Reconfigurable Soft Pneumatic Actuators Using Extensible Fabric-Based Skins

Abstract

The development of the field of soft robotics has led to the exploration of novel techniques to manufacture soft actuators, which provide distinct advantages for wearable assistive robotics. One subset of these soft pneumatic actuators is conventionally developed from silicone, fabrics, and thermoplastic polyurethane (TPU). Each of these materials in isolation possesses limitations of low-stress capacity, low-design complexity, and high-input pressure requirements, respectively. Combining these materials can overcome some limitations and maintain their desirable properties. In this article, we explore one such composite design scheme using a combination of silicone polymer-based bladder and reconfigurable fabric skin made from an anisotropic extensible fabric. The silicone polymer bladder acts as the hermetic seal, while this skin acts as the constraint. Bending and torsional actuators were designed utilizing the anisotropy of these fabrics. The torsional actuator designs can achieve over 540° of twist, significantly larger than previously reported in the literature, owing to the lower mechanical impedance of the extensible fabrics. Actuators with 360° of bending were also fabricated using this method. In addition, the lack of TPU-backed or inextensible fabrics reduces the actuator's stiffness, leading to lower actuation pressures. Skin-based designs also confer the advantage of modularity, reconfigurability, and the ability to achieve complex motions by tuning the properties of the bladder and the skin. For applications with high-force requirements, such as wearable exoskeletons, we demonstrate the utility of multilayer design schemes. A multilayer bending actuator generated 190 N of force at 100 kPa and was shown to be a candidate for wearable assistive devices. In addition, torsional designs were shown to have utility in practical scenarios such as screwing on a bottle cap and turning knobs. Thus, we present a novel fabric-skin-based design concept that is highly versatile and customizable for various application requirements.

Introduction

The field of soft robotics involves using compliant materials, mechanisms and structures to develop naturally compliant robots. Various methods of achieving compliance have been used, including tendon drives,^1–4 pneumatic muscles^5,6 or soft pneumatic actuators,^7–10 hydraulic actuators,¹¹ as well as entirely self-powered soft robots. Soft pneumatic actuators are gaining popularity due to their varied applications in robotics for health care,^12–18 soft manipulation,^19–21 and bioinspired locomotion.^22–26 Conventional pneumatic soft pneumatic actuators are mainly manufactured from three materials: silicone,^27–29 fabrics,^30–34 or 3D printed thermoplastic polyurethane (TPU).^35–37 Each of these materials possesses specific salient characteristics but with certain limitations.

Silicone elastomers have been a popular material of choice for soft robotics. They are known to generate the most strain or deformation. However, their relatively lower Young's Modulus cannot handle high pressure. This is because they develop high strain at low pressures, meaning that they need some form of strain limiting to survive at higher pressures. Various methods have been used to constrain them, such as using fiber reinforcement,³⁸ but these often require high pressures to actuate.

TPU possesses a higher Young's Modulus than most fabrics and tends to be much stiffer. As a result, the TPU-based actuators can provide much larger force output but require much higher operating pressures.^36,39 It was shown that higher stiffness, in general, tends to cause a loss of force output due to losses such as elastic energy storage.³⁹ The authors used the same design but made with different TPUs: one of shore hardness 60A and lower elastic modulus, and another with shore hardness 85A and higher elastic modulus. The softer actuator was seen to generate higher tip force. This is due to less energy in the softer actuator being stored as elastic potential energy.

The other limitation with using entirely 3D printed TPU actuators is the printable wall thickness of fused deposition modeling (FDM) 3D printers. FDM printers tend to be able to print a wall no smaller than half the nozzle diameter, which is around 0.2 mm thick. However, printers differ in finishing quality and the walls may not be perfect hermetic seals. Thus, the actuators often need to be printed at a wall thickness upwards of 1 mm, adding to the stiffness and the operating pressure.

Fabric-based actuators have been a much more recent development, with research groups typically using TPU-backed fabrics for making soft actuators. The TPU serves two purposes: it acts as a hermetic seal and provides a mechanism for sealing the actuators. Usually, using ultrasonic welding or a heat press, the TPU-to-TPU contacts can be heat sealed to create pouch-like actuators. The actuators have been shown to handle pressures up to 150 kPa.^40,41 The typical failure mode, however, involves the rupture of the TPU-TPU hermetic seal.

Composites have been used in engineering to overcome individual material limitations while maintaining their beneficial properties for centuries. Soft robotic reconfigurability has been a recent trend. Shape memory alloy-based skins called OmniSkins were also shown to be applicable for multiple applications.⁴² Silicone-based printing⁴³ shows another promising approach to tuning soft actuators postfabrication, however, it is currently quite complex to implement. In this article, we explore the idea of combining two of the materials to overcome their limitations, while utilizing their inherent strengths. We combine an anisotropic fabric with an internal silicone bladder, showcasing the utility of fabric anisotropy in programming macroscopic deformation in soft pneumatic actuators.

The idea of using composites creates a new design as the fabrics used no longer need to be TPU-backed; we can use fabrics that enable high deformation or force as per the requirement. Recent studies have moved to exploring composite design schemes, including TPU and fabric combination designs.^44,45 These designs focused on using inextensible textile to improve force generation. In this article, we explore design schemes that use an extensible fabric, Kinesiology Tape. This type of fabric provides versatility in manufacturing soft actuators such for bending as well as torsion, due to its inherent extensibility. In addition, we demonstrate that our designs allow for a degree of customizability and reconfiguration after fabrication.

This ability to tune properties can provide advantages in situations where the soft actuator is being used to grasp or manipulate objects of different shapes or sizes, or in cases where the force requirements change based on application. Specifically, we will demonstrate that our proposed design scheme can achieve exceptional levels of deformation, as demonstrated by a torsional actuator or high force output as demonstrated by a bending actuator. Thus, we present a versatile framework for fabrication of reconfigurable soft pneumatic actuators.

Methods

Concept of soft skins: the advantage of using multiple materials in soft actuators

Each material used for conventional soft robots has a specific set of advantages and limitations. These limitations arise due to the material properties and the lack of maturity of fabrication methods. The soft skin-based designs presented herein are intended to overcome the individual limitations of these materials using the material and manufacturing methods currently available to manufacture them. Figure 1a shows a set of properties that soft pneumatic actuators require to be of practical use and where conventional actuators place on this framework. Fabric-based skins utilize a composite design scheme to overcome some of the limitations of single material actuators, although the force versus deformation trade-off persists.

FIG. 1.

(a) Comparison of the pros and cons of conventional materials used for fabrication of soft robots and advantage of fabric-based skins. Fabric-based skins will be shown to achieve one of two, either exceptionally high deformation or large forces, both at moderate pressures without compromising durability. (b) The underlying concept of a fabric-based skin is to use a high-stiffness external skin and a low-stiffness internal bladder. The bladder acts as a hermetic seal while the skin acts as a deformation constraint. (c) Tensile properties of Kinesiology Tape (KTape) in the warp and weft directions. The high-stretch weft helps create motion, while the low-stretch warp direction acts as a constraint and prevents radial bulging.

Figure 1b describes the based concept of a skin-based actuator. The properties of the skin can be programmed using two methods- using materials of varying properties or utilizing varied geometries. In this article, we will explore the former, using anisotropic fabrics as the basis.

We explored a variety of fabrics to manufacture the actuators, including spandex, nylon, and cotton and settled on using Kinesiology Tape (KTape) as the primary fabric of choice. This is due to the ability of KTape to provide large deformations and still be able to take enough stress to fabricate a practically usable actuator. Another important trait is the natural anisotropy of the KTape, which will be explained in the following sections. This anisotropy can be exploited to create designs that can achieve specific motions, such as bending or torsion.

Designs of soft skins

Fabric-based soft skins shown here are designed using KTape (Tarmak-Decathlon, Lille, France). KTape was chosen for its specific anisotropic properties: it is naturally extensible in the weft direction to around a 110% strain, and <20% strain in the warp direction as shown in Figure 1c. These data were acquired using a tensile Test on an Instron machine as per the ASTM D5035 standard for textiles using five samples. In addition, the KTape is observed to be almost twice as strong in the warp direction, only failing at 180 N of tensile load. In the following sections, we utilized these properties to develop a bending and torsional actuator. The low-strain warp direction acts as a practical radial constraint, preventing the inner bladder from expanding radially, whereas the weft direction allows for the deformation and force transfer required to make these soft pneumatic actuators usable in practice.

Bending actuators

Bending is the most used motion mode for soft robots. It has applications from rehabilitation^46–48 and wearable robots^44,49 to grippers^50,51 and manipulation robots,^51–54 as it is fundamental in joint-like behavior. In addition, bending is an essential primitive for locomotion^22–26 in robots. Figure 2a and b show two distinct design schemes for the bending skins, involving a circular and a semicircular cross-section tube, respectively.

FIG. 2.

Bending actuator design schemes. (a) Circular cross-section design involves two layers of KTape wrapped over each other (Steps 1 and 3) with a strain limiter to increase bending moment in between (Step 2). (b) Semicircular Tube design involves a single-layer wrap with an overlap over the base. The inner bladder may also be designed with a layer of stiffer material on the base to increase bending further. (c) The deformation of the actuator shown in (b) when pressurized.

The first scheme shown in Figure 2a uses a circular cross-section bladder and three fabric layers. The first layer is a KTape (extensible fabric) layer. The direction of the wrapping is such that the stretchable (weft) is along the axis of the actuator. A first layer of KTape is wrapped around the inner bladder with the adhesive backed side outwards. This allows for the skins to be removable. This is followed by an inextensible fabric layer, such as nylon which covers half of the diameter of the extensible layer. This forms the strain limiting constraint.

The third layer is another extensible layer, this time wrapped with its adhesive side inwards. This creates a strong adhesive-to-adhesive contact between layers one and three, improving the actuator's resistance to layer shearing. The two extensible layers impart the ability for the actuator to deform, whereas the half-diameter inextensible layer acts as a strain limiter making the actuator bend toward it. It is recommended to have overlaps on the extensible layers to improve shear resistance, and having these overlaps align with the inextensible layer can enhance bending.

The second design, as shown in Figure 2b uses a bladder of half-circular cross section. The skin is formed similar to Figure 2a, using two layers of KTape with the weft along the axis of the actuator. The firs layer is wrapped with the adhesive side outwards, overlapping over the flat side of the half-circular bladder. The second layer is then wrapped over, with the overlap on the apex of the actuator. A strain limiting fabric layer is not used, but may be used in case more extreme bending is required.

Figure 2c shows the design in Figure 2b in action. We will subsequently demonstrate the utility of this device as an assistive actuator for the wrist.

Torsional and helical actuators

The fundamental idea behind the design of the torsional actuator is the inclination of the preferential direction of extension of the fabric at an angle with the axis of the internal bladder. This allows the actuator to twist without significant radial bulging. The bladder in these designs is fabricated first using a moulding method similar to that described in Ambrose et al.⁵⁵

Figure 3a shows the basic single-layer torsional actuator design. The KTape is then wrapped around a metal rod of diameter 2 mm more than the outer diameter of the bladder, with the adhesive-backed side facing outward to ensure that the fabricated skin can easily slide onto the bladder with minimal friction and adhesion. The width of the KTape used is 5 cm, and a 1 cm overlap is used between consequent wraps, as empirically, this gave the best twist angle without sacrificing durability. Markings on the metal rod help orient the fabric at the desired angle (45°) and the rod itself acts as a template providing consistency in the radius of the fabricated skin.

FIG. 3.

Designs of torsional and helical actuators. (a) Single-layer helical actuators can be manufactured using and angle winding pattern of KTape. (b) To increase the durability of the single-layer design a second layer of overlap is added over the first at locations where the tube is prone to be exposed. (c) The single-layer torsional actuator can achieve over 540° of torsional at 70 kPa. (d) Using a bending bladder instead of a homogenous bladder, we can achieve helical motion, a composite motion of bending and twisting.

If durability needs to be further enhanced, the dual layer scheme can be used, in which a 2 cm wide piece of KTape is wrapped over this overlap to provide further resistance to the shearing of the adhesive. This is the dual-layer design as shown in Figure 3b. The second layer in the dual layer scheme can compress the skin slightly making the insertion of the bladder more difficult than the single-layer scheme. However, this can be solved by applying mild vacuum (−40 kPa) on the bladder when inserting.

Figure 3c shows the deformation profile of the single-layer torsional actuator. The actuator achieved over 540° of torsion at a modest pressure of 70 kPa. Supplementary Video S1 shows this actuator in motion. In addition, we can generate helical movement, as shown in Figure 3d. This is done by replacing the homogenous bladder used in the design in Figure 3c and replacing it with a bending bladder. This is also demonstrated in Supplementary Video S1.

Torsional designs with large deformations can find great utility in locomotion robots or increase the effective range of motion of soft grippers or industrial robots. These designs can be scaled up as per requirement, with a caveat that the resistance of the bladder to deformation will increase significantly with thickness.

Reconfigurability and composite motions using multimaterial bladders

The skins have an advantage over most conventional soft actuators as they can be reconfigured. The external skin is removable, and the internal bladder can easily be replaced by one with different properties. Thus, this gives the designer multiple variables to tweak performance; the design of the skin and the design of the bladder, both of which can be tweaked based on geometry and material.

The removability also means that the actuators are effectively reconfigurable. For example, it takes ∼30 s to apply the torsional skin, whereas the other skins easily slide on within 2 s. This is because the torsional skin has a smaller tolerance to ensure force transfer from the bladder, which is much more critical in this case due to the design needing to overcome significant resistance to twisting. In Supplementary Video S2, we demonstrated this ease of transfer, using a homogenous bladder with a torsional, extension, and bending skin.

Experimental setups and characterization

The following subsection describes two main sets of experiments that were conducted on the fabric-skin-based actuators. In the first set of experiments, we used a LIDAR (Intel RealSense L515; Intel Corp.) to extract the deformation profile of the actuators. Based on the profile, we were able to compute the deformation angles at the tip with respect to the initial position at pressure increments of 5 kPa. This setup also enabled automation of a cyclic test, which was used to evaluate the changes in performance of the actuators over a large set of cycles.

Experimental setups

Light Detection and Ranging instrumentation and processing

The flow of the Light Detection and Ranging (LIDAR) instrumentation, data collection, and processing pipeline is shown in Figure 4. We acquired 25 frames per cycle for a set of 6000 actuator cycles using the LIDAR system. The experiment was synchronized via a Python script with an Arduino Uno controlling an on/off solenoid valve that inflated and deflated the actuator.

FIG. 4.

LIDAR instrumentation and postprocessing pipeline. ROI is selected manually before the start of the experiment. Depth-based segmentation allows isolation of the actuator from the background. Actuator profile can be extracted using contour detection by identifying key points (green). The morphological parameters can then be computed for every cycle, and stored for analysis later in a CSV file. CSV, comma separated values; ROI, region of interest.

Once extracted, the goal of the instrumentation process is to cleanly model the pneumatic actuation as a function of certain key pixels on the image representation of our actuator. Certain methods used earlier to represent terrain maps and 2D shapes involve Binary Skeletonization and the Medial Axis Transform. We use a combination of both these methods to populate a list of points whose frame-wise positional displacement is used to represent the actuation in the frame.

We used the following data processing pipeline for the data acquired from the LIDAR.

Cropped the data from the images using a Region of Interest set before the start of the experiment. This is done so that the saved images are smaller in size.

Background removal is done using depth information to generate a mask.

Actuator key points can then be extracted using contour detection.

The morphology geometries are calculated using standardized parameters as mentioned in the figure. These metrics can be used for checking performance changes over time, such as actuator leaks as well as sensing malfunctions.

The bending angle can be calculated with respect to the initial angle knowing the key points at the start and end of the cycle.

Blocked force measurement

The second set of experiments involves measurement of the blocked force output of the skin-based actuators.

The setup consists of Vertical Automatic Handy Tester (JSV H1000; Measuring Instrument Technology, Singapore) with a force gauge to measure the blocked force of the bending actuators (shown in Fig. 5a). The actuator is anchored on a retort stand. The force gauge has an attached contact pad to ensure that the force is applied at the same location every time. The bottom surface of the actuator is constrained by the platform. The bending actuator bends upwards (against gravity) and transfer the force onto the contact pad. We built the setup in this way to ensure that the force measurements take into account the actuator compensating for its own weight, as opposed to the conventional blocked force setups.^18,56 This compressive force is then measured by the force gauge. We measured the force while increasing the input pressure in 5 kPa increments.

FIG. 5.

(a) Blocked force measurement setup. (b) Blocked force measurement of the two bending skin designs. The semicircular tube design generated almost five times the force as the circular tube design of similar diameter at the same pressure.

Characterization

Pressure versus deformation profile for bending actuators

Deformation profiles are crucial for assistive robots as they determine the range of motion that the robot can assist. For example, the range of motion for the wrist is ∼100° for flexion and 70°–90° for extension.⁵⁷ Thus, significant deformations are needed for the actuation of these joints. Figure 6a shows the deformation profile for the actuators with semicircular and circular cross sections. Each color represents a sample set, and each marker represents a data point. Both actuators were able to achieve the required degree of bending for wrist motion, well over 100° bending at the tip.

FIG. 6.

(a) Pressure deformation profiles of the actuators with the circular and semicircular cross section. The semicircular design shows larger deformations at lower pressure; however, the circular cross-section actuator can tolerate higher pressures and as a result has a higher maximum angular displacement. (b) Pressure deformation profiles of the torsional designs. As expected, the single-layer design shows much larger deformations, but the two-layer design can tolerate higher pressures.

The semicircular cross section actuator starts deforming at a higher pressure as the initial pressure inflates the inner bladder to fit the skin. The semicircular cross section actuator deformed more at the equivalent pressure than the circular cross section. The deformation of this actuator tends to saturate around 80–90 kPa (operating pressure is 70 kPa). However, the force generation still increases, as we will see in the following sections.

Blocked force measurement for bending actuators

The blocked tip force of the actuator can be a representative measure of the assistive force that the actuator may be able to provide, minus the transmission losses. Thus, it is a crucial metric to gauge the practical applicability of the actuators.

We tested two actuator designs for blocked force for bending motion. Figure 5a shows the setup used for the measurement, while Figure 5b shows the data. The shaded bars represent one standard deviation of the sample. The blocked force vs pressure relationship is almost linear, save for the initial 10 kPa, during which the bladder expands to the shape of the skin. This makes sense, given that the area of contact is seen to be almost constant after hitting a moderate pressure (∼20 kPa), and the force transfer increases linearly, as given by the equation Force = Pressure × Area, where F is the measured force on the calibrated load cell. As the actuator is deformable, the surface area of contact may see slight variations at higher pressures compared to lower pressures.

Both designs show an initial region where the force output does not increase much with pressure (up to about 20 kPa). This is due to two factors. First, the bladder requires a certain pressure to deform and overcome the material's elastic potential energy storage loss. Second, the bladder-skin fit is not perfectly tight to aid reconfigurability. Hence, part of the deformation in this range is to allow the bladder to conform to the shape of the skin, via radial expansion. This factor could also explain the larger variance in the circular cross-section tube as it probably leads to a degree of inconsistency in the fit of the skin. One potential improvement in the fabrication of these designs is using an automated manufacturing process.

It was observed that the semicircular cross section bending design generated about five times the amount of blocked force (about 190 N at 100 kPa) at the same pressure as the design with the circular cross section (38.5 N at 100 kPa, 75.3 N at 175 kPa). The circular cross-section actuator generates lower force output as it has more surface area for radial expansion, compared to semicircular cross section. The force output of the semicircular cross-section design is good enough for a wearable assistive device, as we will briefly showcase in the application section. Although it is difficult to compare various designs due to size differences and differences in experimental testing conditions, the fabric skins tend to outperform the force output of soft bending actuators based on fabrics at similar or lower pressures.

Pressure versus torsional angle for torsional skins

We measured the twist angle of the torsional actuator at 5 kPa increments up to the maximum tolerable pressure for the torsional actuator, as shown in Figure 6b. Each color represents a sample set and each marker represents a data point. The stiffness of the two-layer design is higher than the single-layer design, owing to double the amount of fabric layers. This translates to higher mechanical impedance to bending. The lower mechanical impedance of the single-layer design means that this actuator generates a significantly higher deformation than the two-layer design.

However, the silicone tube is exposed at 70 kPa as seen from the failed sample (yellow), and although the actuator does not burst until about 100 kPa owing to the elastic properties of the bladder, 70 kPa is considered the failure point. On the contrary, the two-layer design takes a much larger initial pressure to show significant deformation (about 40 kPa) but can sustain higher pressures. This is also important in the context of durability, where the two-layer skins hold up much better than their single-layer counterparts due to better reinforcement at high-strain locations.

The torsional angle is also a function of the size of the actuator. A larger diameter actuator twists less at the same pressure as a smaller one owing to higher resistance to torsion. However, as expected, its ability to do work improves as it can generate larger torques.

Cyclic test and durability

One of the crucial features of any actuator deployed in practice is durability. If the actuator performance deteriorates significantly in a short number of cycles, or worse, if the actuator fails, its practical utility is severely limited. Therefore, we studied the durability of our bending skin to gauge if there was any longitudinal effect of cycling on the actuator performance. The actuator was subjected to 6000 cycles of inflation and deflation, known as a cyclic test, at its expected operating pressure of 80 kPa. The LIDAR was used to record the actuator deformation every cycle, and the data were postprocessed using a segmentation algorithm, as shown in Figure 4, to extract the actuator profile. We were then able to compute key metrics such as the deformation from the mask of the actuator bending profile.

The cyclic test data show the distribution of the bending angle with respect to the cycle count. The actuator initially starts with a minor deformation which increases slightly over time and stabilizes around cycle 1000, as shown in Figure 7a and b, Figure 7a shows the distribution of the angles from the scatter plot using a violin plot for each set of 1000 cycles. It can be observed that the first set of 1000 cycles contains a longer tail, and these outliers may explain the lower initial angle average. This may be due to a slight error in the measurement by the extraction process, and the algorithm.

FIG. 7.

Actuator characterization using a cyclic test. (a) Distribution of the raw peak bending angles of cycles 1–6000. A slight increase is seen after the initial set of 1000 samples after which the angles stabilize. (b) Violin plots of each set of 1000 cycles show the distribution to be similar. Thus, the actuator performance does not degrade significantly with cycle count in the demonstrated range.

Nevertheless, the violin plots show a consistent distribution of angles over each set of 1000 cycles up to 6000 cycles, suggesting that the actuator's performance does not change significantly over this duration of testing. Thus, we can consider the actuator quite durable over the duration studied herein.

Applications

We used the design methodology described above to build a wrist assistive device. The device is wearable and applies part of the force required for lifting objects at the wrist. Due to its high force output, we used the semicircular section actuator that was developed for this application. This is shown in Supplementary Video S3.

It was observed by Ang and Yeow³⁹ that the force transfer is affected by the fit. To maximize force transfer, we used a medical-grade wrist support for mounting the actuator. The wrist support is designed to be a compression fit on its own and can support part of the weight of the wrist. This minimizes transmission losses. In addition, an extreme strength fastener (Velcro Companies, United Kingdom) is used to attach the bending side of the actuator to the wrist support. This makes the design modular, and the actuator can be replaced easily in case of a need for higher force or a different size. These factors often come into play when the wearer changes, as individual variations in size and weight of limbs, are common.

The torsional actuator was shown to generate enough torque to screw on a bottle cap with a rudimentary gripper attachment (Fig. 8a). We also demonstrate an application where the actuator could turn a potentiometer knob that activates air flow to another actuator (Fig. 7b). Both these applications are shown in Supplementary Video S4 and demonstrate the practical utility of the torsional skin-based actuators.

FIG. 8.

Applications of fabric skins. (a, b) Torsional fabric skin-based actuator applications. (a) Tightening of bottle cap and (b) turning of a control knob were demonstrated. The actuator has enough torque to overcome the frictional resistance at approx 40 kPa owing to the larger diameter than shown in the previous figure. (c, d) Fabric skin-based bending actuator used as an assistive device. (c) Shows the provided range of motion at 40 kPa. (d) Shows the potential load-bearing capacity of the device when mounted on a mannequin hand. The device was shown to be able to lift 2.2 kg at 70 kPa.

Figure 8c and d also shows the use of the bending sleeve as an assistive wrist device. The device's range of motion is demonstrated in Figure 8c. Subsequently, we placed the device on a mannequin arm to gauge the output torque. At 70 kPa, the actuator could lift a 2.2 kg dumbbell bar, as shown in Figure 8d. Thus, based on the blocked force output as well as the lifting test, we believe that the fabric-based skins have utility in assistive applications. However, further tests studying the exoskeleton's effect on the wearer's EMG signals will be required to validate this claim and are beyond the scope of the current study.

Discussion

In this article, we demonstrated a design scheme that allows for the fabrication of reconfigurable multimaterial soft pneumatic actuators using silicone bladders and fabric skins. Specifically, we demonstrate an actuator design that achieves torsional deformation significantly greater than what was shown in the literature. We also demonstrated a bending design which can generate force comparable to what is shown in literature at moderate pressure (100 kPa). In addition, the fabrication method presented in this article provides a set of unique features and distinct advantages, modularity, reconfigurability, and the ability, to customize for a wide array of applications based on requirements for large deformations or high force or torque.

The isolation of the pneumatic channel from the external constraint (fabric) provides a multitude of advantages. First, it opens a wide array of materials that can be used as these materials do not need to be airtight as in the case of most conventional soft fabric actuators, that used TPU-coated fabrics. We demonstrate the utility using kinesiology tape, which is one of many materials or composites which can benefit from this design scheme.

The second advantage is that of reconfigurability. The ability to reuse the same designs in different combination to achieve composite motions is helpful in cases where modifications need to be made on the fly. On possible scenario is in a factory setting where the factory may have a set of fabric skins and bladders that in combination can grasp a wide array of objects or perform diverse tasks. This is especially relevant in the consumer goods industry where grippers must be changed based on the product on the supply line at a given time.

Compared to other soft actuators in the literature,^58–62 the fabric skin-based torsional actuator can achieve a significantly higher degree of twisting at moderate pressure of under 100 kPa. We can attribute this to the high extension ratio of the KTape as well as the low mechanical impedance of the composite actuator. Most existing torsional actuators must overcome significant stiffness of their own materials to twist, which reduces their capacity to twist. It must be noted that in our design as well, as the inner bladder increases in size, the mechanical impedance provides scales (proportional to the square of the radius), which means that the resulting actuator requires more pressure to generate the same degree of twist. We believe that the exceptional level of deformation that our design achieves would be ideal for locomotion applications.

The current bending skin-based actuator designed for the wrist demonstrates a force output comparable to existing actuators. We also showed the ability of the actuator to lift a mannequin arm with a dumbbell when mounted on a compression sleeve using Velcros.

The table below (Table 1) contains some representative soft actuators and their force/torque generation capabilities. It is difficult to directly compare all designs directly due to the difference in dimensions and measurement schemes. Hence, the application is mentioned to give a better idea. The review focuses mainly on actuators that could potentially be used for assistive applications.

Table 1.

Soft Robotic Actuators for Rehabilitation Using Positive Pneumatic Pressure and Their Performance Metrics

Material	Joint/Motion	Force (N)	Torque (Nm)	Pressure (kPa)	References
Silicone	Hand	1–3	—	160	⁴⁷
Silicone+braided sleeve	Hand	50	—	500	⁶⁸
Silicone with reinforcement	Hand	11.27	—	250	⁶⁹
Silicone with fabric reinforcement	Hand	2	—	120–160	⁴⁷
Fiber-reinforced silicone	Elbow	22		180	¹⁷
Fiber-reinforced silicone	Bending	—	0.05	200	⁷⁰
Fabric	Hand	9.5	—	120	⁷¹
Fabric	Hand	9.12	—	120	⁷²
Fabric	Wrist (Gripper)	60	—	200	⁵²
TPU-backed fabric	Hand	13	—	70	¹⁸
TPU-backed fabric	Elbow	—	27	300	⁷³
Silicone bladder with TPU-backed fabric	Elbow	—	3.611	110365	^74,75
TPU bladder with fabric	Shoulder	—	60	136	⁴⁵
TPU bladder with fabric	Shoulder	70	3–8	70	⁴⁴
TPU backed fabric	Knee	—	25	40	⁶⁷
TPU backed fabric	Hip	—	31	86	⁷⁶
TPU (3D-printed)	Hand	41	—	200	³⁶
TPU (3D-printed)	Hand	70	—	300	³⁵
TPU (3D-printed)	Hand	5	—	50	⁷⁷
TPU (3D-printed)	Elbow	75–98	—	200	³⁹
Fabric skin based	Wrist	190	—	100	This article

TPU, thermoplastic polyurethane.

It can be seen from this table that our skin-based design can provide comparable force output to similar sized actuators or actuators designed for the same task, but at lower operating pressures. The likely cause is the reduction in stored energy in the lower stiffness extensible fabric, allowing for more output force or work. It must also be noted that our blocked force measurement was performed against gravity, whereas most measurements in literature are performed with gravity assistance. This means that at least a few of the values mentioned in the table have some added effect on the weight of the actuator, whereas our measurement must overcome this effect. Recent studies using inextensible fabrics such as O'Neill et al.⁴⁵ have shown high peak torque output, and it is worth exploring in the future whether a combination of extensible and inextensible material schemes can further improve torque to pressure ratios.

One of the limitations of the current designs is the ability not to take pressures higher than about 110 kPa for single-layer designs and about 200 kPa for dual-layer designs for large durations. This is due to the natural adhesive of the KTape shearing. We currently only use the natural adhesive that comes with the Kinesiology Tape, which cannot sustain the shear forces it is subject to a higher pressures. We believe that applying another layer of adhesive can further strengthen the actuator. Additional layers of KTape on top of the layers presented here can also be used to strengthen the design. We found that using two layers also added to the durability of the skins when subject to a cyclic test as well as the force output.

The torsional skins demonstrated herein are especially difficult to fabricate consistently and require practice. Using a metal rod as a template helps reduce variation in radius, and markings help with orientation, but the authors believe that an automated method is necessary to make these designs viable.

Sewing³⁰ and machine knitting⁶³ are alternative options that have been explored. However, depending on the pattern of the sew, it can become very complex to execute, and the sew adds to the stiffness of the actuator, thus modifying its properties. All three of these solutions come with the trade-off that it adds stiffness to the actuator, meaning that the deformation is likely to be lower at the same pressures. Moreover, the stiffness added tends to be nonsymmetric, and a significant change in the behavior of the actuator from the intended design can often be seen. Thus, there is a trade-off between the deformation and the required pressure for the same, as well as the durability.

It is also possible to use different materials or material combinations for bladders along with the skins apart from the two shown here, which may lead to a wider set of applications. Potential combinations with multichamber designs⁶⁴ and alternative manufacturing methods³² could also be explored for extensible fabric-based designs. A recent study on machine-based manufacturing⁶⁵ shows great promise in creating repeatable fabric-based actuators especially with integrating sensing. Finally, modeling of extensible fabric-based designs, possibly using Finite Element Analysis or SpringForm,⁶² can be explored as current modeling methods assume inextensibility for modeling soft actuators.^45,66,67

Conclusion and Future Work

In this article, we presented a universal design scheme for fabric-based actuators capable of multiple motions, bending and torsion and helical motions. The designs allow for reconfigurability due to a modular bladder and skin design. The properties of the skin can be changed based on the wrapping pattern of an anisotropic fabric (KTape) and the number of layers. This method of mechanical programming relies on utilizing the materials' mechanical properties. We showed that the designs could be tuned to generate exceptional high deformation or large force based on requirements. Multilayer skins trade-off deformation for force and higher durability presented by single-layer designs, enabling optimization for various applications such as assistive exoskeletons or locomotion.

We demonstrated the utility of the high-force multilayer bending skins as a wearable assistive actuator, whereas the torsional skins can be shown to have practical utility in applications that require pure torsion, such as turning knobs or screwing bottle caps.

In the current stage, we use a prototyping method that involves significant human skill, but we believe that it is possible to design skins using automated fabrication, such as fabrication of the skins using a fabric machine. We expect this to reduce the variance in the manufactured design as well as improve the durability even further. In addition, there is potential for fabrics to be designed specifically for a kind of actuator that is being fabricated, allowing for more versatility.

In future work, we will explore 3D printed actuators that also use the composite scheme. Owing to their higher material stiffness and strain withstanding capacities, we expect these designs to outperform the fabric-based skins in terms of force output. Fabric-based skins still present a viable case where comfort and compliance are required, and reconfigurability is desired. In the future, it is possible to study the variations in fabric skin designs using multichamber approaches, such as those shown in previous studies.

Footnotes

Acknowledgments

The authors thank the Integrated Science and Engineering Program (ISEP) at NUS for providing scholarship support to the first author of this article.

Authors' Contributions

A.B.: Conceptualization, Methodology, Formal analysis, Software, Visualization, Article Drafting, Data Curation, and Writing—Review and Editing Final Article. S.S.J.: Investigation, Formal analysis, Software, Writing—Review and Editing, Data Curation, and Visualization. L.L.T.: Investigation, Formal analysis, and Writing—Review and Editing. R.C.-H.Y.: Conceptualization, Supervision, Project administration, Resources, Funding Acquisition, and Writing—Review and Editing.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by the Soft and Hybrid Robotics Phase 2a program (W2025d0243) under the National Robotics Programme—Robotics Enabling Capabilities and Technologies and the A*STAR Industry Alignment Fund—Pre-Positioning (A20H8A0241).

Supplementary Material

References

Choi

, Kang

, Jung

, et al. Exo-Wrist: A soft tendon-driven wrist-wearable robot with active anchor for dart-throwing motion in hemiplegic patients. IEEE Robot Autom Lett, 2019; 4(4):4499–4506; doi: 10.1109/LRA.2019.2931607

Xiloyannis

, Galli

, Chiaradia

, et al. A soft tendon-driven robotic glove: Preliminary evaluation. Biosyst Biorobot, 2019; 21(August):329–333; doi: 10.1007/978-3-030-01845-0_66

Ismail

, Ariyanto

, Perkasa

, et al. Soft elbow exoskeleton for upper limb assistance incorporating dual motor-tendon actuator. Electron, 2019; 8(10):1184; doi: 10.3390/electronics8101184

Panizzolo

, Galiana

, Asbeck

, et al. A biologically-inspired multi-joint soft exosuit that can reduce the energy cost of loaded walking. J Neuroeng Rehabil, 2016; 13(1):43; doi: 10.1186/s12984-016-0150-9

Meller

, Bryant

, Garcia

. Reconsidering the McKibben muscle: Energetics, operating fluid, and bladder material. J Intell Mater Syst Struct, 2014; 25(18):2276–2293; doi: 10.1177/1045389X14549872

Takashima

, Sugitani

, Morimoto

, et al. Pneumatic artificial rubber muscle using shape-memory polymer sheet with embedded electrical heating wire. Smart Mater Struct, 2014; 23(12):125005; doi: 10.1088/0964-1726/23/12/125005

, Alici

. Bioinspired three-dimensional-printed helical soft pneumatic actuators and their characterization. Soft Robot, 2020; 7(3):267–282; doi: 10.1089/soro.2019.0015

. Design and simulation of a soft robotic device for wrist rehabilitation. In: Proceedings—2020 3rd International Conference on Advanced Electron Materials, Computers, Software Engineering (AEMCSE) 2020. 2020; pp. 697–701; doi: 10.1109/AEMCSE50948.2020.00153

Sun

, Yap

, Liang

, et al. Stiffness customization and patterning for property modulation of silicone-based soft pneumatic actuators. Soft Robot, 2017; 4(3):251–260; doi: 10.1089/soro.2016.0047

10.

Rosalia

, Ang

BWK

, Yeow

RCH

. Geometry-Based customization of bending modalities for 3D-Printed soft pneumatic actuators. IEEE Robot Autom Lett, 2018; 3(4):3489–3496; doi: 10.1109/LRA.2018.2853640

11.

Meller

, Chipka

, Volkov

, et al. Improving actuation efficiency through variable recruitment hydraulic McKibben muscles: Modeling, orderly recruitment control, and experiments. Bioinspir Biomim, 2016; 11(6):065004; doi: 10.1088/1748-3190/11/6/065004

12.

Chu

, Patterson

. Soft robotic devices for hand rehabilitation and assistance: A narrative review. J Neuroeng Rehabil, 2018; 15(1):1–14; doi: 10.1186/s12984-018-0350-6

13.

Agarwal

, Besuchet

, Audergon

, et al. Stretchable materials for robust soft actuators towards assistive wearable devices. Sci Rep, 2016; 6(October); doi: 10.1038/srep34224

14.

Xiloyannis

, Chiaradia

, Frisoli

, et al. Physiological and kinematic effects of a soft exosuit on arm movements. J Neuroeng Rehabil, 2018;(February); doi: 10.1186/s12984-019-0495-y

15.

Tsuneyasu

, Ohno

, Fukuda

, et al. A soft exoskeleton suit to reduce muscle fatigue with pneumatic artificial muscles. In Proceedings of the 9th Augmented Human International Conference 2018 Feb 6. Association for Computing Machinery: Seoul, Republic of Korea; 2018; pp. 1–4.

16.

Polygerinos

, Wang

, Galloway

, et al. Towards a soft pneumatic glove for hand rehabilitation. In: Robotics and Autonomous Systems, 2015; pp. 135–143; doi: 10.1016/j.robot.2014.08.014

17.

, Chen

, Liu

, et al. A soft wearable exoskeleton with pneumatic actuator for assisting upper limb. In: 2020 IEEE International Conference on Real-Time Computing and Robotics (RCAR) 2020. 2020;(January 2021); pp. 99–104; doi: 10.1109/RCAR49640.2020.9303269

18.

Yap

, Khin

, Koh

, et al. A fully fabric-based bidirectional soft robotic glove for assistance and rehabilitation of hand impaired patients. IEEE Robot Autom Lett, 2017; 2(3):1383–1390; doi: 10.1109/LRA.2017.2669366

19.

Hughes

, Culha

, Giardina

, et al. Soft manipulators and grippers: A review. Front Robot AI, 2016; 3(November):1–12; doi: 10.3389/frobt.2016.00069

20.

Truby

, Katzschmann

, Lewis

, et al. Soft robotic fingers with embedded ionogel sensors and discrete actuation modes for somatosensitive manipulation. Soft Robot, 2019;(May):322–329.

21.

Odhner

, Dollar

. Stable, open-loop precision manipulation with underactuated hands. Int J Rob Res, 2015; 34(11):1347–1360; doi: 10.1177/0278364914558494

22.

Branyan

, Fleming

, Remaley

, et al. Soft snake robots: Mechanical design and geometric gait implementation. In: 2017 IEEE International Conference on Robotics and Biomimetics (ROBIO) 2017. 2018; 2018(December):282–289; doi: 10.1109/ROBIO.2017.8324431

23.

Shepherd

, Ilievski

, Choi

, et al. Multigait soft robot. Proc Natl Acad Sci U S A, 2011; 108(51):20400–20403; doi: 10.1073/pnas.1116564108

24.

Rafsanjani

, Zhang

, Liu

, et al. Kirigami skins make a simple soft actuator crawl. Sci Robot, 2018; 3(15):eaar7555; doi: 10.1126/scirobotics.aar7555

25.

Cao

, Liu

, Chen

, et al. A novel slithering locomotion mechanism for a snake-like soft robot. J Mech Phys Solids, 2017; 99(July 2016):304–320; doi: 10.1016/j.jmps.2016.11.019

26.

Tawk

, In Het Panhuis

, Spinks

, et al. Bioinspired 3d printable soft vacuum actuators for locomotion robots, grippers and artificial muscles. Soft Robot, 2018; 5(6):685–694; doi: 10.1089/soro.2018.0021

27.

Sun

, Guo

, Miller-Jackson

, et al. Design and fabrication of a shape-morphing soft pneumatic actuator: Soft robotic pad. In: IEEE International Conference on Intelligent Robots and Systems, vol. 2017. 2017; pp. 6214–6220; doi: 10.1109/IROS.2017.8206524

28.

Homberg

, Katzschmann

, Dogar

, et al. Robust proprioceptive grasping with a soft robot hand. Auton Robots, 2019; 43(3):681–696; doi: 10.1007/s10514-018-9754-1

29.

Marchese

, Katzschmann

, Rus

. A recipe for soft fluidic elastomer robots. Soft Robot, 2015; 2(1):7–25; doi: 10.1089/soro.2014.0022

30.

Zhu

, Do

, Hawkes

, et al. Fluidic fabric muscle sheets for wearable and soft robotics. Soft Robot, 2020; 7(2):179–197; doi: 10.1089/soro.2019.0033

31.

Khin

, Yap

, Ang

, et al. Fabric-based actuator modules for building soft pneumatic structures with high payload-to-weight ratio. In: IEEE International Conference on Intelligent Robots and Systems. IEEE; 2017; pp. 2744–2750; doi: 10.1109/IROS.2017.8206102

32.

Connolly

, Wagner

, Walsh

, et al. Sew-free anisotropic textile composites for rapid design and manufacturing of soft wearable robots. Extrem Mech Lett, 2019; doi: 10.1016/J.EML.2019.01.007

33.

Cappello

, Galloway

, Sanan

, et al. Exploiting textile mechanical anisotropy for fabric-based pneumatic actuators. Soft Robot, 2018; 00(00); doi: 10.1089/soro.2017.0076

34.

Liang

, Yap

, Guo

, et al. Design and characterization of a novel fabric-based robotic arm for future wearable robot application. In: 2017 IEEE International Conference on Robotics and Biomimetics (ROBIO) 2017, vol. 2018. 2018; pp. 367–372; doi: 10.1109/ROBIO.2017.8324445

35.

Yap

, Ng

, Yeow

. High-force soft printable pneumatics for soft robotic applications. Soft Robot, 2016; 3(3):144–158; doi: 10.1089/soro.2016.0030

36.

Ang

BWK

, Yeow

. Print-it-Yourself (PIY) glove: A fully 3D printed soft robotic hand rehabilitative and assistive exoskeleton for stroke patients. In: IEEE International Conference on Intelligent Robots and Systems, vol. 2017. 2017; pp. 1219–1223; doi: 10.1109/IROS.2017.8202295

37.

Tawk

, Spinks

, In Het Panhuis

, et al. 3D printable linear soft vacuum actuators: Their modeling, performance quantification and application in soft robotic systems. IEEE/ASME Trans Mech, 2019; 24(5):2118–2129; doi: 10.1109/TMECH.2019.2933027

38.

Connolly

, Polygerinos

, Walsh

, et al. Mechanical programming of soft actuators by varying fiber angle. Soft Robot, 2015; 2(1):26–32; doi: 10.1089/soro.2015.0001

39.

Ang

BWK

, Yeow

. Design and modeling of a high force soft actuator for assisted elbow flexion. IEEE Robot Autom Lett, 2020; 5(2):3731–3736; doi: 10.1109/LRA.2020.2980990

40.

Natividad

, Del Rosario

, Chen

, et al. A hybrid plastic-fabric soft bending actuator with reconfigurable bending profiles. In 2017 IEEE International Conference on Robotics and Automation (ICRA) 2017 May 29. IEEE: Singapore, Singapore; 2017; pp. 6700–6705.

41.

Natividad

, Del Rosario

, Chen

PCY

, et al. A reconfigurable pneumatic bending actuator with replaceable inflation modules. Soft Robot, 2018; 5(3):304–317; doi: 10.1089/soro.2017.0064

42.

Booth

, Shah

, Case

, et al. OmniSkins: Robotic skins that turn inanimate objects into multifunctional robots. Sci Robot, 2018; 3(22):1853; doi: 10.1126/scirobotics.aat1853

43.

Calais

, Sanandiya

, Jain

, et al. Freeform liquid 3D printing of soft functional components for soft robotics. ACS Appl Mater Interfaces, 2022; 14(1):2301–2315; doi: 10.1021/acsami.1c20209

44.

O'Neill

, Phipps

, Cappello

, et al. A soft wearable robot for the shoulder: Design, characterization, and preliminary testing. In: IEEE International Conference on Rehabilitation Robotics, 2017; pp. 1672–1678; doi: 10.1109/ICORR.2017.8009488

45.

O'Neill

, McCann

, Hohimer

, et al. Unfolding textile-based pneumatic actuators for wearable applications. Soft Robot, 2022; 9(1):163–172; doi: 10.1089/soro.2020.0064

46.

Belforte

, Eula

, Ivanov

, et al. Soft pneumatic actuators for rehabilitation. Actuators, 2014; 3(2):84–106; doi: 10.3390/act3020084

47.

Yap

, Lim

, Nasrallah

, et al. A soft exoskeleton for hand assistive and rehabilitation application using pneumatic actuators with variable stiffness. In: Proceedings—IEEE International Conference on Robotics and Automation, 2015; pp. 4967–4972; doi: 10.1109/ICRA.2015.7139889

48.

Polygerinos

, Wang

, Galloway

, et al. Soft robotic glove for combined assistance and at-home rehabilitation. Rob Auton Syst, 2015; 73:135–143; doi: 10.1016/j.robot.2014.08.014

49.

Yang

. Modeling and Analysis of a Novel Pneumatic Artificial Muscle and Pneumatic Arm Exoskeleton. Virginia Tech: Virginia, USA;. 2017.

50.

Velliste

, Perel

, Spalding

, et al. Cortical control of a prosthetic arm for self-feeding. Nature, 2008; 453(7198):1098–1101; doi: 10.1038/nature06996

51.

Miron

, Bédard

, Plante

. Sleeved bending actuators for soft grippers: A durable solution for high force-to-weight applications. Acuators, 2018; 7(3):40; doi: 10.3390/act7030040

52.

Fei

, Wang

, Pang

. A novel fabric-based versatile and stiffness-tunable soft gripper integrating soft pneumatic fingers and wrist. Soft Robot, 2019; 6(1):1–20; doi: 10.1089/soro.2018.0015

53.

Tawk

, Gillett

, In Het Panhuis

, et al. A 3D-printed omni-purpose soft gripper. IEEE Trans Robot, 2019; 35(5):1268–1275; doi: 10.1109/TRO.2019.2924386

54.

Shintake

, Cacucciolo

, Floreano

, et al. Soft robotic grippers. Adv Mater, 2018; 30(29); doi: 10.1002/adma.201707035

55.

Ambrose

, Chiang

NZR

, Cheah

DSY

, et al. Compact multilayer extension actuators for reconfigurable soft robots. Soft Robot, 2022; 00(00):31–33; doi: 10.1089/soro.2022.0042

56.

, Chen

, Wang

, et al. Design, modeling, and evaluation of fabric-based pneumatic actuators for soft wearable assistive gloves. Soft Robot, 2020; doi: 10.1089/soro.2019.0105

57.

Gates

, Walters

, Cowley

, et al. Range of motion requirements for upper-limb activities of daily living. Am J Occup Ther, 2016; 70(1); doi: 10.5014/ajot.2016.015487

58.

Sanan

, Lynn

, Griffith

. Pneumatic torsional actuators for inflatable robots. J Mech Robot, 2014; 6(3):031003; doi: 10.1115/1.4026629

59.

Yan

, Zhang

, Xu

, et al. A new spiral-type inflatable pure torsional soft actuator. Soft Robot, 2018; 5(5):527–540; doi: 10.1089/soro.2017.0040

60.

Olson

, Menguc

. Helically wound soft actuators for torsion control. In: 2018 IEEE International Conference on Soft Robotics, RoboSoft 2018 IEEE: Livorno, Italy;, 2018; pp. 214–221; doi: 10.1109/ROBOSOFT.2018.8404922

61.

Hagihara

, Wakimoto

, Kanda

, et al. Operation of a pneumatic soft manipulator using a wearable interface with flexible strain sensors. In 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2019 Nov 3. IEEE: Macau; 2019; pp. 4949–4954.

62.

Ahlquist

, McGee

, Sharmin

. PneumaKnit: Actuated architectures through wale- and course-wise tubular knit-constrained pneumatic systems. In: Proceedings of 37th Annual Conference of the Association for Computer Aided Design in Architecture, 2022; pp. 38–51; doi: 10.52842/conf.acadia.2017.038

63.

Elmoughni

, Yilmaz

, Ozlem

, et al. Machine-knitted seamless pneumatic actuators for soft robotics: Design, fabrication, and characterization. Actuators, 2021; 10(5); doi: 10.3390/act10050094

64.

Cappello

, Meyer

, Galloway

, et al. Assisting hand function after spinal cord injury with a fabric-based soft robotic glove. J NeuroEng Rehabil, 2018; 15(1):1–10.

65.

Luo

, Wu

, Spielberg

, et al. Digital fabrication of pneumatic actuators with integrated sensing by machine knitting. In: Conference on Human Factors in Computing Systems—Proceedings, 2022; doi: 10.1145/3491102.3517577

66.

Nesler

, Swift

, Rouse

. Initial design and experimental evaluation of a pneumatic interference actuator. Soft Robot, 2018; 5(2):138–148; doi: 10.1089/soro.2017.0004

67.

Fang

, Yuan

, Wang

, et al. Novel accordion-inspired foldable pneumatic actuators for knee assistive devices. Soft Robot, 2020; 7(1):95–108; doi: 10.1089/soro.2018.0155

68.

Al-Fahaam

, Davis

, Nefti-Meziani

. The design and mathematical modelling of novel extensor bending pneumatic artificial muscles (EBPAMs) for soft exoskeletons. Rob Auton Syst, 2018; 99:63–74; doi: 10.1016/j.robot.2017.10.010

69.

Jiang

, Chen

, Liu

, et al. Fishbone-inspired soft robotic glove for hand rehabilitation with multi-degrees-of-freedom. In: 2018 IEEE International Conference on Soft Robotics, RoboSoft 2018 Livorno, Italy. 2018; pp. 394–399; doi: 10.1109/ROBOSOFT.2018.8404951

70.

Tarvainen

, Yu

. Pneumatic multi-pocket elastomer actuators for metacarpophalangeal joint flexion and abduction-adduction. Actuators, 2017; 6(3):27; doi: 10.3390/act6030027

71.

Yap

HK.

Design, Fabrication, And Characterization of Soft Pneumatic Actuators Towards Soft Wearable Robotic Applications. National University of Singapore: Singapore; 2017.

72.

Yap

, Lim

, Nasrallah

, et al. Design and preliminary feasibility study of a soft robotic glove for hand function assistance in stroke survivors. Front Neurosci, 2017; 11(OCT):1–14; doi: 10.3389/fnins.2017.00547

73.

Thalman

, Lam

, Nguyen

, et al. A novel soft elbow exosuit to supplement bicep lifting capacity. In: 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE: Madrid, Spain; 2018; pp. 6965–6971.

74.

Nassour

, Hamker

. Enfolded textile actuator for soft wearable robots. In: 2019 IEEE International Conference on Cyborg and Bionic Systems. IEEE: Munich, Germany; 2019.

75.

Nassour

, Zhao

, Grimmer

. Soft pneumatic elbow exoskeleton reduces the muscle activity, metabolic cost and fatigue during holding and carrying of loads. Sci Rep, 2021; 11(1):1–14; doi: 10.1038/s41598-021-91702-5

76.

Miller-Jackson

, Natividad

, Lim

DYL

, et al. A wearable soft robotic exoskeleton for hip flexion rehabilitation. Front Robot AI, 2022; 9(April):1–17; doi: 10.3389/frobt.2022.835237

77.

. Flexible Fluidic Actuators for Soft Robotic Applications. University of. Wollongong: Wollongong, Australia;. 2019.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

9.77 MB

24.30 MB

25.36 MB

27.74 MB

0.00 MB