Artificial arms are engineered devices that replace a missing upper limb, ranging from simple passive cosmetic covers to sophisticated robotic systems driven by muscle signals or even brain activity. They work by translating some remaining physical input from the user, whether that is shoulder movement, electrical signals from residual muscles, or neural impulses recorded by implanted electrodes, into mechanical motion at the elbow, wrist, or hand. The technology spans a surprisingly wide spectrum, and the way a given prosthesis functions depends heavily on its type, the level of amputation, and the surgical and electronic infrastructure supporting it.
Body-Powered Arms
The oldest and most mechanically straightforward type of artificial arm is the body-powered prosthesis. These devices use a harness and cable system strapped across the opposite shoulder. When you shrug, extend your shoulder, or shift your torso in specific ways, the cable transmits that movement to a terminal device, usually a hook or a mechanical hand, causing it to open or close. There are no batteries, no electronics, and no software. The entire system is mechanical.
That simplicity comes with real advantages. Body-powered arms tend to be lighter, more durable, cheaper to maintain, and quicker to learn. The cable also gives the user a form of feedback: you can feel the tension in the harness change as the terminal device grips an object, which provides a rough sense of how hard you are squeezing. A systematic review comparing the two main prosthesis types found that body-powered devices had advantages in durability, training time, frequency of adjustment, and maintenance, while also providing better physical feedback to the user.1PubMed. Differences in myoelectric and body-powered upper-limb prostheses: Systematic literature review The trade-off is limited grip strength, a narrow range of motion, and the visible harness system, which many users find cosmetically unappealing.
How Myoelectric Arms Work
Myoelectric prostheses represent the most common powered alternative. Instead of cables and shoulder motion, they detect the tiny electrical signals your muscles produce when they contract. Sensors on the inner surface of the prosthetic socket pick up these signals from the muscles remaining in the residual limb. A processor interprets those signals and drives small electric motors that move the hand, wrist, or elbow.
In the simplest configuration, two muscle sites control the device: contracting one muscle group opens the hand, and contracting the other closes it. Switching between hand and wrist movements typically requires a specific gesture like co-contracting both muscle groups at once. This sequential, one-joint-at-a-time control works, but it is slow and mentally taxing. Research has shown that newer controllers allowing simultaneous two-joint control, where the hand and wrist move at the same time, can perform functional tasks two to four times faster than the conventional sequential approach.2PubMed Central. Myoelectric Control Performance of Two Degree of Freedom Hand-Wrist Prosthesis by Able-Bodied and Limb-Absent Subjects
More advanced myoelectric systems use pattern recognition. Instead of relying on one or two muscle sites, they record from several sensors at once and use software to recognize distinct patterns of muscle activity. When you think about rotating your wrist versus closing your fist, your residual muscles fire in subtly different patterns, and the algorithm learns to tell them apart. Although this approach has been studied for decades, its real-world clinical benefits over simpler direct control are still being evaluated.3PubMed Central. Evaluation of EMG pattern recognition for upper limb prosthesis control: a case study in comparison with direct myoelectric control The same systematic review that highlighted body-powered advantages found that myoelectric arms offered better cosmetic appearance and were more accepted for lighter tasks, but concluded there was insufficient evidence to say either type was broadly superior.1PubMed. Differences in myoelectric and body-powered upper-limb prostheses: Systematic literature review
Wrist and Hand Mechanics
One challenge with prosthetic arms that often surprises people is the wrist. A biological wrist bends, rotates, and tilts almost unconsciously, repositioning the hand for every task from turning a doorknob to pouring a glass of water. Most commercial prosthetic wrists offer only a single axis of rotation, which forces users to compensate with awkward shoulder and elbow movements. New designs are tackling this with multi-axis wrists. One recent approach uses a single motor driving a three-axis lockable wrist, allowing users to adjust hand position during the reach phase of a grasp and then lock the wrist in place for stability.4The International Journal of Robotics Research. A three degrees of freedom switchable impedance myoelectric prosthetic wrist
On the hand side, the goal is to replicate as many useful grip patterns as possible without making the device too heavy or complex. A biological hand has over twenty degrees of freedom; a prosthetic hand typically has between one and six. Modern multi-grasp hands use several small motors arranged to provide both precision grips, where the fingertips come together for small objects, and power grasps that wrap around larger items like bottles or tools.5PubMed Central. A Multigrasp Hand Prosthesis for Providing Precision and Conformal Grasps
Surgical Techniques That Expand Control
One of the biggest bottlenecks in prosthetic arm function is not the hardware itself but the control signals available to drive it. After an amputation, many of the nerves that once controlled the hand and fingers are still alive in the residual limb, carrying motor commands from the brain with nowhere to go. Two surgical techniques have emerged to reclaim those signals.
Targeted muscle reinnervation, or TMR, reroutes those orphaned nerves into nearby muscles that have lost their original function. Once the nerves grow into the new muscle targets over several months, the muscles become biological amplifiers: when the person thinks about closing their hand, the reinnervated muscle contracts, and surface sensors on the socket pick up the signal.6PubMed Central. Targeted muscle reinnervation and advanced prosthetic arms The result is more independent control signals and more intuitive movement. Studies of people with below-elbow amputations who underwent TMR showed measurable improvements in hand function tests nine to twelve months after surgery.7PLOS ONE. Myoelectric prosthesis hand grasp control following targeted muscle reinnervation in individuals with transradial amputation
A newer approach, the regenerative peripheral nerve interface (RPNI), wraps a small piece of donor muscle graft around a severed nerve ending. The nerve regenerates into this graft, creating a tiny muscle unit that amplifies the nerve’s electrical output enough for implanted or surface electrodes to detect it clearly. What makes this technique especially promising is its long-term stability. In one study, a participant maintained real-time prosthetic control above 94% accuracy for over 600 days without any recalibration of the control software, and completed a multi-step coffee-making task with 99% accuracy across the same period.8Journal of Neural Engineering. Long-term upper-extremity prosthetic control using regenerative peripheral nerve interfaces and implanted EMG electrodes That kind of durability matters enormously for daily life, since frequent recalibration sessions are one of the things that make advanced prostheses impractical for many users.
Restoring a Sense of Touch
A prosthetic arm that moves but cannot feel is like typing with thick gloves on: you can do it, but you’re clumsy, slow, and never quite sure what you’ve grabbed. The absence of sensory feedback is a persistent complaint among prosthesis users. Several approaches are being developed to bring sensation back.
The most direct method involves implanting tiny electrodes around the peripheral nerves still present in the residual limb. When sensors on the prosthetic fingertips detect contact or pressure, the system converts that information into electrical pulses delivered to the nerves. The user perceives these as touch sensations localized to specific fingers of their missing hand. In testing, this kind of feedback improved people’s ability to discriminate between objects and manipulate them more carefully.9PubMed Central. Sensory feedback by peripheral nerve stimulation improves task performance in individuals with upper limb loss using a myoelectric prosthesis A separate study using a different electrode design found that providing this biomimetic sensory feedback helped a participant grip fragile objects with greater precision.10PubMed. Biomimetic sensory feedback through peripheral nerve stimulation improves dexterous use of a bionic hand
For people who are not candidates for implanted electrodes, a less invasive option is sensory substitution. Here, information from the prosthetic hand’s sensors is translated into vibrations or pressure applied to intact skin on another part of the body, like the upper arm or thigh. The brain learns to interpret those vibrations as touch information from the prosthesis. Research on vibrotactile displays worn on the thigh has found that people can distinguish between two separate vibration points when they are spaced about 25 to 30 millimeters apart, which sets a practical minimum for how densely these feedback devices can encode spatial information.11PubMed. Optimizing Vibrotactile Feedback for Sensory Substitution in the Thigh: Spatial Acuity and Frequency Characteristics
Beyond touch, researchers are also working on proprioception, the sense of where your limb is in space and how it is moving. This is what lets you reach for a light switch in the dark without looking at your hand. Restoring proprioceptive feedback in a prosthetic limb involves stimulating muscle spindles or other deep receptors, which is a different physiological target than the skin sensors used for tactile feedback. The research is still in relatively early stages, but combining both types of sensation into a single system is considered an important goal for making prosthetic limbs feel truly integrated with the body.12PubMed Central. A worldwide research overview of Artificial Proprioception in prosthetics
Brain-Computer Interfaces
For people with very high-level amputations or spinal cord injuries that leave no usable muscle signals, brain-computer interfaces (BCIs) offer an alternative control pathway. Small electrode arrays are surgically implanted on the surface of the motor cortex, the brain region that plans and executes voluntary movement. These electrodes record the firing patterns of nearby neurons while the person thinks about moving their arm. Decoding algorithms translate those neural patterns into commands that drive a robotic limb.
The results, while still largely confined to research settings, are striking. In several cases, people with complete paralysis have used BCIs to reach out, grasp objects, and bring them to their mouths, movements they had been unable to perform for years.13PubMed Central. Review: Human intracortical recording and neural decoding for brain computer interfaces One particularly compelling demonstration added bidirectional capability: the system not only decoded movement intentions from the motor cortex but also stimulated the somatosensory cortex to create tactile sensations when the robotic hand touched something. With this sensory feedback loop active, the participant completed a standard clinical arm assessment roughly twice as fast, cutting median task times from about 21 seconds to about 10 seconds.14PubMed Central. A brain-computer interface that evokes tactile sensations improves robotic arm control
BCIs remain far from everyday clinical use. The implanted electrodes can degrade over time, the supporting hardware is bulky, and the surgery carries inherent risks. But they represent the clearest proof that the brain’s motor commands can be captured and used to drive an artificial limb with multiple independent degrees of freedom, even when no peripheral nerves or muscles are available.
Materials, Sockets, and Attachment
However sophisticated the control system, a prosthetic arm is only as useful as its connection to the body. The socket, the cup-shaped interface that fits over the residual limb, is where most users encounter the most day-to-day frustration. Traditional prosthetic sockets have been made from composites reinforced with glass, carbon, or Kevlar fibers to balance strength with weight.15PubMed Central. Strength Assessment of PET Composite Prosthetic Sockets Getting the fit right requires skilled prosthetists and often multiple adjustments. A poorly fitting socket can cause skin breakdown, pain, and sweating, all of which drive people to stop wearing the device.
Osseointegration bypasses the socket entirely. In this approach, a titanium implant is surgically anchored directly into the bone of the residual limb, and the prosthesis attaches to a small abutment that protrudes through the skin. This eliminates the socket’s pressure points and heat trapping, restores full range of motion at the shoulder, and means the prosthesis always “fits” regardless of fluctuations in limb volume due to swelling or weight changes.16PubMed. Osseointegration amputation prostheses on the upper limbs: methods, prosthetics and rehabilitation The trade-off is the risk of infection at the skin-implant boundary and the need for careful, ongoing monitoring of the bone-implant junction.
Artificial Intelligence and Automated Grasping
Even with good control signals, picking up an unfamiliar object requires the user to think through a sequence of decisions: which grip pattern to select, how wide to open the hand, how much force to apply. Researchers are increasingly turning to computer vision and machine learning to automate parts of this process. A camera mounted on the prosthesis or nearby glasses identifies the object, estimates its shape and size, and pre-selects an appropriate grasp before the user’s hand even reaches it. In simulation testing, reinforcement learning algorithms trained on automated grasping tasks have achieved success rates as high as 99%.17PubMed Central. On Automated Object Grasping for Intelligent Prosthetic Hands Using Machine Learning Translating that performance from simulation to the unpredictable real world is an ongoing challenge, but the idea of a prosthetic hand that sees what you are reaching for and prepares accordingly is no longer purely speculative.
Soft robotics is another emerging approach that sidesteps some of the control problem altogether. Instead of rigid motors and gears, soft pneumatic actuators use air pressure to inflate flexible chambers made of silicone rubber, causing them to bend and curl like fingers. These actuators naturally conform around objects of varying shapes, reducing the need for precise grip planning. The materials are lightweight, inherently safe for human contact, and well suited to handling delicate items.18PubMed Central. Soft Pneumatic Muscles: Revolutionizing Human Assistive Devices with Geometric Design and Intelligent Control
Why Many People Stop Wearing Their Prosthesis
Despite all of this technology, prosthetic arm abandonment rates are high compared with lower-limb devices. A multicenter study found that the most common reason for giving up an upper-limb prosthesis, cited by about four in five users who abandoned theirs, was limited functionality.19PubMed Central. Cross-sectional International Multicenter Study on Quality of Life and Reasons for Abandonment of Upper Limb Prostheses A scoping review examining abandonment reasons over time found that comfort and function have consistently been the dominant factors, with weight, heat buildup, and perspiration being the most persistent comfort complaints. Some users reported feeling more functional without the device altogether.20PubMed. Comfort and function remain key factors in upper limb prosthetic abandonment: findings of a scoping review
This is the uncomfortable truth about the field: a prosthetic arm that looks impressive in a lab demonstration may not survive the daily reality of commuting, cooking, working, and sweating. The gap between what is technically achievable in controlled research settings and what is practically tolerable for eight or more hours a day remains wide.
Prosthetic Arms for Children
Fitting children presents a distinct set of challenges. Kids grow fast, meaning sockets and devices need frequent replacement. Many children with congenital upper limb differences adapt remarkably well with one hand and may resist wearing a device that feels heavier or clumsier than going without. A review of pediatric prosthetics found that most available devices offer only a single open-close grasp and often have non-anthropomorphic appearances, falling short of what children need to perform daily tasks effectively and engage socially with peers.21Prosthetics & Orthotics International. A review of upper limb pediatric prostheses and perspectives on future advancements For a child, the social dimension can matter as much as function. A device that draws unwanted attention or looks robotic in a way the child dislikes may stay in a drawer regardless of how well it grips.
Access and Affordability Around the World
Advanced prosthetic arms with myoelectric control, pattern recognition, or sensory feedback can cost tens of thousands of dollars, and that does not include the ongoing expenses of maintenance, battery replacement, socket refitting, and clinical follow-up. In much of the world, even a basic body-powered device is out of reach. A systematic review of barriers in sub-Saharan Africa found that poverty, limited access to employment after limb loss, and the absence of nearby rehabilitation services all compound to keep prosthetic technology away from the people who need it.22PubMed Central. Exploring barriers to access and use of limb prostheses and orthoses in sub-Saharan Africa: A systematic review The situation is not unique to Africa; rural areas in wealthier countries face similar, if less severe, access gaps. Three-dimensional printing has generated considerable optimism as a way to lower costs and speed up fabrication, particularly for children who outgrow devices quickly, but printed devices still lag behind traditionally manufactured ones in grip strength and durability for demanding daily use.
Waterproofing is another practical concern that rarely makes headlines but matters a great deal in daily life. Standard myoelectric prostheses are vulnerable to moisture, meaning users often remove them for bathing, dishwashing, or getting caught in the rain. Ongoing engineering work on insulated sensors and sealed housings is gradually expanding the environments where powered arms can be worn reliably.23URNCST Journal. What Are Artificial Arms and How Do They Work? For a technology that is supposed to restore normalcy, having to take off your hand before you can wash your dishes is a particularly ironic limitation, and one that engineers are only now beginning to solve seriously.