A More Human Touch: Building the Next-Generation Bionic Hand

by

Author: Denis Avetisyan

Researchers have developed a new 15-degree-of-freedom bionic hand leveraging distributed actuation to enhance dexterity and achieve more natural movements.

The articulated finger demonstrates a kinematic structure enabling complex manipulation through coordinated joint movement, a design ultimately constrained by the realities of mechanical limits and potential for accumulated wear.
The articulated finger demonstrates a kinematic structure enabling complex manipulation through coordinated joint movement, a design ultimately constrained by the realities of mechanical limits and potential for accumulated wear.

This review details the design and development of a tendon-driven, 15-DoF bionic hand with a compact architecture enabled by distributed drive systems in the forearm and palm.

Achieving human-level dexterity in robotic hands remains a significant challenge, often requiring complex mechanisms and numerous actuators. This is addressed in ‘Development of a 15-Degree-of-Freedom Bionic Hand with Cable-Driven Transmission and Distributed Actuation’, which presents a novel design utilizing both forearm-mounted and palm-integrated motors to drive a 15-DoF hand. This distributed actuation system minimizes mechanical complexity while enhancing grasping capabilities and achieving a lightweight 1.4kg structure. Could this bio-inspired approach represent a viable pathway toward more versatile and adaptable robotic manipulation systems?

The Illusion of Dexterity: Why Hands Fail Us

Current prosthetic hands, while offering improved functionality over earlier designs, frequently fall short of replicating the intricate dexterity and adaptability inherent in the human hand. This limitation stems from the challenge of mirroring not just the anatomical structure, but also the complex interplay of muscles, tendons, and neural control that allows for a vast range of grips and manipulations. Consequently, many individuals relying on bionic hands experience difficulties with everyday tasks requiring fine motor skills – tasks most people perform without conscious thought, such as picking up small objects, tying shoelaces, or playing musical instruments. The resulting functional limitations can significantly impact quality of life, highlighting the need for continued innovation in prosthetic design to bridge the gap between artificial and natural hand capabilities.

The human hand, a marvel of biological engineering, presents a compelling model for next-generation prosthetic design. Its intricate architecture – encompassing 27 bones, numerous muscles, and a sophisticated network of nerves – allows for a remarkable range of motion and tactile feedback. This isn’t simply about replicating the hand’s shape, but understanding how it achieves such dexterity. Researchers are meticulously studying the interplay between intrinsic and extrinsic hand muscles, the role of ligaments in providing stability, and the complex feedback loops that enable precise grip control and force modulation. By deconstructing these nuanced control mechanisms, engineers aim to develop prosthetics that move beyond basic grasping functions, offering users the ability to perform delicate tasks, manipulate objects with varying textures, and ultimately, regain a greater sense of natural, intuitive control. The hand’s design serves as a robust blueprint, guiding the development of increasingly sophisticated and adaptable prosthetic limbs.

Replicating the intricate dexterity of the human hand in prosthetic devices presents significant engineering hurdles, primarily concerning the balance between functional complexity and practical limitations of size, weight, and energy use. Traditional prosthetic hands often sacrifice degrees of freedom – the number of independent movements a hand can make – to conserve power and simplify design. However, the IRMV Hand represents a notable advancement by achieving fifteen degrees of freedom, a figure approaching the range of motion found in a natural human hand. This increased articulation allows for a wider variety of grips and manipulations, enabling users to perform tasks requiring subtle adjustments and precise control – from picking up delicate objects to executing complex gestures – while still remaining within reasonable physical and energetic constraints for daily use.

The 15-degree-of-freedom IRMV Hand, measuring approximately 100mm x 100mm x 265mm and closely mirroring the size of an adult human hand, utilizes fifteen motors housed within its forearm and palm to replicate complex grasping and manipulation capabilities.
The 15-degree-of-freedom IRMV Hand, measuring approximately 100mm x 100mm x 265mm and closely mirroring the size of an adult human hand, utilizes fifteen motors housed within its forearm and palm to replicate complex grasping and manipulation capabilities.

Distributed Control: Cramming Less Into More Space

The IRMV Hand employs a distributed drive system characterized by the physical separation of actuation hardware. Instead of consolidating all motors in a single location, such as the wrist or forearm, drive components are strategically positioned within both the forearm and the palm itself. This architectural choice directly addresses size constraints, allowing for a more compact hand design by reducing the volume occupied by driving mechanisms. Furthermore, distributing the motors optimizes available space within the hand, enabling the inclusion of additional sensors, linkages, or compliant elements without increasing the overall device dimensions.

The IRMV Hand employs a cable-driven transmission system to transfer motive force from the motors, located in the forearm and palm, to each finger joint. This mechanism utilizes sheathed cables routed through the hand’s internal structure, functionally analogous to biological tendons. By selectively actuating these cables, precise control over individual finger movements and grip force is achieved. This design minimizes the need for bulky mechanical linkages within the hand itself, contributing to the device’s compact size and allowing for a distributed weight profile. The cable system allows for force transmission with minimal backlash, contributing to the hand’s dexterity and responsiveness.

The IRMV Hand achieves dexterity through independent actuation of each joint, utilizing a dedicated motor for each degree of freedom. This configuration enables a broad spectrum of grasping configurations and complex movements beyond the capabilities of underactuated designs. Each motor contributes to precise control and allows for adaptable grasping forces, culminating in a maximum compressive force of 11 N. This level of force, combined with the independent control, allows the hand to manipulate a variety of objects with differing size, shape, and fragility.

The IRMV Hand utilizes a hardware system integrating actuators, sensors, and a control board to enable dexterous manipulation.
The IRMV Hand utilizes a hardware system integrating actuators, sensors, and a control board to enable dexterous manipulation.

Grasping Reality: Benchmarks and Volume Metrics

The IRMV Hand’s grasping performance was evaluated using the Grasp Taxonomy Benchmark, a standardized assessment of robotic hand dexterity. This benchmark categorizes grasps based on the contact surfaces and the degrees of freedom utilized. Successful completion of 33 distinct grasp types, as defined by the benchmark, indicates a high degree of manipulative capability. These grasps included power, precision, tripod, and lateral grasps, among others, demonstrating the hand’s ability to adapt to a variety of object shapes, sizes, and required grip forces. This performance level suggests the IRMV Hand is capable of performing a wide range of tasks requiring diverse grasping strategies.

Workspace volume for the IRMV Hand was determined to be 99 cm³ per finger, a metric calculated based on the reachable space of each digit during operational testing. This volume represents the three-dimensional space within which the finger’s end-effector can be positioned and manipulated. A larger workspace volume indicates a greater range of motion and improved operational flexibility, allowing the hand to interact with a wider variety of objects and environments without requiring significant repositioning of the base unit. The measurement was consistently achieved across all five fingers, confirming uniform operational capacity and predictable movement throughout the hand’s workspace.

The IRMV Hand utilizes Proportional-Integral-Derivative (PID) control algorithms to ensure accurate and stable finger and wrist positioning. These algorithms are implemented on a STM32F411 microcontroller, which processes position data acquired through an ADS1118 analog-to-digital converter. The ADS1118 provides high-resolution readings of sensor data, enabling the STM32F411 to calculate necessary adjustments to the motor controls. This closed-loop control system continuously monitors and corrects for deviations from desired positions, resulting in precise and repeatable movements.

The GRASP taxonomy categorizes a wide range of grasp types achievable with the IRMV Hand.
The GRASP taxonomy categorizes a wide range of grasp types achievable with the IRMV Hand.

The Illusion of Natural Control: Where Biomimicry Meets Reality

The IRMV Hand achieves a remarkable level of dexterity through a sophisticated interplay of engineering principles. Rather than relying on a single motor to control all movements, a distributed drive system places multiple actuators closer to the joints, mirroring the biological structure of the human hand. This is coupled with a cable-driven mechanism – analogous to tendons – which translates the motor’s power into precise finger and thumb movements. Crucially, this system isn’t simply about power, but precise control; advanced algorithms coordinate these actuators to enable a wide range of grips – from delicate pinching to powerful grasping – allowing the prosthetic hand to perform complex tasks such as manipulating small objects, using tools, and even playing musical instruments with a fluidity previously unattainable in prosthetic devices.

Integrating force sensors into advanced prosthetic hands represents a crucial step toward restoring natural manipulation capabilities. These sensors, strategically positioned across the prosthetic fingertips and palm, would detect the magnitude and distribution of forces during object contact. This information isn’t merely about detecting pressure; it’s about translating that pressure into nuanced signals relayed to the user via neural interfaces or haptic feedback systems. Consequently, the user gains a more comprehensive understanding of how firmly they are grasping an object – preventing slippage or crushing – and enabling delicate tasks like handling fragile items or assembling small components. The addition of tactile feedback not only improves the precision and control of prosthetic hands but also fosters a greater sense of embodiment and reduces the cognitive load associated with everyday manipulation, ultimately leading to more fluid and intuitive prosthetic use.

The IRMV Hand distinguishes itself through a biomimetic design, specifically its reliance on a tendon-driven mechanism that mirrors the functionality of the human hand. Unlike traditional prosthetic designs employing bulky motors and rigid linkages, this approach utilizes cables routed to mimic the flexor and extensor tendons in a biological hand. This anatomical fidelity isn’t merely aesthetic; it allows for a more natural range of motion and a distribution of force that closely resembles human dexterity. Consequently, users may experience a more intuitive connection with the prosthesis, potentially requiring less conscious effort and cognitive load to perform tasks. The design anticipates that such a system will facilitate the development of control schemes based on natural hand movements, effectively translating the user’s intended actions into fluid and precise prosthetic manipulation.

The IRMV Hand utilizes a position sensing circuit to accurately determine finger and joint positions.
The IRMV Hand utilizes a position sensing circuit to accurately determine finger and joint positions.

The pursuit of increasingly complex robotic prosthetics feels perpetually stuck in a cycle. This bionic hand, with its fifteen degrees of freedom and distributed actuation, is certainly an impressive engineering feat. However, one suspects it will soon join the ranks of ‘solutions’ that introduce fresh complications. As Henri Poincaré observed, “Mathematics is the art of giving reasons.” This hand attempts to reason its way to human-like dexterity, but production environments will inevitably reveal unforeseen limitations. The elegance of the distributed drive system-a core concept of this design-will likely be obscured by the pragmatic need for maintenance and repair. Everything new is just the old thing with worse docs, and this hand, despite its innovations, will be no different.

Futureproofing This Hand

The pursuit of anthropomorphism in robotics invariably encounters the harsh realities of practical implementation. This design, with its distributed actuation and fifteen degrees of freedom, represents a logical, if ambitious, step towards a truly dexterous prosthetic. The immediate challenge, predictably, will not be further refinement of kinematics, but longevity. Each additional degree of freedom is a new point of failure, a new cable to stretch, a new motor to overheat. The elegance of the theoretical model will, in time, be measured by the frequency of emergency repairs.

The real bottleneck, however, remains control. Achieving fluid, intuitive manipulation requires more than just mechanical fidelity; it demands a control system capable of translating intent into coordinated movement without inducing jitter or instability. Current solutions, reliant on electromyography or similar interfaces, are still a far cry from the seamless integration of the nervous system. Expect to see the field increasingly focused on machine learning approaches-algorithms trained to anticipate user needs, effectively smoothing over the inevitable imperfections of the hardware.

Ultimately, this hand-and others like it-will not be judged by how closely they resemble a human hand, but by how effectively they extend human capability. It is a pleasant illusion to believe in perfect replication. The long game isn’t about building a better hand; it’s about building a hand that’s reliably useful for a decade, even if that means accepting a few compromises along the way. Legacy, after all, is just a memory of better times.

Original article: https://arxiv.org/pdf/2512.04399.pdf

Contact the author: https://www.linkedin.com/in/avetisyan/

See also:

2025-12-05 08:46