel EMG sensors, control unit (including
the signal processing unit for EMG decoding and the micro-controller for pneumatic actuation),
pumps, valves, solid-state relays, and the power (battery and voltage regulators). The four-channel
EMG sensors (embedded in the socket and mounted on the skin of residual forearm muscles) record
the muscle activities of amputees, which is processed by the readout electronics with the
amplification and Butterworth filtering (20-450 Hz). The signal processing unit receives the
amplified and filtered signals from EMG sensors and classifies the signals into several discrete
classes related to the grasp types of amputees’ intention. Through a universal asynchronous
receiver/transmitter (UART) port, the classification results are sent to a micro-controller (Nano,
Arduino Inc., Italy). The micro-controller employs the classified grasp types to control the pumps
and valves through two solid-state relays, resulting in the intuitive control of the soft neuroprosthetic
hand. The pins (D2-D7) and pins (D8-D13) connect the output pins of the micro-controller relating
to the corresponding pins of pump and valve relays.

27
Extended Data Fig. 4 | Fabrication and assembly of the soft neuroprosthetic hand. A, Molding
of inner elastomeric tubes (fingers, thumb, thumb-palm connection). B, Vacuum defoaming. C,
Cured inner tubes. D, Winding of the finger/thumb. E, Carbon fiber-reinforced plastics (CFRP)
laminates attachment. F, Sleeve wrapping. G, Making joint segments of the sleeve. H, Molding of
the outer elastomeric tube of the finger/thumb. I, Limiting layer attachment of the thumb-palm
connection. J, Winding of the connection pad. K, Molding of the outer tube of the thumb-palm
connection. L, Cured finger/thumb, thumb-palm connection. M, Terminal connectors. N, The 3D-
printed palm skeleton. O, Assembly of the thumb with the thumb-palm connection. P, Assembly of
the opposable thumb and fingers to the palm skeleton. Q, Integrating capacitive touch sensors on
the fingertips. R, Installing the pumps, valves, electrical boards and battery in a waist bag. S,
Installing the pumps, valves, control boards and battery in the palm and socket. T, Connecting the
palm skeleton with the socket to form a soft neuroprosthetic hand.

28
Extended Data Fig. 5 | The four EMG-controlled grasp types (excluding rest) for the soft
neuroprosthetic hand and the kinematic relationship in each finger to achieve the specific
grasp type. A, the six most frequently used grasp types of human hands in daily activities42. B, The
regrouped four grasp types in the soft neuroprosthetic hand based on the results in (A). C, Kinematic
relations of different fingers in the four predefined grasp types in (B).

29
Extended Data Fig. 6 | Still images of a subject wearing the soft neuroprosthetic hand in the
training process for intuitive control. A transradial amputee can quickly adapt to the soft
neuroprosthetic hand and master its functionality by training. For the training algorithms, please
refer to the Methods section. The red button is used for switching the power of the hand on/off and
the silver button for switching to the training mode. As shown in the figure and supplementary
Video 4, we can see that a subject can rapidly put on the soft neuroprosthetic hand within 3 seconds
and master its function to intuitively control it after about 1 min. The training process is repeated
for about 15 min.

30
Extended Data Fig. 7 | Descriptions of the standardized tests to evaluate the function of the
soft neuroprosthetic hand. A, Testing tasks, including the Box and Blocks Test (BBT), all seven
tasks in the Jebsen-Taylor Hand Function Test (i.e., J1-J7), and nine selected tasks in the
Southampton Hand Assessment Procedure (i.e., S1-S9). B and C, Weights, dimensions and still
images of the 17 objects used in the standardized tests.

31
Extended Data Fig. 8 | Evaluation of the soft neuroprosthetic hand on another transradial
amputee with the standardized tests (the same as in Fig. 3A), including the Box and Blocks Test
(e.g., counting the number of blocks per minute), all seven tasks in the Jebsen-Taylor Hand Function
Test (e.g., J1-writing, J2-simulated page-turning, J3-lifting small common objects, J4-simulated
feeding, J5-stacking checkers, J6-lifting large light objects, and J7-lifting large heavy objects), and
nine abstract tasks of the Southampton Hand Assessment Procedure (e.g., grasping nine kinds of
objects, such as S1-spherical light, S2-spherical heavy, S3-tripod light, S4-power light, S5-power
heavy, S6-tip light, S7-tip heavy, S8-extension light, and S9-extension heavy). Values in panel
represent the mean and the standard deviation (n = 3). A p value less than 0.05 (p < 0.05) is
considered statistically significant.

32
Extended Data Fig. 9 | Demonstration of the advantage of the soft neuroprosthetic hand in
handling fragile objects such as a strawberry, a piece of bread and a paper cup filled with
water compared to the rigid i-Limb hand. A, the soft neuroprosthetic hand. B, the i-Limb hand.
Each set of tests have been performed with three experimental trials and the presented images are
from one set of experiments (Supplementary Video 7).

33
Extended Data Fig. 10 | Experimental setup and results of the subject wearing the soft
neuroprosthetic hand to discriminate three cylinders with different diameters. In this test, we
program the stimulation frequencies of the electrical pulses based on the different ranges of ∆C/C0
of the touch sensor on the middle finger (i.e., no stimulation when ∆C/C0≤0.1, 5 Hz when 0.1
<∆C/C0 ≤ 0.3, 20 Hz when 0.3<∆C/C0≤0.4, and 35 Hz when ∆C/C0>0.4). The statistical results
demonstrate that the subject can correctly discriminate the grasped subject with an accuracy of
96.25% (77 successes in all 80 tests).

34
 Extended Data Table 1 | Comparison of the soft neuroprosthetic hand with reported soft
hands mainly composed of elastomers.

Note: *see the website (https://www.festo.com/group/en/cms/13508.htm) for detailed information.

35
Extended Data Table 2 | Detailed information of the mechanical components and electrical
components of the soft neuroprosthetic hand.
Dimension Weight Price
Component Materials and parameters of each component
(mm3) (g) (USD)

Soft neuroprosthetic hand (total weight of 292 g)

Dragonskin 10 for inner elastomeric tube. Ecoflex

Thumb 100×24×20 26 ~3
0030 for outer elastomeric tube. CFRP plates from
Index, middle, ring, 110×24×20 112 TORAY (3K), Polyethylene thread from Yunshangpiao
~12
and little fingers (per piece) (4 pieces) Co. Ltd are attached to the inner tube.
Dragonskin 10 for inner elastomeric tube. Ecoflex
Thumb-palm 0030 for outer elastomeric tube. Polyethylene thread
75×50×20 41 ~3
connection from Yunshangpiao Co. Ltd are attached to the inner
tube.
Palm skeleton is 3d printed with Imagine 8000
Palm with skin 94×78×132 110 ~22
(SOMOS). Skin is fabricated with Ecoflex 0030.
Material: pAAm ionic hydrogel and dielectric
Hydrogel-elastomer 3
13×13×1.9 elastomer. ~5
sensors (5 pieces)
Linearity: ~0.94 and sensitivity: ~0.016 kPa-1.

Hardware of pneumatic control systems in a waist bag (total weight of 444 g)

Pneumatic Control Arduino (Nano).

45×18×7.2 5 ~22
unit 5-12 V, 12-channel input/output pins
MITSUMI (MAP-AM-265).
312
Pump 67×27×27 Maximum pressure: 90 kPa; ~30
(6 pieces)
Flow rate: 2.0 LPM; Voltage: 6 V; Noise: 55 dB
30 Parker (X-valve)
Valve 23×12×8 ~210
(6 pieces) 2-way normal closed, directional; Voltage: 12 V
14 Customized.
Relay 54×30×5 ~20
(2 pieces) Voltage: 12 V; 6 channels
6 Customized.
Voltage regulator 18×10×2.6 ~6
(6 pieces) Input voltage range: 4-36 V
Geshi.
Battery 57×29×22 67 ~10
Capacity: 800 mAh; Voltage: 11.1 V (~12 V)
Signal processing Customized.
unit for EMG 30×30×8 10 Voltage: 3.3 V; A/D: 12 bit. ~22
sensors Bandpass filter: 20-450 Hz

Customized socket (total weight of 375.5 g)

Body of the socket 274×95×90 335.5 Customized

40 Customized.
EMG sensors 35×22×10 ~100
(4 pieces) Material: gold-plated copper

Appendants

Pneumatic tube (PUN-H-3x0.5) from Festo Co. Ltd and electrical wire from ZUOYOU ZHONGGONG company. ~3

36
 Extended Data Table 3 | Comparison of the soft neuroprosthetic hand with representative
commercially-available neuroprosthetic hands. (A) Comparison of the weight. (B) Comparison
of the actuation mechanism, active DoFs, the number of joints, and bending angles of the joints.

Note: NA and NP are abbreviations for not applicable and not provided, respectively.

37
Extended Data Table 4 | Relative prediction errors of the finite-element model for the 1-DoF
finger and 2-DoF thumb.

Note: The relative error e is calculated by 𝑒 100%, where 𝜃 and 𝜃 are

bending angles of the finite-element model and the average experimental results (n = 3),
respectively. Sim and Exp are abbreviations for simulation and experiment, respectively.

38
Extended Data Table 5 | Comparisons of the performances of the soft neuroprosthetic hand
and a conventional rigid neuroprosthetic hand7 on the same subject evaluated with the
standardized tests. (Each item in the standardized tests is performed three times and the mean
values represent the average for n = 3 measurements.) Significant results (p <0.05) are boldface.

39
Extended Data Table 6 | Comparisons of the performances of the soft neuroprosthetic hand
and a conventional rigid neuroprosthetic hand7 on another subject evaluated with the
standardized tests. (Each item in the standardized tests is performed three times and the mean
values represent the average for n = 3 measurements.) Significant results (p < 0.05) are boldface.

40
Description of Supplementary Videos in Supplementary Information
Supplementary Video 1:
Simulation and experiments of individual motion of five soft fingers.
Supplementary Video 2:
Demonstration of independent control of six DoFs with one pump.
Supplementary Video 3:
Demonstration of the durability of a soft finger.
Supplementary Video 4:
Demonstration of fast wearing and training of a soft neuroprosthetic hand. In this video, the EMG
decoder uses a five class classifier for all the four grasps (Power, Precision disk, Tripod and Lateral
pinch) and rest type.
Supplementary Video 5:
Evaluation of the soft neuroprosthetic hand with the standardized tests. In this video, the EMG
decoder uses a two class classifier for the single grasp type and rest type according to different items
of the standardized tests (For a fair comparison, the EMG decoder is same for the rigid
neuroprosthetic hand as shown in Supplementary Video 6). Specifically, the “Power” grasp type is
used for the items J1, J6, J7, S1, S2, S4 and S5; the “Precision disk” grasp type is used for the items
J2, S8 and S9; the “Tripod” grasp type is used for the Box and Blocks Test, items J3, J5, S3, S6 and
S7; the “Lateral pinch” grasp type is used for the item J4.
Supplementary Video 6:
Experimental results of the standardized tests by the same subject wearing a rigid neuroprosthetic
hand. In this video, the EMG decoder uses a two class classifier for the single grasp type and rest
type according to the different items of the standardized tests (For a fair comparison, the EMG
decoder is same for the soft neuroprosthetic hand as shown in Supplementary Video 5). Specifically,
the “Power” grasp type is used for the items J1, J6, J7, S1, S2, S4 and S5; the “Precision disk” grasp
type is used for the items J2, S8 and S9; the “Tripod” grasp type is used for the Box and Blocks
Test, items J3, J5, S3, S6 and S7; and the “Lateral pinch” grasp type is used for the item J4.
Supplementary Video 7:
Demonstration of the compliant advantage of the soft neuroprosthetic hand. A subject wearing the
soft neuroprosthetic hand and a rigid i-Limb hand to grasp fragile objects (e.g., strawberry, bread,
paper cup). The rigid neuroprosthetic hand damages the strawberry and bread, and tends to crush
the paper cup. In this video, the EMG decoder uses a two class classifier for the “Precision disk”
grasp type and rest type.
Supplementary Video 8:
Demonstration of the four EMG-controlled grasp types. In this video, the EMG decoder uses a five
class classifier for all the four grasps (Power, Precision disk, Tripod and Lateral pinch) and rest type.
Supplementary Video 9:
Demonstration of versatile hand functions in daily activities of the subject. In this video, the EMG
decoder uses a two class classifier for the single grasp type and rest type according to different tasks
in daily activities. Specifically, the “Power” grasp type is used for handling water glasses, paper
drinks and clothes; the “Precision disk” grasp type is used for shaking hands, petting a cat, handling

41
cakes and dishes; the “Tripod” grasp type is used for gripping cards and chips; and the “Lateral
pinch” grasp type is used for zippering up the bag.
Supplementary Video 10:
Demonstration of handling objects with different shapes and sizes. A subject can successfully carry
out delicate tasks to handle objects with complex shapes and different sizes and then insert them in
the corresponding slots precisely. In this video, the EMG decoder uses a two class classifier for the
“Tripod” grasp type and rest type.
Supplementary Video 11:
Demonstration of holding heavy payloads. In this video, the subject uses his own EMG signals to
control the soft neuroprosthetic hand to give the “Power” grasp type for lifting a payload of 2.3 kg.
The EMG decoder uses a two class classifier for the “Power” grasp type and rest type.
Supplementary Video 12:
Demonstration of the touch sensation of individual finger and multiple fingers. In a blindfolded and
acoustically-shielded interaction experiment, an experimenter gently compressed the five fingers of
the soft neuroprosthetic hand in random combinations. The subject could almost instantaneously
distinguish any individual finger or multiple fingers being compressed.
Supplementary Video 13:
Demonstration of the closed-loop control capability of the subject. In a blindfolded and acoustically-
shielded experiment, the subject uses his own EMG signals to control the soft neuroprosthetic hand
to give the “Power” grasp type. If the hand firmly grasps a bottle so that the effective pressures on
the five fingertips are above the threshold, the subject is informed by the electrical stimulators to
lift up the bottle. In contrast, if the hand does not grasp anything that applies pressure on the
fingertips, the subject does not lift up but relaxes the hand after a few seconds. The EMG decoder
uses a two class classifier for the “Power” grasp type and rest type.
Supplementary Video 14:
Demonstration of the graded tactile feedback of the subject. In the blindfolded and acoustically-
shielded experiment, the subject uses his own EMG signals to control the soft neuroprosthetic hand
to give the “Power” grasp type. According to the stimulation frequencies (i.e., 5 Hz, 20 Hz or 35
Hz) of the electrical pulses, the subject can discriminate three cylinders with different diameters
(i.e., 60 mm, 70 mm and 80 mm). The EMG decoder uses a two class classifier for the “Power”
grasp type and rest type.

42