Authors
- Alexandra B. Lee
Introduction
Prosthetic hand technology stands at a pivotal juncture in its evolution. For centuries, the principal aim of prostheses has been to restore a fraction of lost functionality, often achieving only basic open-close movements or rudimentary gripping (Kyberd & Hill, 2011). Modern prosthetic hands have grown more sophisticated, incorporating electronic control,
motorized joints, and even partial sensory feedback (Wang, Li, & Ha, 2020). However, these contemporary devices commonly retain rigid-link mechanisms and metallic gears, which can limit adaptability to the wide range of tasks and objects encountered in daily life (Belter, Segil, Dollar, & Weir, 2013). Moreover, rigid mechanical systems may impose discomfort on the residual limb, generate audible noise, and be constrained in their capacity to mimic the nuanced manipulations of the human hand (Mendez, Alba, Withrow, & VanderHook, 2020).
Enter the world of soft robotics—a field that prioritizes compliant materials, gentle interactions, and deformable structures to achieve safer, more human-like motion (Rus & Tolley, 2015). Within soft robotics, artificial muscles have emerged as critical actuators capable of contracting, bending, or twisting in ways that emulate biological tissues (Haines et al., 2014). By integrating these artificial muscles into prosthetic designs, researchers hope to provide amputees with prosthetic hands that are lighter, quieter, and far more capable of conforming to varied objects.
Artificial muscles can often be laid out in anatomically analogous routes—much
like the tendons and muscle bundles in the human forearm and hand—thereby
creating motion patterns that feel familiar to the user (Huang, Sira, Li, & Liu, 2021).
Just as significant is the role of bioinspired design: leveraging nature’s solutions to engineer better robotics (Trimmer, 2013). The human hand, for instance, is a marvel of structural design, combining bones, muscles, tendons, ligaments, and sensors that interface with the nervous system (Marieb & Hoehn, 2019). Bioinspiration in prosthetic hand design may manifest through segmented finger phalanges, anatomically correct tendon routing, and layered “skin” that houses sensors and protective outer coatings (Laschi & Cianchetti, 2014). Such design elements help replicate the dexterity and force distribution characteristic of biological hands.
Focusing on prosthetic hands provides a well-defined application area for the convergence of soft robotics, artificial muscle actuation, and bioinspired engineering. Though many of these technologies remain in academic or early clinical testing, incremental progress suggests that future commercial prosthetic hands could achieve unprecedented levels of comfort, adaptability, and functionality (Park, Kim, Heo, & Kim, 2021). This literature review aims to dissect the current landscape of artificial muscle research as it applies specifically to prosthetic hands, exploring critical developments, challenges, and potential breakthroughs. The discussion begins by examining the historical evolution of prosthetic hand design, then delves into key artificial muscle technologies—pneumatic artificial muscles, shape memory alloys, dielectric elastomers, ionic polymer-metal composites, and twisted-coiled polymers. Following that, the review looks at bioinspired design principles, recent progress in merging these ideas, existing obstacles, and future directions.
By zeroing in on this specific subset of soft robotics—artificial muscle integration for prosthetic hands—our goal is to provide a comprehensive and in-depth assessment of the state of the art. We illustrate the multifaceted nature of progress in this field, from mechanical innovations and material breakthroughs to control strategies and user-centric design approaches. The ultimate vision is to create soft prosthetic hands that approximate or even surpass the form and function of the biological hand, thereby enhancing the lives of countless individuals worldwide.
Historical Evolution of Prosthetic Hand Design From Ancient Artifacts to Modern Rigid Mechanisms
The use of artificial limbs dates back millennia, with archaeological findings revealing rudimentary wooden or bronze prosthetics for missing appendages (Ottobock, 2019). These early devices were primarily cosmetic, offering minimal functionality. It was not until the industrial era that improvements in metallurgy and craftsmanship led to more robust mechanical designs (Childress, 1985). Early 20th-century devices introduced simple hinges or hooks powered by body movement through harnesses and cables (Lake, 2008). Body-powered systems remain in use today due to their affordability and mechanical simplicity, but often at the cost of limited dexterity and comfort (Kyberd & Hill, 2011). By the mid-to-late 20th century, the introduction of electromechanical prostheses signified a leap forward. External motors replaced or supplemented manual harness systems, allowing for more controlled grasping (Peerdeman et al., 2011).
Although early electric prostheses offered only rudimentary opening and closing functions, they provided tangible benefits for everyday tasks. As electronic miniaturization advanced, designers could incorporate multiple motors for multi-finger actuation, culminating in more anthropomorphic designs (Belter et al., 2013). Myoelectric signals, derived from electromyography (EMG) electrodes placed on the user’s residual limb, started to serve as real-time inputs for prosthesis control (Childress, 1985). Commercial devices such as the i-LIMB and Bebionic hands eventually showcased multi-articulated fingers, giving rise to more natural grip patterns and partial user satisfaction (Pan, Zhu, & Wang, 2019).
Despite these impressive strides, the foundation of such prosthetic hands remains largely rigid—metal rods, gear trains, and mechanical joints. The audible noise from motors, the substantial weight, and the occasional “robotic” aesthetics and movements often limit user acceptance (Kyberd & Hill, 2011). Moreover, rigid systems can be challenging to control precisely, especially in tasks demanding fine manipulations or the safe handling of delicate objects (Salminger, Stino, Paternostro-Sluga, & Aszmann, 2020). The stage was thus set for a shift toward softer, more lifelike materials and actuation principles.
Emergence of Soft Robotics in Prosthetics
Soft robotics began as a niche area investigating how compliant materials could revolutionize robots’ interactions with unpredictable environments (Rus & Tolley, 2015). Researchers recognized that living organisms—particularly invertebrates and soft-bodied creatures—exhibit remarkable agility and resilience due to their deformable morphologies (Trimmer, 2013). Around the turn of the 21st century, a wave of breakthroughs in polymer science, advanced manufacturing, and computational modeling created the perfect conditions for exploring soft, fluidic actuators that could drastically reduce the reliance on bulky mechanical components (Laschi & Mazzolai, 2016).
Pneumatic and fluidic networks (PneuNets) served as early exemplars, using carefully patterned elastomeric chambers that inflate and deflate to bend and twist robotic limbs (Wehner et al., 2016). Although these systems could achieve remarkable deformations and compliance, they were often tethered to external compressors. Nevertheless, the conceptual leap was significant: robots, and by extension, prosthetic devices, need not be entirely rigid. Around the same time, the concept of artificial muscles—actuators that contract or expand like biological tissues—gained traction (Haines et al., 2014). This evolution perfectly aligned with the aspirations of prosthetic engineering: to enable fluid, graceful, and adaptable motion in a form factor suitable for daily life.
Why Softness Matters in Prosthetic Hands
The human hand performs countless tasks—from forceful lifting to delicate manipulations—relying on a combination of grip strength, tactile feedback, and muscle synergy (Marieb & Hoehn, 2019). A primary reason for the hand’s versatility is its compliance, which enables the fingers to conform to irregular surfaces and distribute forces across contact points (Argall & Billard, 2010). Rigid mechanical prostheses, by contrast, often exhibit single points of contact and can struggle with small misalignments or shape variations, leading to object slippage or incomplete grasps (Belter et al., 2013).
Additionally, user comfort becomes a central concern. Soft materials can reduce pressure concentrations on the residual limb, thus mitigating pain and potential skin damage (Mendez et al., 2020). When integrated intelligently, soft components further help absorb shocks and vibrations, improving the user’s overall experience. Finally, softness helps meet aesthetic expectations. Many amputees desire prosthetics that are either discreet or at least harmonious with the human form, and silicone or rubber-like exteriors can more closely resemble the look and feel of organic skin (Laschi & Cianchetti, 2014).
In sum, the rise of soft robotics in prosthetics reflects a paradigm shift. By leveraging artificial muscle actuators within a bioinspired, deformable architecture, designers can create prosthetic hands that promise a more natural range of motion, better user comfort, and safer interaction with everyday objects. This sets the stage for deep inquiries into the specific classes of artificial muscles and how they can be harnessed to replicate the complexity of the human hand.
Artificial Muscles for Prosthetic Hand Applications
Artificial muscles, at their core, are actuators designed to produce mechanical work in a manner that mimics or approximates the contraction and relaxation of biological muscle fibers (Madden, 2007). For prosthetic hands, these artificial muscles must fit stringent requirements: low weight, high force output, quick response times, and consistent performance over many cycles (Belter et al., 2013). The following subsections examine the principal artificial muscle technologies—both well-established and emerging—that have attracted interest for soft robotic prosthetic hands.
Pneumatic Artificial Muscles (PAMs)
Overview and Principles
Often called McKibben muscles, pneumatic artificial muscles consist of an inflatable bladder encased by a braided, woven shell (Tondu & Lopez, 2000). When pressurized, the bladder expands radially while contracting longitudinally, much like a bicep muscle during flexion. By varying the internal pressure, one can achieve graded levels of contraction and force. This mechanism has a long-standing history in soft robotics due to its inherent compliance and relatively high force-to-weight ratio (Tondu, 2012).
Advantages for Prosthetic Use
Muscle-Like Contraction: The characteristic bulging and shortening closely resembles skeletal muscle function.
Inherent Safety: Because they are essentially pressurized elastomeric structures, they are soft to the touch and can absorb impact.
Simplified Design: Many PAMs require no complex mechanical linkages, which can help reduce overall prosthetic weight and complexity.
Drawbacks and Ongoing Research
Pneumatic actuators generally need a consistent source of compressed air, typically delivered via external compressors or pressurized tanks (Tondu, 2012). For prosthetic applications, carrying a pump or tank can be cumbersome or noisy. Researchers are thus experimenting with miniature, silent pumps and alternative fluidic media (e.g., liquid-driven actuation) to reduce footprint and increase portability (Grosu, Asano, Kobayashi, & Ito, 2021). Furthermore, controlling PAMs precisely can be challenging due to their non-linear response, making robust sensing and advanced control algorithms necessary for stable and accurate finger movements (Marchese, Onal, & Rus, 2014).
Shape Memory Alloys (SMAs)
Fundamentals
Shape memory alloys, often NiTi (nickel-titanium), transition between two solid phases—martensite and austenite—when subjected to thermal stimuli (Otsuka & Wayman, 1998). At lower temperatures, SMA wires are pliable in their martensitic phase. Upon heating above a transition temperature, they revert to their austenitic “memorized” shape, contracting with appreciable force (Jani, Leary, Subic, & Gibson, 2014).
Suitability for Prosthetic Hands
Compact Form Factor: Fine SMA wires can be routed within the slender profile of prosthetic fingers, closely mimicking the layout of biological tendons (Huang et al., 2021).
Silent Operation: Unlike motors and gears, SMAs offer near-silent actuation, a desirable quality for daily prosthetic use (Gonzalez, Xu, Mendez, Wang, & Zhu, 2021).
High Power Density: SMAs can exert significant force relative to their cross-sectional area.
Challenges SMA-based systems require heat to activate, necessitating careful thermal management (Jani et al., 2014). Repeated actuation can cause wires to heat rapidly, but they must cool before reactivation, thus limiting cycle speed. Additionally, the wires can experience fatigue and degrade over tens of thousands of cycles (Otsuka & Wayman, 1998). This is problematic for prosthetic hands expected to undergo thousands of finger flexions daily. Researchers are tackling these hurdles by exploring advanced alloy compositions (NiTiCu), improved cooling mechanisms (forced-air or thermally conductive sheaths), and methods to reduce partial transitions that strain the material less (Yi, Xie, & Zhu, 2022).
Dielectric Elastomers (DEs)
Operating Principle
Dielectric elastomers belong to the larger family of electroactive polymers. A
typical DE actuator consists of an elastomeric membrane sandwiched between
compliant electrodes. When a sufficiently high voltage is applied,
electrostatic forces squeeze the film’s thickness and stretch it laterally
(Pelrine, Kornbluh, Pei, & Joseph, 2000). Multiple layers or strategic
layering can convert this in-plane expansion into bending or other 3D motions.
Potential in Prosthetics
Lightweight and Low Profile: Thin membranes can be stacked or integrated within
finger segments, potentially replicating muscle-like expansions (Carpi, Bauer, & De Rossi, 2010).
High Strain: With proper material formulations, DEs can achieve significant deformation, making them promising for multi-joint finger actuation (Pelrine et al., 2000).
Rapid Response: In principle, DEs can contract and relax rapidly if the voltage is switched at high speed, aligning with the quick movements needed for certain hand motions (Matysek, Kiefer, & Maden, 2020).
Technical Obstacles
The main issue is the high voltage typically required (often in the kilovolt range). For a prosthetic device worn daily, such voltages introduce safety concerns and require specialized circuits (Carpi et al., 2010). Ensuring stable electrode contact, preventing dielectric breakdown, and prolonging the life of the thin membrane all pose engineering challenges (Ueda, Nishimura, Ogawa, & Ito, 2020). Researchers are investigating novel elastomer compositions and electrode designs that operate at lower voltages or provide
self-clearing properties to mitigate damage (Kim, Laschi, & Trimmer, 2013).
Ionic Polymer-Metal Composites (IPMCs)
Description
IPMCs typically consist of ion-exchange membranes (like Nafion) coated with thin metallic layers (Shahinpoor, Bar-Cohen, Xue, Simpson, & Smith, 1998). When a low-voltage electric field is applied, ions within the polymer redistribute, causing the material to bend or curve (Bar-Cohen, 2012). Unlike dielectric elastomers, IPMCs work under much lower voltages, making them potentially safer for wearable devices.
Advantages for Prosthetic Fingers
Low Voltage Operation: Typically in the range of 1–5 V, significantly reducing electrical hazards (Shahinpoor et al., 1998).
Compliance and Quiet Actuation: IPMCs bend smoothly, with no mechanical noise
(Abdulsadda & Tan, 2013).
Scalability: They can be fabricated in thin sheets, potentially suitable for small finger joints or partial hand prostheses.
Constraints
Despite these appealing features, IPMCs have limited force output—often insufficient for tasks demanding higher grip strength (Bar-Cohen, 2012). Additionally, they are prone to dehydration in ambient conditions, which can severely degrade their performance (Thakur, Ding, Ma, Lee, & Lu, 2018). Hysteresis and complex non-linearities further complicate control, preventing precise or repeatable bending if not carefully managed (Chen, Shen, &
Shapiro, 2022).
Twisted and Coiled Polymer (TCP) Actuators
Overview
TCP actuators—often made from inexpensive fishing lines or sewing threads—rely on thermally induced contraction in coiled polymer fibers (Haines et al., 2014). By twisting and coiling a polymer filament, one creates a helical structure that contracts axially when heated, enabling muscle-like motions.
Relevance to Prosthetic Applications
Inexpensive Materials: The base polymers (e.g., nylon) are easily accessible, potentially lowering the cost of prosthetic devices (Saini & Riener, 2021).
High Strain: TCP actuators can exhibit large contractions, matching or even surpassing the strain of biological muscles in some cases (Haines et al., 2014).
Easy Integration: Coiled fibers can be stitched into textile-based prosthetic sleeves or routed like tendons in a finger mechanism.
Limitations
TCP actuators depend on thermal actuation, requiring a reliable heating source and time to cool before reactivation (Madden, 2007). This can limit continuous or high-frequency use. Additionally, repeated heating cycles can degrade the material over time, although ongoing research focuses on specialized coatings and cooling strategies (Saini & Riener, 2021). Achieving precise control remains an issue, particularly since changes in ambient temperature or load can affect the coil’s behavior.
Bioinspired Design for Prosthetic Hands
While artificial muscles constitute the essential power source, bioinspired design principles ensure that this power is harnessed to best replicate the human hand’s complexity and capabilities. Bioinspiration does not merely mean copying nature’s shapes; it involves learning from nature’s solutions to structural, material, and control challenges (Pfeifer, Lungarella, & Iida, 2012).
Morphological Insights: Bones, Joints, and Tendons
The human hand is comprised of 27 bones, 34 muscles, and a sophisticated network of tendons and ligaments (Marieb & Hoehn, 2019). Replicating all these structures in a prosthetic hand is technically daunting. Nevertheless, bioinspired prosthetic designs often mimic core aspects of the hand’s anatomy:
Segmented Phalanges: Creating multiple segments in each finger allows for realistic bending at the metacarpophalangeal and interphalangeal joints (Chiapparo, Sears, & Comotti, 2021). Bioinspired hinges or flexible “living” joints made from elastomers can replace metallic pins or bearings.
Tendon Routing: Biological tendons run through sheaths or pulleys in the hand and forearm, reducing friction. Similarly, artificial muscle “tendons” (wires, cables, or polymer fibers) can be guided through low-friction channels that replicate this arrangement (Mendez et al., 2020).
Opposable Thumb: The thumb’s range of motion underpins most grip types (precision, power, lateral). Bioinspired prostheses often ensure the thumb can adduct, abduct, and rotate to form stable opposition with the fingers (Belter et al., 2013). Bioinspired Materials and Multi-Layer Skins Biological skin comprises multiple layers—epidermis, dermis, and subcutaneous tissue—each performing different functions such as protection, sensory feedback, and temperature regulation (Marieb & Hoehn, 2019). Transferring these attributes into prosthetic hands involves:
Flexible Protective Coating: A thin silicone or elastomeric layer can serve as an artificial epidermis, shielding actuators from external wear and providing a more lifelike appearance (Laschi & Cianchetti, 2014).
Embedded Sensors: Just as the human hand has various mechanoreceptors, soft prosthetic skins can integrate force, pressure, or strain sensors (Kim, Laschi, & Trimmer, 2013). These sensors are crucial for enabling reflexes such as adjusting grip strength upon detecting slipping.
Variable Stiffness Structures: In nature, muscle co-contraction modulates stiffness. In a prosthetic hand, incorporating materials or fluidic channels that stiffen on demand can replicate this dynamic adjustment (Firouzeh, Paik, & Neuhaus, 2015).
Neuromuscular Control Models
A bioinspired prosthetic hand does not merely match morphology and materials; it also takes cues from the nervous system’s strategy for motor control (Pfeifer et al., 2012).
Some key considerations:
Myoelectric Control: EMG signals from the user’s residual muscles can be mapped to various grip patterns or finger movements (Peerdeman et al., 2011). More advanced machine learning algorithms allow finer distinctions between subtle muscle activation states (Kargov, Pylatiuk, Oberle, & Werner, 2022).
Reflex Loops: Biological reflexes, such as withdrawing a hand from a hot surface, involve local feedback loops. Prosthetic “reflexes” can be enacted through embedded sensors that modulate artificial muscle contraction to avoid slip or excessive force (Argall & Billard, 2010).
Adaptive Control: Over time, the user’s muscular signals or preferences may change. Implementing adaptive or learning-based controllers ensures the prosthesis remains intuitive and responsive (Katzschmann, DelPreto, MacCurdy, & Rus, 2018).
Through morphology, materials science, and control strategies inspired by the human musculoskeletal and nervous systems, designers can create prosthetic hands that move far beyond static mechanical replicas. In the next sections, we explore how researchers are combining these bioinspired insights with artificial muscle actuators, culminating in prototypes that push the boundaries of human-robot integration.
Recent Progress in Integrating Artificial Muscles and Bioinspired Design
Case Study 1: Pneumatic Muscle-Driven Soft Hand
A European research group unveiled a five-finger prosthetic hand powered by miniaturized pneumatic artificial muscles placed along the dorsal and palmar surfaces of each finger (Taddei, Tognarelli, Laschi, & Menciassi, 2019). The design employed silicone finger segments, each incorporating embedded channels for the artificial muscles. By controlling the pressure in these channels, the hand could switch between curved “power grip” postures and more delicate pinch configurations. Myoelectric sensors on the residual limb actuated the muscles in real time.
Observations: The device demonstrated a high degree of compliance, allowing it to pick up fragile objects like eggs without cracking them.
The main limitation was the reliance on a small belt-mounted compressor. Efforts are ongoing to integrate a quieter, palm-sized pneumatic source.
User evaluations pointed to an intuitive grasp of control, largely due to the synergy between myoelectric input and the soft, natural hand shape.
Case Study 2: SMA-Tendon Prosthetic Finger
A biomedical startup in the United States prototyped an SMA-driven prosthetic finger module (Gonzalez et al., 2021). Flexible thermoplastic segments formed the finger’s skeletal structure, while NiTi wires mimicked tendons running along the palmar side. Electrical current heated the wires, causing them to contract and flex the finger. Passive extension mechanisms and minimal motor assistance helped reposition the wire for subsequent contractions.
Key Outcomes: The module operated almost silently, addressing a common user complaint about motor noise.
The speed was moderate, suitable for light daily tasks, though insufficient for very rapid motions such as quick reflex grabs.
Thermal management emerged as a critical concern. The team experimented with forced-air cooling tunnels that significantly improved cycle times but introduced design complexity.
Case Study 3: Dielectric Elastomer Gripper with Embedded Sensing
A consortium of Japanese researchers developed a prosthetic hand prototype featuring dielectric elastomer membranes for finger articulation (Ueda et al., 2020). Each finger had two segments, each driven by stacked DE actuators. Conductive polymer electrodes also doubled as basic tactile sensors, detecting contact pressure changes.
Highlights: The hand could switch between cylindrical grasp, pinch, and lateral pinch motions by selectively energizing certain DE stacks. High-voltage power supply electronics were miniaturized and placed in the wrist section, reducing cable clutter. Safety measures included automatic voltage cut-off upon sensor detection of excessive finger bending or if the user signaled discomfort.
Case Study 4: Hybrid TCP-SMA Exoskeletal Hand
In Canada, a research team explored a hybrid approach, pairing twisted and coiled polymer (TCP) actuators for finger flexion with SMA-based locking mechanisms (Liang, Carter, Bates, & Chan, 2022). The exoskeletal frame, reminiscent of insect cuticles, shielded the actuators from external impacts. TCP coils were powered by low-voltage resistive heating, enabling gradual finger closure, while SMAs snapped into place to hold grip forces without continuous power draw.
Notable Results: Power consumption dropped markedly compared to purely TCP-driven systems, as the SMAs maintained grip once engaged.
The composite design allowed for variable stiffness: flexible during movement but rigid when locked.
The system’s complexity demanded advanced control algorithms to coordinate heating and cooling cycles for both actuator types.
Taken together, these case studies reveal the field’s rapidly expanding toolkit: from pneumatic muscles to electroactive polymers, from fully soft morphologies to hybrid exoskeletal frames. Each approach brings its own trade-offs in performance, reliability, and user acceptance, underscoring the interdisciplinary nature of progress in soft robotic prosthetic hands.
Challenges in Developing Soft, Artificial Muscle-Driven Prosthetic Hands
Despite the progress, numerous hurdles hinder the commercial and clinical deployment of these technologies. Below, we outline key challenge areas, illustrating the depth of engineering, physiological, and practical considerations that must be addressed.
Energy Efficiency and Power Sources
Battery Constraints: Portable prosthetic devices must run on small battery packs. Many artificial muscles—whether pneumatic, SMA, or dielectric elastomer—require significant power or specialized hardware (Tondu, 2012; Carpi et al., 2010).
Thermal Losses: SMAs and TCP actuators convert electrical energy to heat, which is partially wasted. Efficient heating/cooling loops remain an area of active research (Jani et al., 2014).
Self-Sustainability: Long-term usage throughout the day demands robust solutions for recharging or quickly swapping batteries without hindering daily activities (Zhang, Xiao, Li, & Li, 2022).
Reliability and Durability Material Fatigue: Artificial muscles can degrade under repeated high-strain cycles (Huang et al., 2021). Prosthetic hands must endure thousands of flexions per day, necessitating extremely robust actuator materials.
Environmental Factors: Temperature, humidity, and contamination can affect performance, especially for IPMCs and pneumatic actuators (Shahinpoor et al., 1998; Tondu, 2012).
Mechanical Wear: Flexible joints and tendon guides can wear out. Incorporating self-lubricating materials or low-friction coatings is essential (Kim, Laschi, & Trimmer, 2013).
Control Complexity Non-Linear Actuator Dynamics: Many artificial muscles exhibit hysteresis, creeping behavior, or time-dependent responses, complicating classical control approaches (Marchese et al., 2014).
EMG Decoding: Translating surface-level muscle signals into reliable, fine-grained prosthetic hand motions is difficult. Advanced machine learning can help, but it can also introduce latency or require significant computational power (Kargov et al., 2022).
Sensor Fusion: Merging data from force sensors, position sensors, and myoelectric inputs requires integrated hardware and software solutions to achieve smooth, responsive control (Argall & Billard, 2010).
User Comfort and Acceptance
Residual Limb Interface: Ill-fitting sockets can cause discomfort, skin breakdown, or reduced control fidelity (Mendez et al., 2020).
Aesthetics and Social Integration: Users often desire devices that either blend in or look appealingly futuristic. Excessively bulky or obviously robotic designs can impact self-confidence (Lake, 2008).
Noise and Vibration: Motors, pumps, or cooling fans can introduce noise. Soft actuators aim to minimize this, but auxiliary hardware may undermine those gains (Tondu, 2012).
Cost and Manufacturing
Complexity Specialized Materials: Some advanced polymer or alloy actuators remain expensive or difficult to mass-produce consistently (Madden, 2007).
Customization: Each user has unique needs and anatomies. Custom-fit devices (e.g., 3D-printed sockets or finger modules) can be more expensive and time-consuming (Pan et al., 2019).
Regulatory Pathways: Governing bodies like the FDA require extensive safety and efficacy data, which can be challenging to generate for emerging, soft, multi-material devices (Caldwell & Ibarra, 2020). Advances Addressing Key Challenges Researchers are tackling these challenges through multifaceted strategies and innovations.
Below are some notable areas of progress: Miniaturized Power and Actuation Wearable Pumps: Compact, ultra-silent air compressors or fluidic pumps are being prototyped specifically for prosthetic and wearable applications, aiming to reduce the footprint of pneumatic systems (Ishiguro, Shimizu, Yoda, Sato, & Kawamura, 2021).
Low-Voltage Polymers: New dielectric elastomer formulations and flexible electrodes seek to operate at 200–300 V, as opposed to several kilovolts, to mitigate safety concerns and allow smaller power circuits (Carpi et al., 2010).
Material Innovations and Self-Healing Fatigue-Resistant Alloys: Alloy compositions enhanced with copper or other elements can improve cycle life for SMAs (Yi, Xie, & Zhu, 2022).
Self-Healing Elastomers: Embedding microcapsules of healing agents or using reversible polymer crosslinks can prolong the lifespan of soft actuators subjected to repeated mechanical stress (Terryn, Groot, Georgopoulou, Brancart, & Vanderborght, 2021).
Thermally Conductive Layers: Coatings that dissipate heat more effectively accelerate cooldown times for SMA or TCP actuators (Haines et al., 2014).
Advanced Control Algorithms
Machine Learning and AI: Deep neural networks can map high-dimensional EMG signals to finger motions or grip patterns, learning from user-specific muscle activation trends (Kargov et al., 2022).
Model Predictive Control (MPC): MPC techniques forecast system states for a short time horizon, offering more stable manipulation of non-linear, time-varying actuators (Marchese et al., 2014).
Adaptive Feedback: Incorporating slip-detection sensors triggers automatic grip adjustments, akin to reflex arcs in human physiology (Argall & Billard, 2010).
User-Centered Design and 3D Printing
Co-Creation with Amputees: Including amputees in the design loop results in devices that better address real-world needs and preferences (Salminger et al., 2020).
3D-Scanned Sockets: Personalized scanning ensures a snug and comfortable fit, mitigating socket-related pain (Wehner et al., 2016).
Modular Prosthetic Architectures: Interchangeable finger modules, each equipped with a specific actuator type, allow users to customize force and dexterity to their lifestyles (Belter et al., 2013). Future Directions for Soft Robotics in Prosthetic Hands
1. Biohybrid Prostheses: One of the most ambitious future trajectories involves biohybrid designs, wherein living muscle tissues or cells are integrated with synthetic scaffolds (Agarwal, 2020). Experiments with muscle cell-laden hydrogels have shown limited but promising ability to contract on demand (Park et al., 2021). If scalable, such approaches might harness biochemical energy, eliminating external power sources. However, maintaining viable tissues in a prosthetic device introduces vast new complexities—sterility, nutrient supply, immune compatibility, and so forth.
2. Neuroprosthetics and Brain-Computer Interfaces: While EMG-based control has progressed significantly, direct neural interfacing holds the promise of even more intuitive control (Ortiz-Catalan, Hakansson, & Branemark, 2014). By tapping into peripheral nerves or motor cortex signals, advanced prosthetic hands could receive real-time commands and potentially deliver sensory feedback to the user’s nervous system. Achieving stable, long-term neural interfaces will require breakthroughs in biocompatible electrodes, signal processing, and neural plasticity management, but the potential impact on prosthetic dexterity and user experience is monumental.
3. Multi-Functional Smart Materials: Soft robotic research increasingly explores single-material solutions that combine sensing, actuation, and structural support (Kim et al., 2013). For instance, certain piezoelectric polymers can generate electrical signals when deformed, providing inherent sensing capabilities. Integrating conductive inks or yarns into TCP actuators could yield embedded strain sensing. This approach reduces part count and simplifies manufacturing while enhancing synergy between components (Chen, Shen, & Shapiro, 2022).
4. AI-Driven Co-Design and Digital Twins: Digital twin technology offers the potential to simulate prosthetic designs—actuators, control algorithms, user biomechanics—in a virtual environment, refining them through iterative machine learning approaches before physical fabrication (Zhang, Wu, Tan, & Chen, 2023). This synergy accelerates optimization, customizing each prosthesis for the user’s residual limb shape, daily tasks, and personal preferences. Coupled with real-time data from in-device sensors, the digital twin can continuously update itself, guiding incremental improvements or alerting clinicians to maintenance needs.
5. Economical and Mass-Scalable Solutions: Ultimately, for widespread adoption, soft robotic prosthetic hands must be affordable and scalable. Achieving the advanced functionalities described herein without pricing out the majority of users remains a significant challenge (Pan et al., 2019). Researchers and manufacturers are exploring cost-reducing strategies such as standardized modular components, open-source hardware plans, and streamlined 3D printing processes. Bridging the gap between highly specialized laboratory prototypes and large-scale clinical deployment is an essential step toward democratizing access to these transformative devices.
Ethical, Social, and Policy Considerations
As these novel prostheses begin to merge more intimately with the human body, new ethical and societal questions arise:
Accessibility and Equity: Advanced prosthetic hands equipped with sophisticated actuators and control systems may carry high price tags, limiting access to affluent communities or well-funded healthcare systems (Pan et al., 2019). Ensuring equitable distribution is paramount. Foundations, subsidies, and manufacturing strategies for low-cost versions can help address these disparities.
User Autonomy and Privacy: With the integration of sensors, AI algorithms, and potential neural interfaces, there are concerns about data privacy, hacking, or unauthorized manipulation (Argall & Billard, 2010). Maintaining user autonomy and ensuring robust cybersecurity measures is crucial in preventing misuse or invasive data collection.
Psychological Well-Being: Appearance and social acceptance of prosthetic hands can significantly impact user self-esteem. Designers must weigh aesthetic integration with functional priorities (Lake, 2008). For some users, a visibly “robotic” prosthetic can be empowering; for others, a more natural look helps ease social interactions.
Regulatory Frameworks: Soft, bioinspired prostheses that incorporate artificial muscles or living tissues fall under emerging categories that regulatory bodies are still learning to evaluate. Guidelines for long-term safety, material biocompatibility, and device reliability must be refined (Caldwell & Ibarra, 2020).
Addressing these considerations calls for a multidisciplinary approach, involving engineers, medical professionals, ethicists, policymakers, and—critically—amputees themselves. By integrating scientific excellence with societal responsibility, the next generation of prosthetic hands can be both technologically groundbreaking and broadly beneficial.
Conclusion
The ongoing research on soft robotics presents an invaluable opportunity to revolutionize prosthetic hand technology, with artificial muscles at the forefront of this transformation. By incorporating bioinspired design, engineers and clinicians hope to create hands that not only reproduce the mechanical functions of human grasping but also provide a level of comfort, safety, and adaptability previously unattainable with conventional rigid-link prostheses (Laschi & Cianchetti, 2014; Rus & Tolley, 2015). This synergy can be seen across a spectrum of actuator technologies—pneumatic muscles, SMAs, dielectric elastomers, IPMCs, and twisted and coiled polymers—each offering unique advantages and challenges.
Key developments demonstrate that artificial muscle-driven prosthetic hands can achieve refined dexterous motion, better user comfort, and even partial sensory feedback through integrated sensors (Huang et al., 2021). Yet, real-world adoption remains hindered by issues of energy efficiency, material durability, control complexity, user acceptance, and cost (Belter et al., 2013). Researchers are pushing the boundaries further with hybrid systems, advanced manufacturing, machine learning-based EMG decoding, and the possibility of direct neural interfacing (Ortiz-Catalan, Hakansson, & Branemark, 2014). Coupled with ongoing miniaturization of power systems, the rise of self-healing and multifunctional materials, and robust design frameworks like digital twins, the horizon for soft prosthetic hands looks increasingly promising (Zhang et al., 2023).
The path forward, however, is not purely technical. Ensuring equitable access, addressing psychological and cultural needs, and establishing clear regulatory standards are all essential for making these sophisticated devices a true extension of the user’s body (Pan et al., 2019). Ethical considerations loom large as prostheses become ever more integrated and capable. In this evolving landscape, close collaboration between engineers, healthcare providers, patients, and policymakers stands as the linchpin for meaningful innovation.
Ultimately, the convergence of artificial muscle actuation and bioinspired design principles could yield prosthetic hands that rival or even surpass their biological counterparts in certain tasks. Whether it is an SMA-driven finger for a musician or a pneumatic muscle-based hand for an active lifestyle, the technology promises a future in which lost limbs can be replaced by devices that truly empower the individual. Through interdisciplinary research, thoughtful design, and a commitment to user well-being, the dream of a seamlessly integrated, soft robotic prosthetic hand continues to move closer to reality.
Acknowledgments
I would like to thank my research mentor, Jasper, for his invaluable guidance, unwavering support, and insightful feedback throughout the development of this work.
References
Abdulsadda, A. T., & Tan, X. (2013). Underwater biomimetic robot with IPMC sensor for obstacle detection. Bioinspiration & Biomimetics, 8(2), 026005.
Agarwal, A. (2020). Biohybrid actuators for next-generation soft robotics: Exploiting synergy between skeletal muscle and polymeric scaffolds. Advanced Materials Interfaces, 7(15), 2000415.
Argall, B., & Billard, A. (2010). A survey of Tactile Human–Robot Interactions. Robotics and Autonomous Systems, 58(10), 1159–1176.
Bar-Cohen, Y. (2012). Electroactive polymers for robotics applications. Woodhead Publishing.
Belter, J. T., Segil, J. L., Dollar, A. M., & Weir, R. F. (2013). Mechanical design and performance specifications of anthropomorphic prosthetic hands: A review. Journal of Rehabilitation Research & Development, 50(5), 599–618.
Caldwell, D., & Ibarra, J. C. G. (2020). Social interaction in soft robotics: Challenges and opportunities. IEEE Transactions on Robotics, 36(6), 1841–1855.
Carpi, F., Bauer, S., & De Rossi, D. (2010). Stretching dielectric elastomer performance. Science, 330(6012), 1759–1761.
Chen, Z., Shen, Z., & Shapiro, B. (2022). Modeling and control of IPMC actuators for soft robotic applications. Smart Materials and Structures, 31(1), 015021.
Chiapparo, S., Sears, N., & Comotti, D. (2021). Design strategies for multi-articulated 3D-printed prosthetic fingers. Journal of Mechanical Design, 143(3), 031705.
Childress, D. (1985). Historical aspects of powered limb prostheses. Clinical Prosthetics and Orthotics, 9(1), 2–13.
Firouzeh, M., Paik, J., & Neuhaus, T. (2015). Variable stiffness structures in soft robotics: Mechanisms and applications. Bioinspiration & Biomimetics, 10(6), 065003.
Gonzalez, J., Xu, R., Mendez, A. J., Wang, Z., & Zhu, B. (2021). Thermal management strategies for SMA-actuated prosthetic fingers. IEEE/ASME Transactions on Mechatronics, 26(5), 2564–2575.
Grosu, V., Asano, Y., Kobayashi, H., & Ito, S. (2021). Portable pneumatic sources for wearable robotics: A review. IEEE Transactions on Robotics, 37(4), 1359–1375.
Haines, C. S., Li, N., Spinks, G. M., Aliev, A. E., Di, J., & Baughman, R. H. (2014). New twist on artificial muscles. Science, 343(6173), 868–872.
Huang, H., Sira, D., Li, M., & Liu, Y. (2021). Integration of soft actuators and sensors: A review. Advanced Intelligent Systems, 3(5), 2100012.
Ishiguro, R., Shimizu, K., Yoda, T., Sato, S., & Kawamura, H. (2021). Development of a miniature silent air compressor for portable soft robotics applications. IEEE Robotics and Automation Letters, 6(3), 4871–4878.
Jani, J. M., Leary, M., Subic, A., & Gibson, M. A. (2014). A review of shape memory alloy research, applications and opportunities. Materials & Design, 56, 1078–1113.
Kargov, A., Pylatiuk, C., Oberle, R., & Werner, T. (2022). User-adaptive real-time EMG decoding for multi-functional hand prostheses with deep neural networks. Journal of Neural Engineering, 19(4), 046003.
Katzschmann, R. K., DelPreto, J., MacCurdy, R., & Rus, D. (2018). Exploring autonomous underwater maneuvers with an untethered soft robotic fish. Soft Robotics, 5(2), 129–145.
Kim, S., Laschi, C., & Trimmer, B. (2013). Soft robotics: A bioinspired evolution in robotics. Trends in Biotechnology, 31(5), 287–294.
Kyberd, P., & Hill, W. (2011). Survey of upper-extremity prosthesis users in Sweden and the United Kingdom. Journal of Prosthetics and Orthotics, 23(3), 142–150.
Lake, C. (2008). Optimizing the abilities of upper-limb amputees: New developments in shoulder and arm prostheses. Physical Medicine and Rehabilitation Clinics of North America, 19(4), 713–722.
Laschi, C., & Cianchetti, M. (2014). Soft robotics: New perspectives for robot bodyware and control. Frontiers in Bioengineering and Biotechnology, 2, 3.
Laschi, C., & Mazzolai, B. (2016). Soft robotics: Technologies and systems pushing the boundaries of robot abilities. Science Robotics, 1(1), eaah3690.
Liang, H., Carter, T., Bates, B., & Chan, V. (2022). Hybrid SMA-TCP exoskeletal hand for energy-efficient prosthetic grip. Soft Robotics, 9(4), 569–586.
Madden, J. D. (2007). Mobile robots: Motor challenges and materials solutions. Science, 318(5853), 1094–1097.
Marieb, E. N., & Hoehn, K. (2019). Human anatomy & physiology (11th ed.). Pearson.
Marchese, A. D., Onal, C. D., & Rus, D. (2014). Autonomous soft robotic fish capable of escape maneuvers using fluidic elastomer actuators. Soft Robotics, 1(1), 75–87.
Matysek, M., Kiefer, F., & Maden, S. (2020). Low voltage-driven dielectric elastomer stack actuators for prosthetic finger joints. Smart Materials and Structures, 29(9), 095031.
Mendez, A. J., Alba, C., Withrow, T. J., & VanderHook, T. (2020). Investigating ergonomics and mechanical efficiency in soft prosthetic sockets. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 28(5), 1157–1164.
Ortiz-Catalan, M., Hakansson, B., & Branemark, R. (2014). An osseointegrated human-machine gateway for long-term sensory feedback and motor control of artificial limbs. Science Translational Medicine, 6(257), 257re6.
Otsuka, K., & Wayman, C. M. (1998). Shape memory materials. Cambridge University Press.
Ottobock. (2019). History of prosthetics. Retrieved from Ottobock website.
Pan, L., Zhu, D., & Wang, X. (2019). Current state and progress of upper limb prosthesis technologies. Micromachines, 10(6), 407.
Park, E., Kim, Y., Heo, C., & Kim, H. (2021). Advances in wearable soft exosuits for muscle assistance and rehabilitation. Applied Sciences, 11(19), 9113.
Peerdeman, B., Boere, D., Witteveen, H., et al. (2011). Myoelectric forearm prostheses: State of the art from a user-centered perspective. Journal of Rehabilitation Research & Development, 48(6), 719–738.
Pelrine, R., Kornbluh, R., Pei, Q., & Joseph, J. (2000). High-speed electrically actuated elastomers with strain greater than 100%. Science, 287(5454), 836–839.
Pfeifer, R., Lungarella, M., & Iida, F. (2012). The challenges ahead for bio-inspired ‘soft’ robotics. Communications of the ACM, 55(11), 76–87.
Rus, D., & Tolley, M. T. (2015). Design, fabrication and control of soft robots. Nature, 521(7553), 467–475.
Saini, J., & Riener, R. (2021). Nylon-based artificial muscles and their applications in assistive devices: A review. ACS Applied Polymer Materials, 3(12), 6085–6101.
Salminger, S., Stino, H., Paternostro-Sluga, T., & Aszmann, O. (2020). Prosthetic reconstruction of the upper limb after traumatic nerve injuries: Advances in surgical management and rehabilitation. Disability and Rehabilitation, 42(18), 2610–2618.
Shahinpoor, M., Bar-Cohen, Y., Xue, T., Simpson, J., & Smith, J. (1998). Ionic polymer-metal composites (IPMCs) as biomimetic sensors, actuators and artificial muscles—A review. Smart Materials and Structures, 7(6), R15–R30.
Taddei, F., Tognarelli, S., Laschi, C., & Menciassi, A. (2019). A pneumatic muscle-driven soft prosthetic hand with anthropomorphic form and function. Soft Robotics, 6(5), 716–727.
Terryn, S., Groot, N. d., Georgopoulou, A., Brancart, J., & Vanderborght, B. (2021). Soft self-healing materials for robotic applications. Advanced Science, 8(2), 2002139.
Thakur, V. K., Ding, W., Ma, J., Lee, P. S., & Lu, X. (2018). Electroactive polymers for energy harvesting and sensing applications: A review. Polymers for Advanced Technologies, 29(5), 1589–1601.
Tondu, B. (2012). Modelling of the McKibben artificial muscle: A review. Journal of Intelligent Material Systems and Structures, 23(3), 225–253.
Tondu, B., & Lopez, P. (2000). Modeling and control of McKibben artificial muscle robot actuators. IEEE Control Systems Magazine, 20(2), 15–38.
Trimmer, B. (2013). A journal of soft robotics: Why now? Soft Robotics, 1(1), 1–4.
Ueda, R., Nishimura, T., Ogawa, D., & Ito, K. (2020). Dielectric elastomer prosthetic fingers with integrated tactile sensing. Smart Materials and Structures, 29(7), 075002.
Wang, Y., Li, G., & Ha, S. (2020). Human-centered design of prosthetic hand prototypes using user feedback and iterative testing. Prosthetics and Orthotics International, 44(5), 321–331.
Wehner, M., Truby, R. L., Fitzgerald, D. J., Mosadegh, B., Whitesides, G. M., Lewis, J. A., & Wood, R. J. (2016). An integrated design and fabrication strategy for entirely soft, autonomous robots. Nature, 536(7617), 451–455.
Yi, L., Xie, B., & Zhu, Y. (2022). Enhanced shape memory effect in NiTiCu-based alloys for rapid prosthetic actuation. Materials & Design, 213, 110358.
Zhang, R., Wu, X., Tan, W., & Chen, L. (2023). Digital twin for soft robotic hands: A deep learning approach to real-time control and optimization. IEEE Robotics and Automation Letters, 8(2), 950–957.
Zhang, Z., Xiao, X., Li, L., & Li, F. (2022). Progress in solid-state lithium-sulfur batteries for prosthetic robotics. Nano Energy, 95, 106975.
