Resistive Sensing in Soft Robotic Grippers: A Comprehensive Review of Strain, Tactile, and Ionic Sensors

Abstract

Soft robotic grippers have emerged as crucial tools for safe and adaptive manipulation of delicate and different objects, enabled by their compliant structures. These grippers need embedded sensing that offers proprioceptive and exteroceptive feedback in order to function consistently. Resistive sensing is unique among transduction processes since it is easy to use, scalable, and compatible with deformable materials. The three main classes of resistive sensors used in soft robotic grippers are systematically examined in this review: ionic sensors, which are emerging multimodal devices that can capture both mechanical and environmental cues; tactile sensors, which detect contact, pressure distribution, and slip; and strain sensors, which monitor deformation and actuation states. Their methods of operation, material systems, fabrication techniques, performance metrics, and integration plans are all compared in the survey. The results show that sensitivity, linearity, durability, and scalability are all trade-offs across sensor categories, with ionic sensors showing promise as a new development for multipurpose soft grippers. There is also a discussion of difficulties, including hysteresis, long-term stability, and signal processing complexity. In order to move resistive sensing from lab prototypes to reliable, practical applications in domains like healthcare, food handling, and human–robot collaboration, the review concludes that developments in hybrid material systems, additive manufacturing, and AI-enhanced signal interpretation will be crucial.

1. Introduction

1.1. Robotic Grippers: Definition and Importance

Robotic grippers are the main connection between a robotic system and its environment [1,2]. They are specialized end-effectors made to grasp, hold, and manipulate items [3,4]. Grippers allow robots to carry out necessary activities like choosing, placing, assembling, or moving objects by transforming actuator inputs into defined contact forces [5]. Grippers can have everything from basic grasping systems to complex multi-fingered hands that imitate human holding, depending on how they are made [6]. From delicate handling in food processing, healthcare, and dangerous situations such as space exploration or deep-sea operations to high-precision assembly in electronics and vehicle manufacturing, they are essential in a wide range of industries [7,8,9].

The significance of grippers is seen in their capacity to blend accuracy and flexibility [10]. A reliable gripper must be strong enough to hold things in place without breaking delicate things such as thin chips, fruits, or medical equipment [11]. In service robots and assistive technologies, where safety and human-robot interaction are crucial factors, Grippers play a role that extends beyond industrial automation [12]. In the next years, the world market of robotics is expected to grow remarkably, overcoming the market of industrial robotics [13,14]. The capabilities of grippers have been further enhanced by recent developments in materials and actuation techniques, including soft elastomers, shape-memory alloys, and additive manufacturing, which enable them to reliably manipulate delicate objects and complex shapes [15,16,17]. Grippers continue to be essential as robotics advances, and sensors and clever control schemes are improving their functionality [18,19].

1.2. Soft Robotics and Soft Grippers

Soft robotics is an emerging field that employs compliant and deformable materials to create robots capable of safe, adaptive, and human-friendly interactions [20,21,22]. Soft robots, as opposed to conventional rigid robots, are able to continually deform, adapt to uneven surfaces, and work securely around people or delicate goods [23,24]. Because of their versatility, they are extremely useful in fields where careful handling, adaptation, and safety are essential, such as healthcare [25], agriculture [26], food handling [27], and environmental investigation [28]. Recent developments in compliant actuation and morphological adaptability have significantly advanced the capabilities of soft robots. Comprehensive analyses of these trends can be found in recent reviews [29,30], which further emphasize the role of soft robotic grippers in safe human–robot collaboration and delicate manipulation tasks.

One of the most researched and significant applications in this field is soft grippers [31]. The ability of soft grippers to cover and conform to objects of different sizes, shapes, and textures without knowing their geometry beforehand was inspired by the adaptability of the human hand [32]. Soft grippers, which are powered by actuation techniques like tendons, pneumatics, or shape-memory alloys, and constructed from materials like silicone elastomers [33], hydrogels [34], or fabric composites [35], provide a degree of versatility that rigid designs simply cannot match [33]. Their exceptional value is emphasized by their capacity to handle delicate objects like eggs, fruits, or biomedical samples without causing harm, which establishes them as crucial instruments for automation of the future [35,36].

1.3. The Role of Sensing in Robotic Grippers

Sensing is essential for robotic grippers to execute dependable and skillful manipulation [37]. Similar to how the human hand uses nerves to sense motion, pressure, and texture, grippers also require sensors to understand their own condition and the characteristics of the objects they are handling [38]. Grippers that lack embedded sensing function in an open-loop method, which raises the possibility of overuse, slippage, or unsuccessful grasps [13,39,40]. In order to estimate object shape, weight, compliance, and position parameters that are crucial for modifying grasping techniques in real time, sensors offer the essential feedback required [41,42].

In addition to object characterization, sensing improves safety and grip stability [43]. These sensors measure internal deformation or finger bending, enabling the gripper to track its own actuation state [44]. In contrast, tactile sensors detect contact, pressure distribution, and slip events to provide sensory feedback. When combined, these feedback systems enable intelligent interaction between robotic grippers, particularly soft grippers, and complex, unpredictable environments [7]. By adding sensing, grippers can do more tasks, such as industrial automation, biomedical support, and human-robot collaboration, in addition to increasing grasp reliability [45,46,47].

1.4. Flexible and Resistive Sensors: An Overview

Several transduction mechanisms have been explored to enable sensing in robotic grippers, including capacitive [48], optical, magnetic [49], piezoelectric [50], and resistive approaches [51]. An overview of these sensing approaches is illustrated in Figure 1. Despite their high sensitivity, capacitive sensors frequently need sophisticated shielding against electromagnetic interference [52,53]. Although optical sensors have high resolution, their integration into soft materials is limited by their reliance on large light sources and detectors [54]. Large deformations can be detected by magnetic and inductive sensors, although they usually require more power and have rigid parts [55]. Even though each of these approaches offers useful features, many of them are still too expensive, inflexible, or challenging to scale for broad application in soft robotics [56].

Resistive sensors have become especially popular among these choices because of their ease of use, affordability, mechanical compliance, and simplicity of signal readout [57]. A voltage divider or Wheatstone bridge is an example of basic circuitry that can be used to measure the variation in electrical resistance that occurs when a resistive element is subjected to strain, pressure, or deformation [58,59]. The deformable structures of soft robotic grippers are also highly compatible with resistive sensors because they can be made from a variety of soft, stretchable, and printable materials, such as carbon-based composites, conductive polymers, TPU, PLA, textiles, and hydrogels [60,61,62,63]. Resistive sensing is one of the most promising methods for obtaining flexible, scalable, and robust sensing in soft robotics because of these features [64].

In addition to resistive sensing, capacitive, optical, and hybrid multimodal sensing techniques for soft robotics have advanced quickly in the last few years (2021–2024). While optical sensors have shown high-resolution deformation mapping using fiber Bragg gratings and elastomer wave guides, capacitive devices have been tuned for low-power and stretchable applications [65]. More significantly, multimodal sensing techniques that combine resistive components with magnetic, capacitive, or piezoelectric transducers are becoming more popular [66]. Typical performance ranges reported in the literature include resistive sensors (gauge factor 2–200), capacitive sensors (0.05–1 pF/kPa), and piezoelectric sensors (0.1–10 V/N). Despite these representative values, systematic quantitative benchmarking across modalities remains inconsistent, as protocols for strain amplitude, loading frequency, and calibration reference differ across studies. Establishing standardized benchmarking methods remains an open challenge for fair comparison. These techniques allow for enhanced robustness and richer feedback in unstructured settings. These advancements emphasize the significance of comparative studies across modalities and offer useful standards by which resistive sensors can be assessed.

These sensing principles and material platforms are schematically illustrated in Figure 2.

1.5. Classification of Resistive Sensors

Resistive sensors in soft robotic grippers can be grouped into three main categories: Ionic sensors differ from conventional resistive sensors primarily in their charge transport mechanism. Whereas traditional resistive sensors rely on electron conduction through metallic or carbon-based networks, ionic sensors generate signals via the migration of mobile ions in a gel or liquid medium, coupling mechanical deformation with ionic displacement. Strain sensors monitor how the finger bends, stretches, or compresses to provide proprioceptive feedback [67]. They are frequently made from liquid-metal microchannels [68], conductive elastomers [66], and carbon-based nanocomposites like graphene or carbon nanotubes (CNTs) [69]. These designs preserve mechanical compliance while enabling grippers to estimate actuation states and curvature [70]. By identifying normal and shear forces, slip events, and distributed pressure, tactile sensors, on the other hand, provide perceptual information [71,72]. Velostat-based resistive arrays, silicone-embedded pneumatic resonant tubes [73], liquid-metal microfluidic channels [74], and printed nanocomposite [75] e-skins are a few examples that strike a different balance between cost, sensitivity, durability, and fabrication complexity [76].

A more recent subcategory of resistive sensing that moves toward multipurpose use is ionic sensors [77]. Constructed from hydrogels or ionogels, they provide characteristics, like stretchability, transparency, antifreezing resilience [78], and self-healing, while taking advantage of changes in ionic conductivity under pressure or strain [79]. Ionic sensors are excellent candidates for next-generation soft robotic applications because, in contrast to conventional resistive devices, they allow for the simultaneous detection of strain, pressure, and occasionally environmental cues [78,80]. By concentrating on these three classes, this review covers the complementary roles of proprioceptive and exteroceptive feedback in facilitating dexterous and adaptive manipulation, highlighting both the well-established foundations (strain and tactile sensors) and the cutting-edge limits (ionic sensors) of resistive sensing in soft grippers [31].

1.6. Applications of Resistive Sensors in Soft Grippers

The integration of resistive sensors into soft grippers has enabled a variety of uses that extend beyond mere grabbing [81]. Strain sensors enable sensory monitoring, which enables grippers to measure applied forces, actuation states, and bending angles [56]. This feature ensures that the grip force is enough to secure objects without causing damage by supporting real-time feedback control [82]. This capability is increased by tactile sensors, which offer exteroceptive feedback that enables the detection of surface texture, contact location, pressure distribution, and slip [83,84]. These features are essential for tasks where accuracy and safety are equally crucial, like handling delicate foods, manipulating biomedical samples, and assembling electronics [31,85].

Multimodal resistive sensing has led to more complex applications [86]. For example, ionic sensors provide special chances for adaptive manipulation in unstructured environments by allowing the simultaneous detection of strain, pressure, and environmental conditions [87]. Resistive sensing has been used in a variety of fields, including force control [88], slip detection [89], object recognition [90], classification, and safe human–robot interaction [91]. These illustrations demonstrate how resistive sensors are crucial in converting soft grippers from passive, obedient machines into intelligent systems that can manipulate objects deftly, independently, and in context [7,20]. Beyond laboratory demonstrations, resistive sensing technologies are increasingly transitioning toward real-world and industrial applications. In the food industry, soft grippers equipped with resistive tactile sensors enable delicate handling, sorting, and packaging of produce without causing damage. In healthcare and rehabilitation, stretchable hydrogel-based sensors integrated into assistive gloves and soft exosuits provide lightweight, low-cost, and biocompatible solutions for motion monitoring. Furthermore, resistive e-skin arrays play a pivotal role in human–robot interaction, enhancing safety, compliance, and adaptive grip control in collaborative robotic systems.

1.7. Materials and Fabrication Considerations

The performance of resistive sensors in soft grippers is highly dependent on the materials used in their design [66,92]. Numerous alternatives have been investigated, each with unique mechanical and electrical characteristics [93]. Because of their low cost [66], flexibility [94], and adjustable conductivity [95], carbon-based composites (like carbon nanotubes, graphene, and carbon black) and conductive polymers (like PEDOT: PSS and polypyrrole) [96] are widely used. Stretchable, customizable sensor geometries are made possible by the compatibility of elastomer blends like TPU and PLA with 3D printing [97,98]. Additional material strategies include liquid metals (e.g., EGaIn [99]) that offer remarkable stretchability but have durability and leakage issues, conductive foams for pressure sensing [100], and textiles and fabrics (Velostat, EeonTex) for affordable tactile [101]. The stretchability, transparency, antifreezing properties, and potential for multimodal sensory perception of hydrogels and ionogels have caused interest in recent years [102].

Fabrication techniques play an additionally important role in determining scalability and integration [66,103]. For the production of thin-film or foam-based resistive elements, conventional techniques like screen printing, spray coating, and molding are still helpful [104]. The ability to directly pattern ionogels, hydrogels, or conductive inks within elastomer matrices through embedded 3D printing (EMB3D) has made it possible to create intricate sensor architectures like fingertip tactile pads or U-shaped curvature traces [97,105,106]. Enhancing durability, adhesion, and resistance to environmental effects like dehydration or leakage also requires packaging techniques, which frequently use silicone elastomers (such as EcoFlex) [107]. The sensitivity, stretchability, and long-term stability of resistive sensors are determined by the combination of material selection and fabrication technique, which directly influences how well they work in soft robotic grippers [64,94,108]. Fabrication reproducibility is strongly influenced by parameters such as curing time, temperature, and humidity during elastomer crosslinking or hydrogel polymerization. These parameters affect both mechanical compliance and the electrical stability of the sensors and should be carefully controlled during fabrication.

The curing time of the elastomeric matrices is an important reproducibility parameter that has a direct impact on the mechanical and electrical stability of the sensors [109]. For instance, Dragon Skin™ 30 requires about 60–90 min under ambient conditions [110], while Ecoflex™ 00-30 silicone is generally reported to cure within 3–4 h at room temperature [111]. Depending on the additives used, hydrogel-based devices, like polyacrylamide–alginate double networks, typically take one to two hours to polymerize and crosslink [112]. Reporting these details is crucial to guaranteeing that sensor properties can be reliably reproduced in other labs, even though curing times may differ depending on formulation and temperature [107].

The timeline shows Figure 3 the evolution of resistive sensing in soft robotics from 2012 to 2024 to give historical context. Early works (2012–2015) focused on low-cost Velostat tactile arrays and strain films made of carbon nanotubes. Hybrid fabrication techniques and 3D-printed strain and tactile sensing architectures were introduced in mid-decade advancements (2016–2020). More recent advancements (2021–2024) emphasize liquid-metal composites, accessible direct-ink-writing techniques, and ionic hydrogel sensors with self-healing and antifreezing capabilities, signaling a shift toward scalable, multipurpose platforms appropriate for industrial deployment (

Table 1. Comparison of materials used for resistive sensors in soft robotic grippers.

Material Type	Examples	Key Properties	Advantages	Limitations/Typical Use
Carbon-based Composites	CNTs, Graphene, Carbon black	High conductivity, tunable sensitivity, flexible	High sensitivity, versatile fabrication	Hysteresis, nonlinear response; used for strain sensors and curvature monitoring
Conductive Polymers	PEDOT:PSS, Polypyrrole	Stretchable, processable, chemically stable	Good integration with soft substrates	Lower conductivity vs metals; used in flexible strain/tactile pads
TPU/PLA Elastomers	Thermoplastic polyurethane (TPU), PLA blends	3D-printable, customizable geometries	Additive manufacturing compatibility	Limited durability under cyclic strain; used in custom 3D-printed strain sensors
Textiles and Fabrics	Velostat, EeonTex	Low-cost, conformable, scalable	Scalable for large-area tactile skins	Low sensitivity, durability issues; used for low-cost tactile arrays
Conductive Foams	PU foams, CNT foams	Compressible, pressure-sensitive, porous	Simple fabrication, good pressure response	Hysteresis, mechanical fatigue; used in pressure pads for contact detection
Hydrogels/Ionogels	Polyacrylamide hydrogels, Ionogels	Stretchable, transparent, antifreezing, self-healing	Multimodal sensing, biocompatibility	Dehydration, mechanical fragility; used in multimodal strain/tactile sensors
Liquid Metals	EGaIn, Galinstan	Highly conductive, liquid phase, deformable	Excellent stretchability, wide strain range	Leakage, oxidation, stability issues; used in stretchable strain and tactile sensors

).

Flexible and stretchable sensors have become central to soft robotics applications (Practical Considerations—Carbon composites and textile-based sensors are cost-effective and easily scalable, while liquid metals and conductive polymers generally exhibit higher cost but superior sensitivity. Shelf life depends mainly on polymer oxidation or hydrogel dehydration during storage. Materials compatible with additive manufacturing or roll-to-roll processing show strong potential for large-scale production.

1.8. Challenges and Future Directions

Despite significant progress, the integration of resistive sensors into soft robotic grippers still faces several challenges [113,114]. Nonlinearity and hysteresis are a common problem that restricts accurate calibration [115], especially in carbon-based composites and conductive foams. Because many resistive elements undergo breakdown, mechanical fracture, or leakage after repeated deformation, durability under cyclic loading is still an issue [56,116]. Dehydration and unstable environments may affect the long-term performance of hydrogels and ionogels [117]. However, issues like oxidation and leakage make it challenging to use liquid metals in useful, closed designs [118]. Large-scale tactile arrays require a high channel number and signal processing, which may limit flexibility [119]. In addition to materials, challenges include sensor wiring and data gain complexity [120]. In addition, long-term stability under cyclic loading remains a significant challenge. The industrial potential of many resistive materials is limited by their mechanical breakdown, signal drift, or delamination after thousands of deformation cycles. Another crucial issue is environmental robustness: liquid-metal systems experience oxidation and leaking problems, and hydrogels, ionogels, and conductive foams are significantly impacted by temperature changes, humidity, and dehydration [115,117,118]. In order to scale resistive sensors from lab prototypes to dependable, real-world deployment, these durability factors must be addressed.

Quantitative evaluation of durability parameters—such as dehydration or ionic leakage under ambient conditions—remains limited. Reported studies indicate gradual conductivity decay of 5–15% over several days of air exposure, suggesting that encapsulation strategies remain essential for long-term stability.

Moreover, recent advances in liquid-metal microchannel confinement, oxide-layer stabilization, and polymer encapsulation have effectively minimized leakage risks while maintaining flexibility [121].

To promote fair comparison among sensor technologies, a standardized benchmarking framework could include parameters such as gauge factor, hysteresis percentage, response/recovery time, and cycle stability under defined deformation amplitudes and frequencies.

Finally, signal interference among multimodal hybrid sensors can be mitigated through electrode decoupling, frequency-domain filtering, or AI-based feature separation, enabling reliable multimodal sensing in complex tasks [122].

In the future, resistive sensing in soft grippers could advance in a number of ways. By integrating tactile feedback and strain on the same material platform, hybrid multimodal sensors have the potential to provide richer data while simplifying wiring [123]. Sensor durability and lifetime can be increased under real-world operating conditions using encapsulation techniques and self-healing materials [124]. Customized sensor geometries that can be directly incorporated into soft actuators will be made possible by developments in additive manufacturing and 3D printing. Furthermore, soft grippers may become intelligent [125], adaptable systems if resistive sensors are integrated into closed-loop control architectures and combined with AI-driven signal processing. Resolving these issues is essential to moving resistive sensing from lab prototypes to reliable, wide implementation in sectors like food handling and healthcare [126].

1.9. Roadmap of This Review

This review has been arranged to provide a structured overview of resistive sensing in soft robotic grippers, focusing on the three main sensor categories: strain, tactile, and ionic sensors. The history, operation, and significance of resistive sensors in relation to soft robotic grippers are covered in Section 2. The working mechanisms, materials, fabrication techniques, representative designs, and integration strategies of strain sensors are reviewed in Section 3. Tactile sensors are covered in Section 4, with a focus on their transduction mechanisms, manufacturing processes, performance indicators, and uses. Ionic sensors are highlighted as a new multimodal class in Section 5, which also discusses their integration with soft grippers, special material systems, and structural issues. In order to draw attention to remaining research questions, each section ends with challenges and future directions.

In addition to these technical reviews, this paper also identifies overarching challenges in durability, scalability, and signal accuracy, which are addressed in Section 1.8. Quantitative comparisons of resistive sensors with other well-known sensing modalities, such as capacitive, optical, and piezoresistive systems, should be a part of future studies. It is still difficult to make systematic comparisons with standardized metrics (such as gauge factor, response time, hysteresis, error margins, and repeatability across multiple trials), even though many studies offer subjective demonstrations. A more thorough assessment of resistive sensors’ sensitivity, robustness, and scalability, as well as an explanation of their relative advantages and disadvantages in comparison to other methods, would be possible with the inclusion of such benchmarking [61,115,127]. Translating resistive sensing from lab prototypes to dependable, industrial-grade soft robotic applications will require this approach.

2. Background and Working Principles of Resistive Sensors

2.1. Fundamentals of Resistive Sensing

Resistive sensors operate on the principle that the electrical resistance of a material changes under mechanical or environmental stimuli [81,128]. The conductive channels in the sensing material are changed by strain, pressure, or deformation, which causes a detectable variation in resistance [129]. This response can be described using Ohm’s principle and basic geometric relationships, where resistance is proportional to the material’s length and inversely proportional to its cross-sectional area and conductivity [130,131]. This simple transfer method is especially attractive in the context of soft robotics since it may be used with flexible, compliant materials without losing flexibility [81].

2.2. Relevance to Soft Robotic Grippers

Resistive sensing provides two essential features for soft grippers. First, feedback on grip force, actuation states, and finger bending is provided by proprioception, which is accomplished using strain sensors [132]. Second, contact events, pressure distribution, slide, and object compliance can all be detected using feedback, which is provided by tactile sensors [133]. By integrating multimodal detection of strain, pressure, and environmental changes into a single device, emerging ionic sensors further expand these capabilities [87,134]. When combined, these features are crucial for manipulating things safely and adaptively, especially when handling delicate or unusually shaped objects [135].

2.3. Integration Considerations

Because resistive sensing is so simple, it can be read out using circuits like Wheatstone bridges or voltage dividers [51,127]. Practical integration in soft grippers, however, involves consideration of sensor placement, fabrication methods, and material selection [64,66]. Sensitivity, stretchability, durability, and scalability are trade-offs that have a direct impact on performance in practical applications [92]. Later sections, which include a detailed examination of strain, tactile, and ionic resistive sensors, will go further into these factors.

3. Strain Sensor

3.1. Working Principle

Strain sensors used in soft robotic grippers primarily operate based on the piezoresistive effect, where mechanical deformation causes a measurable change in electrical resistance [136]. The conductive path inside the sensor changes length or cross-section when a gripper finger compresses, stretches, or bends, thereby altering the resistance [103]. This variation can be linked to applied force, curvature, or bending angles [137]. Other strain-sensing methods include pneumatic sensing, which uses implanted air passages to sense strain by detecting changes in internal pressure or airflow resistance [138], and capacitive sensing, which uses deformation to alter the distance between electrodes [61].

3.2. Materials and Fabrication Techniques

The performance of resistive strain sensors in soft robotic applications depends on a carefully chosen combination of compliant materials, conductive phases, and fabrication techniques that keep flexibility while ensuring reliable electrical response [107,139]. Soft gripper sensors need to be able to resist significant deformations, stay attached to flexible surfaces, and demonstrate consistent electromechanical performance across several cycles Table 2 [140,141].

Recent studies highlight the unique opportunities that additive manufacturing, particularly multimaterial 3D printing, offers to mix conductive phases with soft matrices, allowing complicated bending geometries with dependable production [142]. The significance of silicone rubbers, TPU, carbon-based composites, and architectural features like wrinkled or spiral patterns that improve stretchability and linearity are highlighted in broader studies on flexible piezoresistive sensors [143]. Furthermore, because of their simplicity and speed of integration, resistive sensors are well known within the family of smart materials that enable soft robotic systems [81].

Table 2. Representative resistive strain sensors for soft robotic grippers, comparing materials, structures, and performance metrics.

Material System	Structure / Fabrication	Electrical Range	Dimensions	Linearity	Ref.
Dragon Skin 30 elastomer + MWCNTs (non-conductive elastic pillar)	Spiral conductive fiber wound on an elastic base pillar	–	Spiral fiber ⌀ 0.6 mm	Linearity	[137]
Conductive PLA	Direct 3D printing	∼15–27 k$\Omega$	52 × 19 mm; track thickness 0.3 mm	Most	[144]
Carbon Nanotubes (CNTs)	Serpentine-shaped stretchable interconnects (curved arc of 260°)	Resistance changes up to 1300% at 11% strain	CNT density: 3 CNT/$\mu$m2 @ 5 V; 5.67 CNT/$\mu$m2 @ 15 V	Linear behavior up to 9% strain	[145]
Conductive TPU-based filament	Piezoresistive layer-by-layer FDM printing on TPU substrate	$2.5\times {10}^4$–$8\times {10}^4 \Omega$	130 × 10 × 0.3 mm	Linear up to 5% strain	[103]
Conductive TPU (CTPU)	3D printing	14.48 k$\Omega$ (rest) → 12.45 k$\Omega$ (at 300 kPa)	130 × 10 × 0.3 mm	Linear up to 5% strain	[146]
Soft silicone elastomers	Molded air-filled microchannels (pneumatic strain gauge)	Supply pressure 60 kPa; resistance 0.71–1.2 (normalized)	183 mm length; channel 200 $\mu$m × 200 $\mu$m	Nonlinear behavior	[147]

Table 2. Representative resistive strain sensors for soft robotic grippers, comparing materials, structures, and performance metrics.

Material System	Structure / Fabrication	Electrical Range	Dimensions	Linearity	Ref.
Dragon Skin 30 elastomer + MWCNTs (non-conductive elastic pillar)	Spiral conductive fiber wound on an elastic base pillar	–	Spiral fiber ⌀ 0.6 mm	Linearity	[137]
Conductive PLA	Direct 3D printing	∼15–27 k$\Omega$	52 × 19 mm; track thickness 0.3 mm	Most	[144]
Carbon Nanotubes (CNTs)	Serpentine-shaped stretchable interconnects (curved arc of 260°)	Resistance changes up to 1300% at 11% strain	CNT density: 3 CNT/$\mu$m2 @ 5 V; 5.67 CNT/$\mu$m2 @ 15 V	Linear behavior up to 9% strain	[145]
Conductive TPU-based filament	Piezoresistive layer-by-layer FDM printing on TPU substrate	$2.5\times {10}^4$–$8\times {10}^4 \Omega$	130 × 10 × 0.3 mm	Linear up to 5% strain	[103]
Conductive TPU (CTPU)	3D printing	14.48 k$\Omega$ (rest) → 12.45 k$\Omega$ (at 300 kPa)	130 × 10 × 0.3 mm	Linear up to 5% strain	[146]
Soft silicone elastomers	Molded air-filled microchannels (pneumatic strain gauge)	Supply pressure 60 kPa; resistance 0.71–1.2 (normalized)	183 mm length; channel 200 $\mu$m × 200 $\mu$m	Nonlinear behavior	[147]

3.3. Representative Sensor Materials and Structures

Different material–structure combinations reported in research illustrate these trade-offs. Because of its suitability and compliance with fused deposition modeling (FDM), conductive TPU has been shown to be a successful option. Reliable performance is maintained through repetitive bending of multimaterial printed sensors based on TPU, which exhibit linear resistance changes under strain [137,146]. Despite being more rigid, conductive PLA allows for quick and cheap prototyping; when implanted in flexible substrates, it shows very linear bending responses but has little stretchability [144]. With helical and spiral structures enhancing resistance fluctuations by more than 1000% at relatively tiny stresses, carbon nanotube (CNT) networks achieve substantially better sensitivity due to flexible diffusion paths [137,145]. Silicone elastomers can be molded with microchannels to create pneumatic strain gauges; these systems allow extreme stretchability but often display nonlinear responses that require calibration [147].

3.4. Discussion of Trends

These examples demonstrate clear design trade-offs, which are summed up in Table 1. Although they require difficult and expensive production processes, CNT-based sensors offer extremely high sensitivity. Robotic fingers can use TPU-based sensors because they provide a good balance between compliance and manufacturing. Although PLA-based sensors offer consistent, linear outputs and are useful for quick prototyping, their rigidity limits their use in highly deformable actuators. Silicone elastomers support extreme strains but introduce nonlinearities that require calibration or compensation. When developing strain sensors for soft robotic grippers, these comparisons highlight the need to strike a balance between sensitivity, durability, linearity, and simple manufacture Figure 4.

Sensitivity, linearity, hysteresis, and durability are critical factors in evaluating resistive strain sensors for soft robotic grippers because they determine accuracy and dependability in repeated grasping tasks. Sensitivity, linearity, hysteresis, and durability are critical factors in evaluating resistive strain sensors for soft robotic grippers because they determine accuracy and dependability in repeated grasping tasks.

• Sensitivity (Gauge Factor, GF): (GF stands for gauge factor, which is the ratio of applied strain $\epsilon$ to relative resistance change $\Delta R/R_0$) The gauge factor is commonly expressed as

  $$GF={\displaystyle \frac{\Delta R/R_0}{\epsilon }},$$

  (1)

  representing the capacity to detect minute deformations. Due to conductive network rearrangements, CNT-based sensors exhibit maximum sensitivity, with resistance variations of 1000% under strains of only 10–11% [137,145]. On the other hand, sensors built on PLA and TPU have moderate sensitivity but offer consistent performance across a wider range of strains [103,144,146].
• Linearity: Conductive PLA sensors demonstrate excellent linearity across bending motions [144], and TPU-based sensors maintain predictable linear responses up to 30% strain [103,146]. By comparison, CNT and silicone-based systems often exhibit nonlinear behavior, especially at higher strains [137,145,147].
• Hysteresis: Adjusting the conducting network and elastic relaxation in the polymer matrix are common ways to introduce hysteresis. While TPU-based sensors often exhibit reduced hysteresis, ensuring repeatability across cycles [103,146], CNT composites and silicone elastomers generally show higher hysteresis [145,147].
• Durability: TPU-based printed sensors demonstrate strong cycle stability and tolerate repeated bending without significant drift [103,146]. Although CNT sensors are extremely sensitive, they might fail if stretchy interconnects are not used to support them [145]. Long-term operation is demonstrated by pneumatic silicone gauges, although extended usage may cause signal drift [147].

These findings, compiled in Table 1, show different trade-offs: silicone elastomers reach high strains at the cost of nonlinearity, TPU achieves a balance between compliance and manufacturability, PLA produces outputs that are predictable and linear, while CNT composites give exceptional sensitivity.

3.5. Integration with Soft Robotic Grippers

For strain sensors to be functionally useful, they must be effectively integrated into soft grippers in a way that maintains compliance while enabling accurate deformation tracking. Several approaches have been explored, as illustrated in Figure 5. To continuously monitor actuator expansion and bending during grabbing, a particular approach is to directly attach strain sensors to the surface of soft pneumatic grippers [103]. Using this method, the sensor is guaranteed to fit closely against the gripper body and record changes in local strain without limiting movement (Figure 5a). Like this, soft robotic fingers with CNT-based strain sensors integrated into them offer real-time bending feedback, enhancing motion control and permitting the manipulation of fragile objects (Figure 5b) [145].

Actuators and strain sensors are combined into a single structure in a more integrated design. For example, sensorized actuators, which maintain structural continuity while providing direct input during deformation, are created by combining printed bending actuators with welded strain sensors (Figure 5c) [103,146]. To accurately estimate the opening and closing states during object handling, strain sensors can be positioned differently across the gripper body to capture various deformation modes (Figure 5d).

Soft pneumatic strain gauges can be used as force and curvature sensors in pneumatic actuators, in addition to electrical sensing. By enabling the decoupling of actuator pressure and bending angle, these designs expand the capabilities of multimodal sensing (Figure 5e) [147]. Lastly, resistive sensing can be immediately included in gripper designs using completely sensor-integrated soft pneumatic actuators (SPAs), which offer consistent feedback for gripping a variety of objects (Figure 5f). Sensor placement strongly influences signal accuracy. Surface-mounted sensors often provide higher sensitivity but may experience slippage or localized strain peaks, whereas embedding sensors within the actuator wall enhances mechanical coupling and improves signal stability with negligible impact on actuation performance.

3.6. Applications, Challenges, and Future Directions

The integration of resistive strain sensors transforms soft robotic grippers into adaptive manipulators capable of intelligent, safe interactions [51]. As an illustration of efficient, practical handling of delicate goods, soft grippers made using multimaterial FDM with integrated TPU strain-sensing elements have been utilized to package delicate fruits, such as strawberries and clementines [11,148]. Biocompatible silicone-based fluidic sensors in a different field show promise for wearable and human-assistive robotic applications due to their good linearity and low hysteresis (5%) [149].

Even with these developments, there are still several challenges. Closed-loop control [150] is made more difficult by the nonlinear nature and signal drift that many resistive sensors still display [151]. Pneumatic sensors based on silicone demonstrate particularly noticeable nonlinearities, whereas CNT-based systems provide great sensitivity but have issues with durability and fabrication complexity [152]. Cutting-edge designs such as 3D crack-enabled conductive elastomer networks can enhance signal fidelity and dramatically reduce hysteresis (2.9%), providing a route toward more dependable deformation detection in soft robotic systems. Furthermore, complicated sensor-morphology interactions present intrinsic hurdles for soft robots, necessitating advanced mapping, calibration, and data processing to appropriately interpret deformation, which further affects real-time use [64]. Possible potential fields include multimodal sensors (e.g., integrating resistive with optical or capacitive senses to offer richer proprioceptive data) and hybrid material systems that mix silicone or TPU with CNT networks for both compliance and sensitivity. Tighter sensor–actuator integration may be possible with improvements in additive manufacturing processes. Soft grippers may be able to adaptively adjust their grasp strategies by incorporating machine learning techniques into sensor inputs, which would enable more reliable operation in unstructured conditions.

4. Tactile Sensor

4.1. Transduction Mechanisms of Tactile Sensing in Soft Robotic Grippers

Tactile sensing in soft grippers is achieved via mechanisms that convert mechanical contact, such as normal force, shear, or contact area, into electrical or pneumatic signals compatible with compliant structures [64]. One method involves sandwiching flexible electrodes between piezoresistive films, like Velostat, where pressure reduces local resistance and makes it possible to measure force distribution over a sensor matrix [153]. Another technique uses pneumatic acoustic impact, which allows for contact detection at the fingertip without the need for electronics. A flexible internal chamber or tube acts as a resonator, and touch-induced deformation changes its acoustic resonance frequency [154,155]. Liquid-metal microchannels, usually made of EGaIn, are used in advanced designs. These sensors are soft, conformable, and may be shaped using microfluidics or lithography. They react to indentation by changing resistance through modified channel geometry [156]. Distributed pneumatic taxels, soft, 3D-printed air chambers whose interior pressure varies upon contact, are used in another method to provide a tactile response that is nearly linear throughout a surface [157]. Finally, printed multimodal skins, such as graphene-based or nanocomposite layers sprayed or printed onto elastomers, can sense temperature, humidity, and pressure (through resistive or capacitive change) all at once, allowing for multipurpose tactile feedback [7].

4.2. Materials, Fabrication, and Structural Considerations of Tactile Sensors

The tactile sensors summarized in

Table 3. Comparison of representative tactile sensor designs integrated with soft robotic grippers.

Material/Composite	Structure and Fabrication	Range/Output	Sensitivity	Size/Geometry	Response Time	Durability	Cross-Sensitivity	Ref.
Velostat + FPC	Matrix-shaped Velostat between two flexible printed circuits	Resistance up to 50 k$\Omega$	–	5 × 10 array (∼50 sensing points), ∼1 mm thick	–	Moderate; requires calibration	Moderate temperature sensitivity	[159]
Silicone (Dragon Skin™ 30)	Pneumatic resonant tube with lateral opening	Frequency-based measurement via resonance	–	Ø 6 mm × 10 cm	Tens of ms (reported qualitatively)	High, robust	Low mechanical cross-sensitivity	[154]
Silicone Elastomers + EGaIn	Microfluidic channels with an indenter layer	Resistivity of EGaIn: $29.4\times {10}^{-8}$ $\Omega$ · m	Higher with narrow indenters (2 mm best)	Length 6 mm; Widths 2–6 mm	<100 ms	Moderate; risk of leakage	Low mechanical cross-sensitivity	[161]
Silicone Rubber (Shore A40)	3D-printed pneumatic chambers	Pressure sensing	2.31 kPa/N	Raised 4 mm; Thickness 1.5 mm	–	High cycle stability	Low cross-sensitivity	[162]
GNPs/MWCNT/PEO + rGO/GO	Printed multimodal sensors on PDMS (pressure, temperature, humidity)	Pressure: 0–100 kPa; Temp: 20–65 °C; Humidity: 25–90% RH	Pressure: 1.1 MPa−1; Temp: 0.65% °C−1; Humidity: $\Delta C$ 3.5–5.8 pF	2 × 2 array (pressure); Thin films (temperature/humidity)	50–60 ms	≥100 cycles	Moderate cross-talk between modalities	[163]
Silicone + EGaIn Microchannels	Spiral microchannel (tactile) + linear channels (curvature)	Tactile: up to ∼0.5 mV; Curvature: 0.6–2.4 mV	High	Tip diameter 9 mm	<50 ms	Moderate; stable but fatigue sensitive	Low mechanical cross-sensitivity	[164]

employ a wide range of materials and structural concepts, each chosen to balance compliance, sensitivity, and simplicity of integration with soft robotic grippers. Velostat, a carbon-loaded polymer film, is often combined with a flexible printed circuit (FPC) layer [158] to form matrix arrays, as demonstrated by Dong et al. [159]. Its flexibility, low cost, and capacity to generate distributed taxels for contact localization are its primary benefits. However, only basic coating can be used for fabrication, and the electrical output that results is nonlinear and drift-sensitive, requiring calibration. Another option is provided by silicone elastomers, like Dragon SkinTM 30, which have tissue-like mechanical qualities and work well with pneumatic structures. In Li et al. [154], contact detection via changes in acoustic frequency was made possible by molding a resonant tube inside a silicone finger [160], demonstrating how straightforward cavity structures can offer reliable tactile cues with little need for embedded electronics.

More complex tactile sensors use soft microfluidic architectures filled with liquid metal, such as EGaIn. According to Deng et al. [161], sensitivity and flexibility are greatly increased when EGaIn channels [165] are embedded in elastomers together with indenter layers that focus stress into small geometries. Excellent linearity and force resolution are provided by this design, but weaknesses and possible leaks offer long-term reliability issues, and fabrication is quite complex. On the other hand, tactile features can now be directly printed into soft robotic hands through additive manufacturing. Shorthose et al. [162] fabricated distributed pneumatic chambers [166] in a 3D printed silicone hand, achieving linear pressure responses with good durability and allowing designers to modify sensing zones rapidly. Moreover, multimodal techniques incorporate nanocomposites like carbon nanotubes, graphene nanoplatelets, and rGO/GO into printed films on PDMS. Wu et al. [163] showed that electro-spraying and direct ink writing could create thin, flexible skins that could sense temperature, humidity, and pressure all at once, with stable performance over 100 cycles. By incorporating linear channels along the actuator body and spiral channels at the fingertip, Hao et al. [164] expanded liquid-metal microchannel designs, allowing for both tactile and curvature sensing in a small package.

Taken together, these strategies highlight the trade-offs between simplicity, sensitivity, and scalability. Printed nanocomposite skins indicate future multifunctional tactile systems; liquid-metal channels achieve high sensitivity but raise reliability concerns; 3D printing prefers durability and customization at reduced resolution; and Velostat and pneumatic designs are robust and affordable but have limited precision. Therefore, the practicality of implementing tactile feedback in soft robotic grippers is directly influenced by the material and fabrication choices, as well as the sensing characteristics (

Table 4. Comparison of tactile sensors: merits, demerits, and cost.

Tactile Sensor	Merits	Demerits	Cost
Velostat-based resistive sensor arrays	Low-cost, simple, flexible, suitable for classification	Nonlinearity, drift, limited durability	Low
Resonant pneumatic sensor	Robust, mechanically protected, stable response	Bulky, low spatial resolution	Medium
Programmable soft gripper	Adaptive, programmable length, versatile	Structural complexity, slower response	Medium–High
Direct-writing nanocomposite e-skin	Conformal, multimodal, scalable, fast response	Mechanical stability issues, durability concerns	Medium
3D-printed hand with distributed sensor	Integrated, scalable, customizable	Moderate sensitivity, design-dependent	Medium
Microfluidic liquid metal	High sensitivity, flexibility, conforms to complex geometry	Leakage risk, fatigue, complex fabrication	Medium–High

).

4.3. Performance Metrics and Sensor Characteristics

We evaluate tactile sensors for soft grippers by sensitivity, response/recovery time, linearity, spatial resolution, durability, and cross-sensitivities (e.g., temperature/humidity). Sensitivities, response times, and qualitative linearity for each paper or design presented are among the key values for the architecture reviewed here that are compiled in Table 3.

Different performance profiles appear across designs (for fabrication and integration contexts that significantly impact these metrics, see Figure 6a–f). Velostat + FPC matrices provide distributed sensing but require calibration due to their nonlinear output and drift (Figure 6a). Pneumatic resonant tubes embedded in silicone fingers deliver coarse spatial detail with stable, nearly linear frequency shifts under touch, exhibiting tens-of-millisecond dynamics (Figure 6b). Narrow indenter layers in microfluidic sensors allow for high, adjustable sensitivity while maintaining compliance; however, compromises include fabrication complexity and potential leakage or fatigue (Figure 6c).

Although spatial detail depends on chamber tiling, 3D-printed pneumatic chambers offer strong robustness, easy layout scaling, and good linearity with a reported sensitivity of 2.31 kPa N⁻¹ (Figure 6d). Printed multimodal skins on PDMS combine pressure, temperature, and humidity sensing, demonstrating e-skin-like functionality with a fast response (≈50–60 ms) and ≥100 stable cycles (Figure 6e). Dual-mode feedback with typically linear behavior and millivolt-scale outputs is achieved by designs that co-embed spiral (tactile) and linear (curvature) liquid-metal channels (Figure 6f).

Consequently, Velostat or pneumatic schemes are appealing for low-cost contact detection; EGaIn microchannels and printed nanocomposites provide faster dynamics and higher sensitivity for precise force estimation or classification; and 3D-printed chambers are suitable for reliable, scalable coverage. The selection process should consider calibration effort and long-term stability in the intended integration, while matching the application’s requirements for resolution, bandwidth, robustness, and multimodal sensing (as shown in Figure 6 and summarized in Table 3). A normalized comparison of these tactile sensor types, summarizing the trade-offs between sensitivity, response time, durability, and linearity, is illustrated in Figure 7.

4.4. Integration with Soft Robotic Grippers

The practical value of tactile sensors comes from how effectively they can be integrated into soft robotic grippers without compromising compliance, adaptability, or robustness. Sensors are usually integrated into internal spaces, attached as flexible skins, or embedded in the finger body, as shown in Figure 5. The mechanical structure of the gripper, the kind of tactile information needed (such as contact force, slip, and shape), and the durability requirements of repeated actuation are all closely related to the integration strategy selection.

In order to provide distributed contact detection, Velostat arrays are usually laminated onto the inside gripping surfaces (Figure 8a [159]). Although this method provides coverage at a low cost, it presents wiring difficulties and may marginally decrease surface compliance. Although spatial resolution is constrained, mechanically protected and robust sensing with near-linear frequency shifts is achieved through the embedding of pneumatic resonant tubes within elastomeric finger bodies (Figure 8b [154]). Cast into PDMS or Ecoflex substrates, liquid-metal microchannels (EGaIn) allow for seamless embedding that maintains compliance and offers high sensitivity; however, they may experience fatigue or long-term loss (Figure 8c [161]). Direct fabrication of 3D-printed pneumatic chambers as structural components inside the finger allows for long-lasting and repeatable sensing performance while preserving design flexibility (Figure 8d [162]). Although their long-term stability depends on the quality of the printed interface, printed nanocomposite e-skins conform to the surface of the finger and offer high sensitivity and multifunctionality (pressure, temperature, and humidity) with quick response times (Figure 8e [163]). Lastly, multimodal closed-loop manipulation tasks are supported by dual-mode spiral and linear liquid-metal channels, which enable simultaneous detection of tactile load and finger curvature, albeit with moderate endurance under repeated bending (Figure 8f [164]). All things considered, integration strategies illustrate a balance between sensing resolution, wiring complexity, durability, and mechanical compliance. While surface-mounted techniques like Velostat arrays and printed skins enable modular replacement and multifunctionality, embedded techniques like liquid-metal microchannels and pneumatic chambers provide superior compliance and robustness. Co-designing the gripper and sensor architecture is necessary for optimal integration in order to provide long-term functional performance, as shown in Figure 8.

A comparison of different integration methods, including their advantages, limitations, and related references, is summarized in

Table 5. Comparison of integration methods: advantages, limitations, and references.

Integration Method	Advantages	Limitations	Reference
Velostat lamination	Low-cost, easy to apply, distributed contact sensing	Reduces compliance, wiring complexity	[159]
Pneumatic resonant tube	Robust, mechanically protected, near-linear response	Low spatial resolution	[154]
EGaIn microchannels	Highly flexible, conformal, high sensitivity	Leakage, fatigue over cycles	[161]
3D-printed chambers	Directly integrated, robust, scalable	Design-dependent, moderate sensitivity	[162]
Printed nanocomposite e-skin	Multimodal sensing, fast response	Long-term stability issues	[163]

.

4.5. Applications and Emerging Directions of Tactile Sensors for Soft Robotic Manipulation

A vital aspect in enabling soft grippers to perform more agile and adaptable manipulation is tactile sensing [166]. Object recognition, force control, and handling delicate or deformable objects are some examples of applications. For instance, high-accuracy object classification has been achieved using Velostat-based sensor arrays [167]. However, resonant pneumatic sensors have shown resilience and sensitivity in soft environments when used to track grasp forces during food handling. Like this, embedded sensing has been used by programmable soft grippers with adjustable finger length to manipulate objects of different sizes adaptively. Further demonstrating how design–fabrication co-optimization can produce scalable integration strategies for real-time grasp feedback are 3D-printed soft hands with distributed tactile sensors.

Even with these developments, there are still many obstacles to overcome. Nonlinearity and calibration drift plague resistive materials like Velostat [168], and fatigue and leaking under repeated cycles plague liquid-metal microchannels [169]. In practical applications, direct-written e-skins still lack long-term mechanical stability, despite their potential for conformal integration [166]. Furthermore, the complexity of wiring and data processing brought about by high-resolution tactile sensing may restrict the portability and scalability of small grippers.

Therefore, further research should focus on: (i) low-cost, scalable fabrication techniques like additive manufacturing for large-area e-skins; (ii) multimodal sensing that integrates temperature, pressure, and slip detection; and (iii) closer integration with machine learning frameworks to allow adaptive object property classification and prediction from tactile signatures. Advancements in these areas will speed up the shift from experimental demonstrations to practical uses in service robotics, healthcare, and industrial automation for tactile sensing in soft grippers.

5. Ionic Sensor

Hydrogel and ionogel sensors are generally biocompatible, using ingredients such as poly(vinyl alcohol), alginate, and glycerol. Their low toxicity and moisture content make them suitable for food-handling and healthcare applications, although long-term cytocompatibility and sterilization studies remain limited.

5.1. Multimodal Sensing Principles of Ionic Sensor

Ionic sensors are unique in that they can incorporate several sensing modalities into a single soft gripper [30]. Ionic conductors can support both proprioceptive and exteroceptive feedback by acting as tactile detectors, inflation monitors, or curvature gauges, depending on their architecture and placement [147,170,171]. Ionic systems are different from other resistive sensors that usually only record one stimulus because of their multimodal ability.

The fundamental principle behind ionic sensing is the change in electrical properties of ion-conducting media, either ionogels or ionic hydrogels, when subjected to deformation [172]. Capacitive architectures use ionic hydrogel electrodes separated by a conducting elastomer, where pressure changes the dielectric thickness and increases capacitance [173]. In resistive designs, stretching or compression changes the conductor’s geometry, resulting in detectable resistance variations. These mechanisms enable surface contact, bending, and pneumatic actuation states to be monitored by a [174] single soft gripper.

Through embedded 3D printing [175], ionogel-based resistive sensors have been shown to exhibit nonlinear but extremely sensitive contact detection, inflation sensors responsive to internal pneumatic pressure, and near-linear resistance changes for curvature sensing [176]. Recent developments in ionic hydrogels, on the other hand, expand on this idea by fusing capacitive tactile sensors with resistive strain gauges to provide antifreezing and ambient-stable performance appropriate for challenging conditions [177]. When taken as a whole, these transfer techniques position ionic sensors as robust and adaptable instruments for facilitating multimodal somatosensory feedback in soft robotic grippers. A summary of the working principles, sensing functions, fabrication methods, and structural configurations of representative ionic sensors is presented in

Table 6. Principles and design of ionic sensors in soft robotic grippers, showing the working principle, sensing function, required characteristics, fabrication method, and structural configuration for different ionogel and hydrogel devices.

Principle	Sensing Function	Required Characteristics	Forming of Ionogel/Hydrogel	Device Structure	Ref.
Resistive sensing (ionogel curvature sensor)	–	High ionic conductivity, stretchability, linear $\Delta R/R_0$ response	Embedded 3D printing (EMB3D) of ionogel in elastomer channels	U-shaped traces inside bending actuator	[178]
Resistive sensing (inflation sensor)	Pressure sensitivity, repeatability, stable, resistance under deformation	High ionic conductivity, stretchability, linear $\Delta R/R_0$ response	EMB3D of ionogel along actuator chamber	Ionogel channel integrated into pneumatic actuator wall
Resistive tactile sensing (ionogel contact sensor)	Measures fingertip compression (fine/deep touch) and slip	High compliance, surface sensitivity, nonlinear but discriminative $\Delta R$	EMB3D traces at fingertip regions	Ionogel pads at anterior fingertip	[179]
Resistive strain sensing (hydrogel)	Tracks bending/strain during grasping	Antifreezing, stretchability, strong adhesion	Double-network hydrogel (Alg-PAAm) with glycerol	U-shaped hydrogel bonded to EcoFlex elastomer	[180]
Capacitive tactile sensing (hydrogel)	Detects applied pressure through capacitance change	Stable dielectric properties, encapsulation for durability	Hydrogel electrodes with EcoFlex dielectric encapsulation	Multilayer stacked capacitive element in fingertip

.

5.2. Materials, Fabrication, and Structural Considerations

The connection between soft ionic conductors and elastomeric encapsulants is a key component of ionic sensors, and the selection of material and manufacturing process has a direct impact on the sensors’ durability, performance, and integration into robotic systems [181].

Materials

Hydrogels and ionogels are the two main types of materials. Ionic liquids (like EMIM-ES) are distributed throughout a silica matrix to form ionogels, like those created by Truby et al. [178,179]. They are appropriate for embedding into soft actuators because they combine mechanical compliance with high ionic conductivity. Double-network architectures and glycerol-based additives are used in hydrogels, like the Alg-PAAm networks described by Zhou et al. [180], to improve their stretchability, water retention, and antifreezing characteristics. Given that poor adhesion can result in delamination under cyclic strain, the hydrogel–elastomer interface is especially crucial.

Fabrication

One popular method for adding ionogels straight into elastomeric matrices is embedded 3D printing, or EMB3D. In order to enable complex sensor geometries (such as U-shaped curvature sensors, inflation sensors, and fingertip tactile pads), Truby et al. [179] showed how to precisely deposit ionogel inks into pre-cast silicone channels. This method ensures that electrical pathways stay compliant with actuator deformation by enabling direct multimaterial integration. Simultaneously, Zhou et al. [180] created hydrogel sensors by encapsulating them in EcoFlex layers after in situ polymerization within elastomeric molds. This approach strikes a balance between ionic mobility preservation over wide temperature ranges and mechanical robustness.

Structural considerations

The sensing function determines the device architecture. Pressure sensors are inserted around pneumatic chambers to measure internal pressure, whereas curvature sensors usually use long tracks aligned along the bending axis. Since mechanical deformation is concentrated at the fingertip, tactile sensors frequently employ localized pads there. By varying electrode thickness and dielectric spacing, multilayer stacking (hydrogel electrodes + EcoFlex dielectric) in capacitive hydrogel sensors offers tunable sensitivity. Structural compliance and durability under repeated actuation continue to be major challenges in all designs, especially when it comes to preserving stable interfaces and preventing ionic liquid leakage. These material and fabrication strategies for ionic sensors are summarized in Table 7, which compares different ionogel and hydrogel sensor types, their compositions, fabrication methods, and structural configurations for soft robotic applications.

Table 7. Summary of ionogel and hydrogel-based sensors for soft robotic applications.

Sensor Type	Material Composition	Fabrication Method	Structural Configuration	Ref.
Curvature Sensor (Ionogel)	Ionogel (EMIM-ES + silica) in silicone elastomer	Embedded 3D printing (EMB3D)	U-shaped printed trace along bending axis	[178]
Inflation Sensor (Ionogel)	Ionogel matrix (same as above)	EMB3D deposition in pneumatic chamber wall	Circumferential trace around actuator chamber
Tactile Sensor (Ionogel)	Ionogel fingertip pads	EMB3D deposition at fingertip regions	Localized fingertip pads integrated in fingertip	[179]
Strain Sensor (Hydrogel)	Double-network Alg-PAAm hydrogel + glycerol additive	In situ polymerization + EcoFlex encapsulation	U-shaped bonded hydrogel trace
Capacitive Tactile Sensor (Hydrogel)	Hydrogel electrodes + EcoFlex dielectric layer	Multilayer assembly	Stacked capacitive sandwich structure	[180]

Table 7. Summary of ionogel and hydrogel-based sensors for soft robotic applications.

Sensor Type	Material Composition	Fabrication Method	Structural Configuration	Ref.
Curvature Sensor (Ionogel)	Ionogel (EMIM-ES + silica) in silicone elastomer	Embedded 3D printing (EMB3D)	U-shaped printed trace along bending axis	[178]
Inflation Sensor (Ionogel)	Ionogel matrix (same as above)	EMB3D deposition in pneumatic chamber wall	Circumferential trace around actuator chamber
Tactile Sensor (Ionogel)	Ionogel fingertip pads	EMB3D deposition at fingertip regions	Localized fingertip pads integrated in fingertip	[179]
Strain Sensor (Hydrogel)	Double-network Alg-PAAm hydrogel + glycerol additive	In situ polymerization + EcoFlex encapsulation	U-shaped bonded hydrogel trace
Capacitive Tactile Sensor (Hydrogel)	Hydrogel electrodes + EcoFlex dielectric layer	Multilayer assembly	Stacked capacitive sandwich structure	[180]

5.3. Performance Metrics and Sensor Characteristics

Ionic sensors’ sensitivity, linearity, hysteresis, durability, and environmental stability are commonly used to assess their performance [87]. The efficiency with which the sensors in soft robotic grippers convert mechanical stimuli into electrical signals is determined by these parameters [182]. Reported fingertip ionic sensors demonstrate high sensitivity to mechanical interactions. Specifically, they exhibit minimum detectable slip velocities of approximately 3–5 mm/s and force resolutions in the range of 10–100 mN, depending on the gel composition and electrode design [179].

Curvature and inflation sensing

Ionogel-based curvature and inflation sensors integrated into pneumatic actuators were demonstrated by Truby et al. [179]. They demonstrated dependable pressure detection over tens of kPa and linear resistance changes with bending angles up to approximately 90°. Because of their low hysteresis during cyclic bending, the sensors can be used to continuously monitor actuator deformation.

Tactile sensing

Truby et al. [178] reported fingertip-integrated ionogel pads that can distinguish between slip, deep compression, and fine touch. Although the tactile sensors’ resistance response was nonlinear, this nonlinearity helped them distinguish between different surface textures and force levels. The system demonstrated the potential of tactile sensing for intelligent manipulation tasks by enabling object recognition through machine learning.

Hydrogel strain and capacitive sensors

Hydrogel-based strain sensors with stretchable and antifreezing qualities were created by Zhou et al. [180] and maintained a steady electrical response at −20 °C. Strong adherence of the double-network hydrogel to elastomer substrates permitted repeated strain cycles without delamination. Additionally, linear capacitance–pressure relationships were offered by capacitive hydrogel tactile sensors, which demonstrated consistent performance over a broad load range.

General trends

Ionic sensors provide a unique edge in mechanical compliance and multifunctionality across all designs, allowing for simultaneous tactile, pressure, and curvature detection in a single device. Long-term stability, especially for hydrogel-based systems that are subject to dehydration, and achieving high signal-to-noise ratios in practical manipulation are still difficulties, in any case. A schematic overview of the structural configurations, fabrication strategies, and comparative performance of these ionic sensors is illustrated in Figure 9.

Figure 9. (a–c) Schematic illustration of ionic sensor structures and fabrication strategies, and a radar plot comparing ionic sensor types across technical performance metrics.

Figure 9. (a–c) Schematic illustration of ionic sensor structures and fabrication strategies, and a radar plot comparing ionic sensor types across technical performance metrics.

5.4. Integration with Soft Robotic Grippers

The ability of ionic sensors to smoothly integrate into compliant actuator systems without affecting mechanical flexibility determines their practical usefulness in soft robotics. Soft grippers can obtain multimodal feedback during interaction thanks to ionogel and hydrogel-based sensors, which, in contrast to rigid sensing modules, can be embedded, surface-mounted, or encapsulated within elastomeric structures. Soft grippers have effectively incorporated ionic sensors using a variety of techniques [178] and showed how to 3D print ionogel channels embedded in elastomeric fingers, allowing for curvature and inflation detection without compromising compliance [179]. To improve adaptive grasping, localized ionogel pads at fingertip regions are used to detect compression and slip events. More recently, ref. [180] presented hybrid hydrogel–elastomer systems that preserved softness while enhancing durability against mechanical fatigue and dehydration by encasing strain and capacitive sensors in EcoFlex.

Challenges and outlook

Despite these advances, integration challenges remain. Long-term stability is hindered by water loss in hydrogels and interfacial degradation under cyclic loading. Moreover, routing multiple sensing channels through compact grippers complicates electrical interconnect design. Future work should focus on robust encapsulation strategies, wireless data transmission, and modular architectures that balance sensing density with mechanical compliance.

5.5. Applications and Emerging Directions

Ionic sensors enable soft robotic grippers to perform multimodal sensing [183] by simultaneously monitoring bending, internal pressure, and contact forces, which are critical for dexterous object manipulation [33]. For instance, curvature and inflation sensors offer real-time feedback for force-regulated grasping, and integrated ionogel pads have been utilized in learning-based frameworks to categorize objects based on size and stiffness [184]. These capabilities are extended to subzero temperature environments by hydrogel-based tactile elements, which enable stable performance in settings that are unsuitable for traditional electronic sensors [185]. Figure 10 illustrates the integration of these ionic sensors into multi-finger soft robotic grippers, demonstrating how curvature, tactile, and strain sensing can be achieved simultaneously within compliant actuators. A comparative overview of different ionogel and hydrogel-based sensors used in soft robotic grippers is summarized in

Table 8. Comparison of ionogel and hydrogel-based sensors.

Sensor Type	Material/Structure	Electronic Range/Signal	Dynamic Profile	Actuation Coupling	Ref.
Curvature Sensor (Ionogel)	Ionogel (EMIM-ES + silica) printed as U-shaped traces inside the bending actuator	Resistance increases linearly with curvature ($\Delta R/R_0$ vs. angle)	Minimal hysteresis; stable under repeated cycles; good linearity	Bending angles up to ∼90° measurable with linear response	[178]
Inflation Sensor (Ionogel)	Ionogel embedded in pneumatic actuator chamber wall	Resistance correlates with internal pneumatic pressure	Slight viscoelastic hysteresis due to the elastomer matrix	Detects actuation pressure ranges (tens of kPa typical for soft grippers)
Tactile/Contact Sensor (Ionogel)	Ionogel traces embedded in the fingertip regions of soft fingers	Resistance change nonlinear vs. applied force	Sensitive to surface texture; distinguishes fine vs. deep touch	Force range up to ∼10 N; enables slip detection and texture discrimination	[179]
Resistive Strain Sensor (Hydrogel)	Double-network Alg-PAAm hydrogel with glycerol; bonded to EcoFlex elastomer	Large resistance variation with strain; maintains function at −20 °C	Linear $\Delta R/R_0$ vs. bending; antifreezing (−95 °C); stable under repeated cycles	Bending angle measurement (proprioception) during grasping
Capacitive Tactile Sensor (Hydrogel)	Hydrogel electrodes with EcoFlex dielectric; encapsulated multilayer structure	Capacitance increases linearly with applied pressure	Stable tactile feedback at ambient and subzero (−20 °C)	Pressure sensing across mN–N range; supports closed-loop force control	[180]

, highlighting their material composition, electrical characteristics, and actuation coupling performance.

There are still opportunities in the face of these advancements. Combining ionic sensors with machine learning algorithms offers adaptive manipulation strategies that change with experience, while embedding them into high-density arrays may allow texture recognition and detailed contact mapping. Furthermore, the current issues of energy dependence and wiring complexity would be resolved by the creation of wireless and self-powered ionic sensing systems. Finally, ionic sensors may be incorporated into modular gripper architectures in future designs to enable reconfigurable hands with task-specific sensing capabilities

Table 9. Benefits, limitations, and cost considerations of ionogel and hydrogel-based sensors.

Sensor Type	Benefits	Limitations	Cost Considerations
Curvature Sensor (Ionogel)	Linear response to bending; reliable proprioception; easy embedding in actuator	Requires EMB3D fabrication; limited scalability	Moderate—depends on 3D printing process and ionogel synthesis
Inflation Sensor (Ionogel)	Differentiates internal actuation from external load; stable pressure sensing	Viscoelastic hysteresis from elastomer matrix; less precise under high cycles	Moderate—similar cost to curvature sensors with added fabrication complexity
Tactile Sensor (Ionogel)	Sensitive to surface contact, slip, and texture; supports ML-based recognition	Nonlinear signal complicates calibration; fabrication at fingertip regions is complex	Higher—multiple embedded traces increase fabrication and assembly cost
Strain Sensor (Hydrogel)	High stretchability; antifreezing down to −95 °C; strong adhesion to elastomer	Risk of dehydration without additives; long-term stability issues	Low to moderate—hydrogels are inexpensive, but encapsulation adds cost
Capacitive Tactile Sensor (Hydrogel)	Linear capacitance–pressure response; robust tactile feedback even at subzero	Encapsulation and multilayer assembly increase fabrication complexity	Moderate to high—fabrication is costlier than resistive hydrogels

.

6. Conclusions

Soft robotic grippers represent one of the most promising directions in robotics, as they combine adaptability with the ability to handle delicate objects safely. For these systems to reach their full potential, sensing is not optional but essential. This review has explored resistive sensing as a central enabler, examining strain, tactile, and ionic sensors as the three main approaches currently shaping the field. Strain sensors offer reliable proprioceptive feedback by tracking deformation and actuation states, while tactile sensors enrich interactions by detecting contact, slip, and pressure distribution. More recently, ionic sensors have emerged as multifunctional devices that combine stretchability, transparency, and even self-healing properties, making them well-suited for next-generation applications. Together, these technologies illustrate how resistive sensing can transform soft grippers from passive mechanical tools into intelligent, responsive systems.

At the same time, important challenges remain. Nonlinearity, hysteresis, signal drift, and limited long-term stability continue to limit industrial adoption. Addressing these issues will require not only improvements in materials and fabrication but also more systematic benchmarking and integration with advanced signal processing methods.

Looking ahead, several opportunities stand out. Hybrid multimodal sensors that combine resistive elements with capacitive, optical, or piezoelectric mechanisms could provide richer feedback while reducing wiring complexity. Additive manufacturing and 3D printing will enable scalable, customizable co-design of sensors and actuators. Advances in encapsulation and self-healing materials may overcome durability concerns, while AI-driven data interpretation can unlock adaptive and intelligent control strategies.

In conclusion, resistive sensors remain a highly promising pathway for embedding intelligence into soft robotic grippers. By bridging material innovation, fabrication advances, and machine learning, they hold the potential to transform soft robotics from experimental prototypes into robust, widely deployed tools across healthcare, food handling, and human–robot collaboration.