A Sensorized Soft Robotic Hand with Adhesive Fingertips for Multimode Grasping and Manipulation

Abstract

Soft robotic grippers excel at achieving conformal and reliable contact with objects without the need for complex control algorithms. However, they still lack in grasp and manipulation abilities compared with human hands. In this study, we present a sensorized multi-fingered soft gripper with bioinspired adhesive fingertips that can provide both fingertip-based adhesion grasping and finger-based form closure grasping modes. The gripper incorporates mushroom-like microstructures on its adhesive fingertips, enabling robust adhesion through uniform load shearing. A single fingertip exhibits a maximum load capacity of 4.18 N against a flat substrate. The soft fingers have multiple joints, and each joint can be independently actuated through pneumatic control. This enables diverse bending motions and stable grasping of various objects, with a maximum load capacity of 28.29 N for three fingers. In addition, the soft gripper is equipped with a kirigami-patterned stretchable sensor for motion monitoring and control. We demonstrate the effectiveness of our design by successfully grasping and manipulating a diverse range of objects with varying shapes, sizes, and curvatures. Moreover, we present the practical application of our sensorized soft gripper for remotely controlled cooking.

Introduction

Inspired by the dexterity of human hands, soft robotic grippers, which can manipulate a wider variety of objects, have been actively explored in recent years toward the development of more versatile and universal grippers.^1–8 Several types of soft grippers have been developed so far, which include jamming grippers, multi-fingered grippers, and adhesion-based grippers.^9–11 The jamming grippers grip objects by forming mechanical interlocking with the target objects, which makes them suitable for holding objects with complex or arbitrary shapes.^9,12,13

However, they are not well-suited for grasping flat or large objects. The multi-fingered grippers, one of the most widely explored grippers, have the capability to grasp objects of various shapes through either form-closure grasping.^14–17 However, the size of the object that can be grasped is inevitably predetermined by the working range of the fingers, and grasping flat objects is challenging. Adhesion-based grippers that utilize van der Waals or suction forces are useful for picking up flat or soft deformable object.^18–21 However, the adhesion performance is highly influenced by surface conditions such as curvature, roughness, and contaminants.

Hybrid strategies that integrate different grasping mechanisms into a single soft gripper system can address the limitations of each gripping mechanism.^22–24 In particular, a multi-fingered gripper integrated with an adhesive grasping mechanism has a great potential for more versatile and universal grasping and manipulation.^25–27 To this end, Yang et al. proposed a jamming layer integrated fingered gripper in which the jamming layer acts as interphalangeal joints.²³ This hybrid design enables control over the bending shape and stiffness of the gripper, enhancing gripping performance.

However, this gripper still has the limitations of the typical finger-type grippers described earlier. Ruotolo et al. reported a multi-fingered gripper that integrates a gecko-inspired adhesive suspension layer.²⁶ The adhesive layer with buckling rib structures provides shear load sharing and normal compliance, so the gripper could stably grasp a wide range of objects and also manipulate them in arbitrary directions.

However, the buckling-rib adhesive provides only shear forces without exhibiting normal adhesion, which limits the grasping and manipulation of flat objects as with the previous jamming or finger-type grippers. Despite the remarkable potential, studies on soft grippers toward more universal and high-performance manipulation have not yet been sufficiently explored, and require further development.^28–31 In addition, many previous studies on grippers have mainly focused on gripping mechanisms at the component level without demonstrating system-level manipulation.^32–34

In this study, we propose a hybrid design that incorporates sensorized multi-fingered soft grippers and bioinspired adhesive fingertips for grasping and manipulating various objects as an integrated system for practical scenarios. Our sensorized soft gripper offers multiple grasping modes, depending on the types of target objects: fingertip-based adhesion grasping and finger-based form closure grasping. The adhesive fingertip has bioinspired microstructures over the surface and provides strong normal adhesion through uniform load shearing on the target objects.

The finger comprises two joints, allowing for diverse bending motions and ensuring stable grasping of various objects. The versatile bending features also allow the finger to adjust the angle between the adhesive tip and the object surface, maximizing normal load sharing and adhesion grasping performance. The soft gripper is further equipped with a kirigami-patterned stretchable soft sensor to monitor and control the actuating motion of the fingers. Stretchable soft sensors have been extensively employed to measure the movement of grippers and human hands in recent studies.^35–37

We demonstrate that the sensorized soft gripper can grasp and manipulate a wide range of objects with different shapes and sizes through a combination of adhesion grasping and form closure grasping modes. In addition, we show that the proposed soft gripper can be used for remotely controlled cooking.

Materials and Methods

Fabrication of the bioinspired adhesive fingertip

The bioinspired adhesive fingertip comprises a convex chamber, rigid bones, and an adhesive membrane (Fig. 1a). The convex chamber was fabricated using a silicon elastomer through replica molding. Twenty rigid frames, referred to as “rigid bones,” were then embedded along a side wall of the convex chamber. Finally, the adhesive membrane, made of polydimethylsiloxane (PDMS), was integrated with the convex chamber. This resulting fingertip chamber features mushroom-like adhesive microstructures on its surface, with a tip diameter of 29.3 μm, neck diameter of 21.1 μm, height of 24.8 μm, and pitch of 40 μm (see Supplementary Description S1 and Fig. S1 in Supplementary Data for more details).³⁸

FIG. 1.

Fabrication methods for the soft robotic hand with bioinspired adhesive fingertips. (a) Fabrication method for kirigami-inspired soft sensor through DIW of the liquid metal: (i) detailed section and (ii) configuration of the kirigami-inspired soft sensor. (b) Fabrication and configuration of the bioinspired adhesive fingertip. (c) Fabrication for the multi-joint soft finger with the rigid structures through (i) the 3D printed molds, (ii) assembling the rigid structures, and (iii) integrating with inextensible layer and bottom layer. DIW, direct ink writing.

Fabrication of the kirigami-patterned soft sensor with liquid metal

The soft sensor was fabricated by the direct ink writing (DIW) method using eutectic gallium indium (eGaIn) (Fig. 1b). (see Supplementary Description S2, Supplementary Figs. S2 and S3, and Supplementary Table S1 for more details).^39,40

Fabrication of the multi-joint gripper

The soft gripper, featuring multi-joint fingers, was created using a combination of soft and rigid materials.⁴¹ The soft parts and rigid parts were fabricated using Smooth-sil 940 and Acrylonitrile Butadiene Styrene (ABS) 3D printed material, respectively. The soft finger body, including concave chambers, was fabricated by the 3D printed mold (Fig. 1c-i). The rigid parts were fixed to the soft finger body by bolts (Fig. 1c-ii). The air hoses made of rubber were inserted into the soft finger body (Fig. 1c-iii). At the bottom part of the soft finger body, an inextensible layer was attached (Fig. 1c-iii).

Results and Discussion

Design of the robotic gripper

The proposed robotic hand comprises three fingers, fingertips, and stretchable sensors (Fig. 2a). The three-finger gripper provides improved accuracy, stability, and dexterity when manipulating various 3D objects, compared with the two-finger gripper.^42,43 The fingers are constructed using soft inflatable chambers and rigid structures, as illustrated in Fig. 2b-i. By supplying compressed air into the soft chambers, the fingers can have diverse bending motions.

FIG. 2.

Schematic illustrations of a sensorized soft gripper with 2 DoF fingers and adhesive fingertips. (a) Overall configuration of the sensorized soft gripper with the adhesive fingertip integrated with the robotic manipulator. (b) (i) Detailed structure of the 2 DoF finger, consisting of soft chambers and rigid bones, and integration of (ii) a kirigami-patterned soft sensor and (iii) bioinspired adhesive fingertips with the finger. (c) Detailed schematics showing the versatile grasping capabilities of the system through (i) the finger-grasping mode, where the fingers conform to the object's shape, and (ii) the fingertip-grasping mode, where the bioinspired adhesive tips adhere to the object's surface. DoF, degrees of freedom.

The round-edge-shaped rigid structures (the “Rigid bone” in Fig. 2b-i) are inserted between the soft chambers to provide greater force to the finger with faster bending speeds.⁴¹ In addition, stretchable soft sensors were attached to the fingers to measure and provide feedback control for the bending motions of the fingers.^44,45 The sensors have a kirigami pattern and can be stretched without interfering with the actuating movements of the fingers (Fig. 2b-ii).⁴⁶

The fingertips have an adhesive layer consisting of bioinspired mushroom-shaped micropillars on the surface (Fig. 2a, 2b-iii, and Supplementary Fig. S1).⁴⁷ The mushroom microstructures exhibit strong normal adhesion owing to the uniform stress distribution and load sharing at the contact interface. In addition, these microstructures enable the suppression of continuous crack propagation and increase tolerance to surface roughness, contributing significantly to the enhancement of adhesion.^48–50

However, the mushroom micropillars are limited in their ability to conform to the macroscopic curvature of nonplanar objects. To address this limitation, the adhesive layer was formed on an inflatable chamber whose curvature can be adjusted through pneumatic control. This feature enables the fingertips to make conformal contact even with nonplanar objects, providing an adhesive fingertip-grasping mode (Fig. 2c-i). Three soft robotic fingers, each with independently actuated two joints, can provide a form closure-grasping mode through pneumatic control (Fig. 2c-ii).

Adhesion properties of robotic fingertips

Fig. 3a illustrates the pick-and-place grasping motion of the fingertip. The fingertip comprises a flexible convex chamber connected to a tube and valve for internal pressure control. As the fingertip approaches the target object with an open valve (Step I), the object applies a preload to the chamber, passively deforming the chamber shape to conform to the object's macroscopic shape (Step II). Simultaneously, the adhesive microstructures on the fingertip's surface form conformal contacts with the object at the microscopic level.⁵¹ The rigid frames embedded on the side wall of the chamber play a crucial role in facilitating the conformal contacts of the microstructures with the object by ensuring even distribution of the preload at the contact interface (Step II).

FIG. 3.

Adhesion properties of the bioinspired fingertips. (a) Schematic illustration of the grasping and release process of the fingertips. (b) Force–distance curves of convex fingertips with and without mushroom microstructures against a flat acrylic plate. Both fingertips have embedded side frames (frame number: 20). (c) Force–distance curves of convex fingertips with different numbers of side frames (0, 5, 10, and 20). (d) Schematic illustrations showing the contact behavior during the loading process of (i) convex chamber with 20 side frames, (ii) convex chamber without side frames, and (iii) flat chamber with 20 side frames. (e) Frustrated total internal reflection images showing the contact interface during the approach and retraction of (i) convex chamber with 20 side frames, (ii) convex chamber without side frames, and (iii) flat chamber with 20 side frames. (f) Contact area of the three different chambers evaluated by frustrated total internal reflection. (g) Normal adhesion of convex and flat chambers against planar and cylindrical substrates. Data are presented as mean ± s.d. (n = 5 measurements from the same sample measured repeatedly). (h) Photographs showing the contact between the (i) convex chamber and cylindrical substrate and (ii) flat chamber and cylindrical substrate.

Once the fingertip establishes firm contact with the object at both the macroscopic and microscopic levels, the valve is closed to maintain the deformed intimate contact state of the fingertip (Step III). This enables the fingertip to stably pick up the object through the adhesion grasping mode. Unlike many previous pneumatic grippers that require an application of negative pressure using a pump for grasping, our fingertip does not require negative pressure for operation.

However, when negative pressure was applied to the fingertips, it facilitated quicker grasping of the target object and demonstrated slightly higher maximum adhesion strengths (5.25 N) compared with instances without negative pressure (Supplementary Fig. S5). To release the grasped object, one can simply inflate the chamber, causing the chamber to expand, and peeling begins at the edge of the contact area (Step IV) when positive air pressure is applied to the chamber.

To quantitatively assess the adhesion performance of the bioinspired fingertip with mushroom microstructures, we conducted adhesion force measurements using a custom-built adhesion measurement setup (Supplementary Fig. S4). The fingertip sample had 20 side frames. We performed the measurements by approaching and retracting the fingertip against a flat acrylic plate with a preload of 4 N and a retraction speed of 0.5 mm s⁻¹. We also used a fingertip sample without mushroom microstructure as a control for comparison.

The force–distance curves shown in Fig. 3b clearly indicate that the mushroom fingertip exhibits significantly higher adhesion (4.18 N) compared with the bare fingertip, which showed negligible adhesion (0.05 N). We also conducted adhesion measurements for mushroom fingertips with and without side frames. As illustrated in the “approach” steps of the force–distance curves in Fig. 3c, the mushroom fingertips with embedded frames (Frame 5, Frame 10, and Frame 20) were able to apply appropriate preloads to the substrate, whereas the fingertip without frames (Frame 0) failed to apply sufficient preloads.

Moreover, the frames prevent deformation of the chamber caused by the weight of the grasped object, thus avoiding interfacial peeling during the retraction step (Supplementary Fig. S6 and Supplementary Movie S1). Consequently, mushroom fingertips with side frames exhibit higher adhesion forces (4.14–4.21 N) and adhesion energy (12.74–14.18 mJ) compared with the fingertip without frames (adhesion force: 2.08–2.32 N, adhesion energy: 4.23–5.57 mJ). In particular, the fingertips with a higher number of frames showed better adhesion performance (Fig. 3c).

To investigate the mechanisms underlying the enhanced adhesion of framed convex chambers, we employed frustrated total internal reflection to measure the contact area of convex chambers with and without frames during the contact process with a flat substrate (experimental setup is shown in Supplementary Description S4 and Supplementary Fig. S7). As depicted in Fig. 3d-i and e-i, the frames-embedded convex chamber fingertip achieved conformal contact across the contact interface, resulting in a significant contact area (up to ∼1141.5 mm²) during the approach process (see Supplementary Figs. S8 and S9).

This highlights that the frames embedded in the side wall of the chamber facilitate the uniform distribution of the preload for conformal contact formation (Fig. 3d-i). Moreover, the high contact was sustained during the retraction step (Fig. 3e-i, f). In contrast, the convex chamber fingertip without frames was not able to distribute the preload uniformly across the contact interface and could only apply a concentrated preload through the neck region of the fingertip (Fig. 3d-ii). As a result, the frameless convex chamber was unable to establish uniform contact with the substrate, leading to a smaller contact area (589.3 mm²) during the approach process (Fig. 3e-ii, f and Supplementary Fig. S8d).

Furthermore, during the retraction step, the contact area decreased rapidly. This can be attributed to the chamber without frames being unable to sustain the deformation caused by the weight of the attached substrate (Fig. 3e-ii, f).

Next, we compared the adhesive performance of a convex chamber design and a flat chamber design, both of which featured mushroom microstructures and 20 frames (Fig. 3g). The convex chamber exhibited a maximum normal adhesion of 4.18 N against a flat acrylic plate, whereas the flat chamber had an adhesion of 2.76 N. This was attributed to the convex chamber's ability to make conformal contacts with the flat substrate, starting from the central region of the contact interface during the approach stage, resulting in a larger interfacial contact area (Supplementary Fig. S9c).

In contrast, the flat chamber only formed partial contact with the flat plate. The contact interface between the flat chamber and the substrate primarily occurred along the side rim region (peripheral region) of the chamber fingertip, with the central region unable to establish intimate contacts (Fig. 3d-iii, e-iii), resulting in a limited contact area (Fig. 3f and Supplementary Fig. S9d, e).

In addition to flat objects, the convex chamber fingertip demonstrated significantly higher adhesion performance for nonplanar objects, such as cylinders, compared with the flat chamber fingertip. The convex chamber fingertip exhibited strong normal adhesion ranging from 4.36 to 4.85 N against acrylic cylinders with varied diameters (diameter: 20, 30, and 50 mm) (Fig. 3g). Conversely, the normal adhesion of the flat chamber fingertip was only in the range of 0.62–1.51 N for the same cylinders.

This is due to the high deformability of the convex chamber, which allows it to adapt to the macroscopic shape of the curved objects and establish seamless, full contact with them (Fig. 3h-i). In contrast, the limited deformability of the flat chamber prevents it from establishing seamless, full contact with the curved objects, resulting in lower adhesion performance (Fig. 3h-ii). The fingertip with mushroom microstructures demonstrated consistently high adhesion performance throughout over 100 repeated attachment and detachment cycles (Supplementary Fig. S10).

When the fingertip's surface was contaminated with dust, adhesion decreased from 4.18 N (against a flat acrylic plate) to 1.20 N. However, cleaning the fingertip with adhesive tape or water resulted in a full recovery of its original adhesion performance (Supplementary Fig. S10).

Design and working principle of the multi-joint finger

The proposed finger structure consists of two main components: soft structures made of silicone rubber and rigid structures made of ABS (Fig. 4a). The soft portion features inflatable chambers that can be inflated or deflated through pneumatic control. The bending direction of the finger is determined by the shape of the soft chambers and an inextensible layer made by inserting a paper sheet (Fig. 4a, b). The rigid structures, represented by the black round-edge-shaped region in Fig. 4a, provide more powerful and faster movement than a purely soft finger.³¹

FIG. 4.

The multi-joint finger and bending motion. (a) Individual pneumatic control of the finger joints and fingertip. (b) Schematic illustration of the bending mechanism (i) manipulation system, (ii) detailed image of the distal bending mode, and (iii) overall image of the distal bending mode. (c) Bending motions of (i) distal bending, (ii) proximal joint bending, and (iii) two joints bending.

To achieve versatile bending modes, the soft fingers were designed with separate control of two joints: the proximal and distal sections. Three air hoses were connected to each finger through the finger body to enable independent control of the proximal, distal sections, and inflatable fingertips (Fig. 4a-i, ii, and iii). The blue, orange, and pink portions in Fig. 4a denote the air paths to the proximal, distal sections, and fingertips, respectively.

Independent pneumatic control of the proximal and distal joints is achieved by supplying compressed air into the respective sections through tubes (see Fig. 4c, and Supplementary Fig. S11). For example, when compressed air is supplied solely to the distal section, the finger bends at the distal joint due to inflation of the distal chambers (Fig. 4c-i). Similarly, inflating the proximal section results in bending at the proximal joint (Fig. 4c-ii).

Full arc-shaped bending behavior is achieved when compressed air is supplied to both sections, leading to bending at both proximal and distal joints (Fig. 4c-iii). These versatile bending motions enable the finger to grasp objects of varying shapes and sizes through form closure grasping, while also enabling intimate contact between the adhesive fingertips and objects in the fingertip-grasping mode. It is noted that the finger can exhibit rapid bending and releasing motions. The actuation times for full bending and releasing were 1.0 and 0.8 s, respectively.

The bending angles of proximal and distal joints (θ_prox and θ_dis) were evaluated depending on the air pressure supplied to each segment by measuring the positions of four specific points of the finger (gray, blue, orange, and pink dots in Fig. 4c). The bending angles were calculated by assuming the finger as three-segmented linkage structure with two joints. The maximum bending angles for the distal and proximal bending modes were 74° and 55°, respectively, with a maximum allowable pressure of 100 kPa for each distal and proximal section (see Supplementary Description S5 and Supplementary Fig. S12).

Estimation of joint angles by a kirigami-patterned soft sensor

To measure and control the bending motions of the finger, we integrated a stretchable sensor with a kirigami pattern on the upper side of the finger (Fig. 5a). Previous efforts have introduced various soft sensors in soft grippers to overcome the limited stretchability of conventional rigid sensors.^52–56 However, grippers embedded with these prior soft sensors may encounter restricted movement due to their inherent limitations in stretchability.⁵⁷ To address this issue, we integrated a soft sensor with kirigami patterns. This integration effectively reduces the elastic modulus and significantly enhances stretchability. As a result, the kirigami-patterned soft sensor seamlessly covers the complete range of bending motion of the finger, concurrently enhancing reliability and durability.(see Supplementary Description S6, Supplementary Figs. S12 and S13).^39,58–61

FIG. 5.

The kirigami-patterned soft sensor and bending angle estimations through an ANN model. (a) Detailed design of the kirigami-patterned soft sensor (b) Multi-joint angle estimation at free motion with (i) initial state, (ii) distal joint bending, and (iii) distal and proximal joints bending, and (iv) contact with an object. ANN, artificial neural network.

The longitudinal strain applied to the sensor causes a reduction in the cross-section of the microchannel where the eGaIn is located, leading to a change in the electric resistance of the soft sensor.^39,40,62 The bending angles of both the distal and proximal parts were measured by the longitudinal strain of the finger joints. However, the microchannels of the distal sensing part also pass through the proximal sensing parts, leading to an undesired change in the resistance of the distal sensing part when the finger has only proximal bending. To address this issue, a decoupling algorithm was developed for the soft sensors based on an analytical equation that considers the distal, proximal, and both bending motions (see Supplementary Description S7 and Supplementary Fig. S14).

The bending angles in the finger was estimated using the attached sensor with a simple artificial neural network model (Supplementary Fig. S15). The results were compared with those obtained through a vision-based motion capture system (Supplementary Fig. S16). Fig. 5b depicts the estimation of bending angles during the finger's unrestricted motion and upon collision with a dummy object. The proximal and distal angles, as estimated by the soft sensor, exhibited excellent agreement with the angles measured by the motion capture system in both free motion (Fig. 5b(i)–(iii)) and collision states (Fig. 5b(iv)).

The root-mean-squared error between the estimated joint angles and the reference angles for proximal and distal joints were 1.48° and 2.65°, respectively. Using the evaluated bending angles of the finger, we analyzed the reachable area of the gripper (Supplementary Fig. S17).

Grasping and manipulation performance of the soft robotic gripper

To evaluate the multimode grasping capabilities of our soft robotic gripper, we conducted grasping tests on a range of objects with varying shapes and sizes by integrating the gripper into a robotic manipulator (Supplementary Description S8 and Supplementary Fig. S18).⁶³ We used LabVIEW to pre-program the three-dimensional positions of the target objects and the trajectory of the manipulator and gripper system. The manipulator can also be remotely controlled through a Bluetooth connected sensing glove, the Mollisen Hand.⁶⁴ To ensure a secure grasp, a preload was applied to the contact interface between the adhesive microstructures of the fingertip and the object.

Our results demonstrate that the soft gripper can make intimate and firm contact to a wide range of flat or 3D objects (Fig. 6a). Large thin flat objects such as an acrylic sheet (300 × 300 mm) or a Si wafer (10 inch) could be firmly grasped with the adhesive fingertips (Fig. 6a-i, ii). Deformable bumpy 3D objects such as packed ramen or chips (Fig. 6a-iii, iv) and heavy spherical objects such as glass balls (0.8 kg) (Fig. 6a-v) could also be successfully grasped only with the fingertips.

FIG. 6.

Grasping and manipulation performance of the soft gripper. (a) Fingertip-grasping mode for a variety of objects, including (i) an acrylic sheet, (ii) a wafer, (iii) ramen, (iv) chips, and (v) a glass ball. (b) Finger-grasping mode for diverse unstructured objects, including (i) a grape, (ii) a pumpkin, (iii) a radish, and (iv) a doll. (c) Advanced manipulation test of the soft gripper when an object hindered by obstacles.

This indicates that the fingertip of the soft gripper can make firm contact with a wide range of flat or 3D objects. In situations where more complex and unstructured 3D objects need to be grasped, the finger-grasping mode can be employed (Fig. 6b). The multi-joint fingers successfully grasped a variety of arbitrarily shaped unstructured 3D objects by properly actuating the proximal and distal sections of the fingers (Fig. 6b). In addition, the multi-joint finger demonstrated a maximum load capacity of 28.29 N with good repeatability and reliability (Supplementary Description S9 and Supplementary Fig. S19).

We conducted additional tests to evaluate the manipulation capabilities of our soft gripper in more complex scenarios. In this particular scenario, a packaged ramen was selected as the target object and placed it in a location initially obscured by an obstacle (Fig. 6c). Conventional grippers have limited range of motion, making it challenging to access the target object. However, our soft gripper overcame this challenge by employing both adhesion grasping and form-closure grasping modes.

To further demonstrate the versatility of our soft gripper, we conducted a practical experiment demonstrating its ability to perform remotely controlled cooking with a 5 degrees of freedom (DoF) robotic manipulator (Supplementary Description S10, Supplementary Fig. S18, and Supplementary Movie S2). Throughout the entire cooking process, our soft gripper demonstrated reliable grasping and manipulation performance in both the adhesion grasping and form-closure grasping modes for diverse objects of varied shapes, volumes, and even high surface temperatures (Fig. 7 and Supplementary Description S10).

FIG. 7.

Cooking with the soft gripper equipped with a 5-DoF robotic manipulator. (a) Experimental setup for cooking. (b) Pouring a cup of water into the pot on the induction heater through finger-grasping mode. (c, d) Pouring ramen and ramen sauce into the pot using fingertip adhesion and bending motion of the fingers. (e) Handling the hot pot lid with the fingertip. (f) The final step of pouring a cup of egg into the pot through the fingertip-grasping mode.

Conclusions

In conclusion, we have introduced a sensorized multi-fingered soft gripper with adhesive fingertips, capable of providing both fingertip-based adhesion grasping and finger-based form closure grasping modes. The incorporation of adhesive microstructures over the fingertips enabled robust normal adhesion. The maximum force exerted by a single fingertip against a flat acrylic plate was measured at 4.18 N, and this force could be increased to 5.25 N by applying a vacuum (−12 kPa).

The soft fingers, equipped with multiple joints, offered diverse bending motions, ensuring stable grasping of various objects, with a maximum load capacity of 28.29 N for three fingers. In addition, the inclusion of a kirigami-patterned stretchable sensor endowed the soft gripper with motion monitoring and control capabilities.

Our proposed soft gripper integrated different grasping mechanisms into a single soft gripper system, effectively addressing the limitations of each gripping approach. Consequently, it demonstrated the capability to grasp and manipulate a wide range of objects with different shapes, sizes, and curvatures. Furthermore, the practical application of our sensorized soft gripper in remotely controlled cooking highlighted its potential for real-world applications. To further advance the capabilities of our gripper, we plan to develop a more compact design for the adhesive fingertips that offers improved adhesion tunability and adaptability.

In addition, we also aim to integrate pressure or strain sensors into the fingertips for detecting contact and grasping objects, while updating the finger design and control algorithm to enhance control and sensing accuracy. Moreover, the integration of phase change materials could be explored to enhance the gripper with additional functionalities, such as stiffness modulation.^65–67 These advancements will enable more complex manipulation tasks and broaden the range of objects that can be grasped and handled by the soft gripper.

Footnotes

Author's Contributions

W.P and S.P. conceived the research, performed the experiments and analysis of the experimental results, and wrote the article. H.A. and M.S. assisted with the experiments. J.B. and H.E.J. conceived and supervised the research and also wrote the article.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was supported by the National Research Foundation (NRF) of Korea Grant funded by the Korean Government (MSIT; NRF-2021R1A2C3006297, NRF-2019R1A2C2084677, and RS-2023-00208052). This study was also supported by the Technology Innovation Program (00144157) funded By the Ministry of Trade, Industry & Energy (MOTIE) of Korea.

Supplementary Material

References

Fei

, Wang

, Pang

. A novel fabric-based versatile and stiffness-tunable soft gripper integrating soft pneumatic fingers and wrist. Soft Robot, 2019; 6:1–20.

Jain

, Dontu

, Teoh

JEM

, et al. A multimodal, reconfigurable workspace soft gripper for advanced grasping tasks. Soft Robot, 2022; 10:527–544.

Lee

, Chung

, Rodrigue

. Long shape memory alloy tendon-based soft robotic actuators and implementation as a soft gripper. Sci Rep, 2019; 9:1–12.

, Xie

, Wang

, et al. Stiffness-tunable soft gripper with soft-rigid hybrid actuation for versatile manipulations. Soft Robot, 2022; 9:1108–1119.

Liu

, Duo

, Liu

, et al. Touchless interactive teaching of soft robots through flexible bimodal sensory interfaces. Nat Commun, 2022; 13:5030.

Shintake

, Cacucciolo

, Floreano

, et al. Soft robotic grippers. Adv Mater, 2018; 30:1707035.

Terryn

, Brancart

, Lefeber

, et al. Self-healing soft pneumatic robots. Sci. Robot, 2017; 2:eaan4268.

Zhang

, Ni

, Wu

, et al. Pneumatically actuated soft gripper with bistable structures. Soft Robot, 2022; 9:57–71.

Gotz

, Santarossa

, Sack

, et al. Soft particles reinforce robotic grippers: Robotic grippers based on granular jamming of soft particles. Granul Matter, 2022; 24:1–9.

10.

Haninger

, Radke

, Vick

, et al. Towards high-payload admittance control for manual guidance with environmental contact. IEEE Robot Autom Lett, 2022; 7:4275–4282.

11.

Linghu

, Zhang

, Wang

, et al. Universal SMP gripper with massive and selective capabilities for multiscaled, arbitrarily shaped objects. Sci Adv, 2020; 6:eaay5120.

12.

Coulson

, Stabile

, Turner

, et al. Versatile soft robot gripper enabled by stiffness and adhesion tuning via thermoplastic composite. Soft Robot, 2022; 9:189–200.

13.

Sandoval

, Xu

, Adibnazari

, et al. Combining suction and friction to stabilize a soft gripper to shear and normal forces, for manipulation of soft objects in wet environments. IEEE Robot Autom Lett, 2022; 7:4134–4141.

14.

Manti

, Hassan

, Passetti

, et al. A bioinspired soft robotic gripper for adaptable and effective grasping. Soft Robot, 2015; 2:107–116.

15.

Nguyen

, Lopez-Arellano

, Zhang

, et al. Design, characterization, and mechanical programming of fabric-reinforced textile actuators for a soft robotic hand. In: 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE: Macau; 2019; pp. 8312–8317.

16.

Shin

, Park

, Kim

, et al. A universal soft gripper with the optimized fin ray finger. Int J Precis Eng Man-G.T. 2021; 8:889–899.

17.

Wang

, Ahn

. Shape memory alloy-based soft gripper with variable stiffness for compliant and effective grasping. Soft Robot, 2017; 4:379–389.

18.

Jiang

, Hawkes

, Fuller

, et al. A robotic device using gecko-inspired adhesives can grasp and manipulate large objects in microgravity. Sci Robot, 2017; 2:eaan4545.

19.

Koivikko

, Drotlef

, Sitti

, et al. Magnetically switchable soft suction grippers. Extreme Mech Lett, 2021; 44:101263.

20.

Tian

, Li

, Shao

, et al. Gecko-effect inspired soft gripper with high and switchable adhesion for rough surfaces. Adv Mater Interfaces, 2019; 6:1900875.

21.

Wang

, Liu

, Zheng

, et al. In: Intelligent Robotics and Applications: 11th International Conference, ICIRA 2018, Newcastle, NSW, Australia, August 9–11. Proceedings, Part II 11. Springer: New York, USA; 2018; pp. 112–121.

22.

Fang

, Sun

, Wu

, et al. Multimode grasping soft gripper achieved by layer jamming structure and tendon-driven mechanism. Soft Robot, 2022; 9:233–249.

23.

Yang

, Zhang

, Kan

, et al. Hybrid jamming for bioinspired soft robotic fingers. Soft Robot, 2020; 7:292–308.

24.

Zhang

, Zhang

, Gao

, et al. Finger-palm synergistic soft gripper for dynamic capture via energy harvesting and dissipation. Nat Commun, 2022; 13:7700.

25.

Bircher

, Morgan

, Dollar

. Complex manipulation with a simple robotic hand through contact breaking and caging. Sci Robot, 2021; 6:eabd2666.

26.

Ruotolo

, Brouwer

, Cutkosky

. From grasping to manipulation with gecko-inspired adhesives on a multifinger gripper. Sci Robot, 2021; 6:eabi9773.

27.

Zhou

, Ren

, Chen

, et al. Bio-inspired soft grippers based on impactive gripping. Adv Sci, 2021; 8:2002017.

28.

Chin

, Barscevicius

, Lipton

, et al. Multiplexed manipulation: Versatile multimodal grasping via a hybrid soft gripper. In: 2020 IEEE International Conference on Robotics and Automation (ICRA). IEEE: Paris, France (virtual conference); 2020; pp. 8949–8955.

29.

Han

, Hajj-Ahmad

, Cutkosky

. Bimanual handling of deformable objects with hybrid adhesion. IEEE Robot Autom Lett, 2022; 7:5497–5503.

30.

Hao

, Biswas

, Hawkes

, et al. A multimodal, enveloping soft gripper: Shape conformation, bioinspired adhesion, and expansion-driven suction. IEEE Trans Robot, 2021; 37:350–362.

31.

Tawk

, Sariyildiz

, Alici

. Force control of a 3D printed soft gripper with built-in pneumatic touch sensing chambers. Soft Robot, 2022; 9:970–980.

32.

Gwon

, Park

, Hong

, et al. Soft directional adhesion gripper fabricated by 3D printing process for gripping flexible printed circuit boards. Int J Precis Eng Man-G.T. 2022; 9:1–13.

33.

Lee

, Baik

, Hwang

, et al. An electronically perceptive bioinspired soft wet-adhesion actuator with carbon nanotube-based strain sensors. ACS Nano, 2021; 15:14137–14148.

34.

Song

, Drotlef

, Majidi

, et al. Controllable load sharing for soft adhesive interfaces on three-dimensional surfaces. PNAS, 2017; 114:E4344–E4353.

35.

Kim

, Kim

, Pyun

, et al. A substrate-less nanomesh receptor with meta-learning for rapid hand task recognition. Nat Electron, 2023; 6:64–75.

36.

, Kim

, Lee

, et al. A liquid metal based multimodal sensor and haptic feedback device for thermal and tactile sensation generation in virtual reality. Adv Funct Mater, 2021; 31:2007772.

37.

Kim

, Ha

, Kim

, et al. A deep-learned skin sensor decoding the epicentral human motions. Nat Commun, 2020; 11:2149.

38.

, Lee

, Ko

, et al. Ultra-adaptable and wearable photonic skin based on a shape-memory, responsive cellulose derivative. Adv Funct Mater, 2019; 29:1902720.

39.

Kim

, Oh

, Jeong

, et al. Direct wiring of eutectic gallium-indium to a metal electrode for soft sensor systems. ACS Appl Mater Interfaces, 2019; 11:20557–20565.

40.

Kim

, Oh

, Jeong

, et al. Consistent and reproducible direct ink writing of eutectic gallium-indium for high-quality soft sensors. Soft Robot, 2018; 5:601–612.

41.

Park

, Seo

, Bae

. A hybrid gripper with soft material and rigid structures. IEEE Robot Autom Lett, 2019; 4:65–72.

42.

Pagoli

, Chapelle

, Corrales

, et al. A soft robotic gripper with an active palm and reconfigurable fingers for fully dexterous in-hand manipulation. IEEE Robot Autom Lett, 2021; 6:7706–7713.

43.

Samadikhoshkho

, Zareinia

, Janabi-Sharifi

. A brief review on robotic grippers classifications. In: 2019 IEEE Canadian Conference of Electrical and Computer Engineering. IEEE: Paris, France (virtual conference); 2019; pp. 1–4.

44.

Park

, Seo

, Oh

, et al. A sensorized hybrid gripper to evaluate a grasping quality based on a largest minimum wrench. IEEE Robot Autom Lett, 2020; 5:3243–3250.

45.

Yang

, Wu

, Li

, et al. Direct ink writing of fluoropolymer/CNT-based superhydrophobic and corrosion-resistant electrodes for droplet energy harvesters and self-powered electronic skins. Nano Energy, 2021; 86:106095.

46.

Guo

, Ni

, Li

, et al. Designing mechanical metamaterials with Kirigami-inspired, hierarchical constructions for giant positive and negative thermal expansion. Adv Mater, 2021; 33:2004919.

47.

Hwang

, Seong

, Yi

, et al. Low-Resistant electrical and robust mechanical contacts of self-attachable flexible transparent electrodes with patternable circuits. Adv Funct Mater, 2020; 30:2000458.

48.

Arzt

, Enders

, Gorb

. Towards a micromechanical understanding of biological surface devices. Int J Mater Res, 2022; 93:345–351.

49.

Maksuta

, Dalvi

, Gujrati

, et al. Dependence of adhesive friction on surface roughness and elastic modulus. Soft Matter, 2022; 18:5843–5849.

50.

Zhang

, Ma

, Li

, et al. Gecko's feet-inspired self-peeling switchable dry/wet adhesive. Chem Mat, 2021; 33:2785–2795.

51.

Seong

, Hwang

, Park

, et al. Enhanced thermal transport across self-interfacing van der waals contacts in flexible thermal devices. Adv Funct Mater, 2021; 31:2107023.

52.

Bai

, Li

, Barreiros

, et al. Stretchable distributed fiber-optic sensors. Science, 2020; 370:848–852.

53.

Frutiger

, Muth

, Vogt

, et al. Capacitive soft strain sensors via multicore-shell fiber printing. Adv Mater, 2015; 27:2440–2446.

54.

Jin

, Sun

, Li

, et al. Triboelectric nanogenerator sensors for soft robotics aiming at digital twin applications. Nat Commun, 2020; 11:5381.

55.

Karipoth

, Pullanchiyodan

, Christou

, et al. Graphite-based bioinspired piezoresistive soft strain sensors with performance optimized for low strain values. ACS Appl Mater Interfaces, 2021; 13:61610–61619.

56.

Yong

, Chapman

, Aw

. Soft and flexible large-strain piezoresistive sensors: On implementing proprioception, object classification and curvature estimation systems in adaptive, human-like robot hands. Sens Actuator A Phys, 2022; 341:113609.

57.

Zuo

, Zhou

, Ying

, et al. A soft robotic gripper with anti-freezing ionic hydrogel-based sensors for learning-based object recognition. IEEE Int Conf Robot Autom, 2021:12164–12169.

58.

Jung

, Koirala

, Lee

, et al. Solvent-assisted filling of liquid metal and its selective dewetting for the multilayered 3D interconnect in stretchable electronics. ACS Nano, 2022; 16:21471–21481.

59.

Truby

, Della Santina

, Rus

. Distributed proprioception of 3D configuration in soft, sensorized robots via deep learning. IEEE Robot Autom Lett, 2020; 5:3299–3306.

60.

Wang

, Zhang

, Duan

, et al. Kirigami-patterned highly stretchable conductors from flexible carbon nanotube-embedded polymer films. J Mater Chem C, 2017; 5:8714–8722.

61.

Choi

, Luo

, Olson

, et al. Highly stretchable and strain-insensitive liquid metal based elastic Kirigami Electrodes (LM-eKE). Adv Funct Mater, 2023; 33:2301388.

62.

Kim

, Lim

, Ko

. Liquid metal patterning and unique properties for next-generation soft electronics. Adv Sci, 2023; 10:2205795.

63.

Park

, Jung

, Bae

. An interactive and intuitive control interface for a tele-operated robot (AVATAR) system. Mechatronics, 2018; 55:54–62.

64.

Soft sensor-based glove “Mollisen HAND” announced (2023). Available from: www.ftsame.com [Last accessed: February 24, 2024].

65.

Long

, Cheng

, Ren

, et al. Latest advances in development of smart phase change material for soft actuators. Adv Eng Mater, 2022; 24:2100863.

66.

Yoon

, Park

, Lee

, et al. Bioinspired untethered soft robot with pumpless phase change soft actuators by bidirectional thermoelectrics. Chem Eng J, 2023; 451:138794.

67.

, Duan

, Lv

, et al. Soft actuators based on liquid-vapor phase change composites. Soft Robot, 2021; 8:251–261.

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.

1.74 MB

22.48 MB

1.89 MB

0.00 MB