Making a prosthetic hand at home is a realistic project thanks to 3D printing and open-source designs. The most accessible route uses freely available designs from the e-NABLE community, a consumer 3D printer, and about $15 to $20 in materials for a basic body-powered hand. More advanced builds can incorporate muscle sensors and servo motors for electrically powered grip, though complexity and cost rise quickly.
Choosing a Design Type
Prosthetic hands fall into two broad categories: body-powered and electrically powered. The type you build determines the materials, cost, and skill level required.
Body-powered hands use the wearer’s own movement to open and close the fingers. A cable runs from a harness on the wrist or forearm to tendons inside the fingers. When the wearer flexes their wrist, the cable pulls the fingers closed. These devices are low cost, lightweight, and straightforward to repair. The trade-off is that they don’t mimic natural hand movement closely, and some users find the cable tension tiring over long periods.
Electrically powered (myoelectric) hands use small sensors placed on the skin over forearm muscles. When the wearer contracts those muscles, the sensors detect the electrical signal and translate it into a command that drives small motors inside the fingers. These hands look more natural and can perform more grip patterns, but they require batteries, electronic components, and programming. A professionally made myoelectric hand can cost tens of thousands of dollars, while a DIY version with consumer electronics and 3D-printed parts can be built for a few hundred.
Starting With an Open-Source Design
The easiest entry point is the e-NABLE project, a global volunteer network that shares free prosthetic hand designs. Their most popular models, like the Raptor Reloaded and Phoenix Hand, are body-powered designs meant for 3D printing. You download the STL files, print the parts, and assemble them with hardware-store components.
These designs are sized parametrically, meaning you can scale the files up or down to match the user’s limb. The critical measurement is the length of the residual limb and its circumference at several points. For a below-wrist design, you’ll measure the distance from the end of the limb to the wrist crease, then the circumference at the widest point and again halfway along the limb. Conventional prosthetic fitting uses plaster casting to capture the exact shape, but for a 3D-printed device, careful caliper and tape measurements or a 3D scan work for initial prototyping. Expect to print and reprint the socket portion several times to get a comfortable fit.
Materials for 3D Printing
A functional prosthetic hand uses two types of plastic. The rigid structural parts, including fingers and the dorsal frame, print well in PLA (polylactic acid), the most common and forgiving 3D printing filament. PLA is stiff, inexpensive, and prints at relatively low temperatures on any consumer printer.
The palm and any parts that contact the residual limb benefit from a flexible material. TPU (thermoplastic polyurethane) at 85A shore hardness is a well-tested choice. It’s firm enough to hold its shape but flexible enough to conform to the limb and absorb impact. Printing TPU requires a direct-drive extruder or at least a well-tuned Bowden setup, since the flexible filament can jam in the feed path. A printer like the Prusa i3 MK3 handles both PLA and TPU without modification.
Print settings matter for strength. Fingers should be printed with higher infill (40% or more) and oriented so the layer lines run along the length of the finger, not across it. A finger printed with layers perpendicular to its length will snap under grip force.
Non-Printed Parts You’ll Need
The printed shell is only part of the build. A body-powered hand needs tendons, elastic return cords, fasteners, and padding. Here’s what a typical e-NABLE build requires:
- Tendon line: Braided Spectra fishing line (such as Power Pro) or Dyneema cord. These are extremely strong for their diameter and resist stretching, which keeps finger movement crisp.
- Elastic cord: 2mm elastic line provides the return force that opens the fingers after each grip.
- Screws: Small machine screws and stop nuts form the pin joints at each knuckle. Sizes range from #2 to #8 depending on the hand’s scale. Chicago screws (binding screws) are used where parts need to pivot smoothly.
- Beads: Small craft beads anchor the tendon and elastic lines inside the fingertips.
- Velcro straps: Secure the device to the limb. Pre-cut 12-inch by 1-inch strap kits with buckles are available from 3D Universe.
- Padding: Medical-grade thermoformable plastic (like Aquaplast) lines the socket for comfort. Foam padding from suppliers like Patterson Medical adds cushioning over bony areas.
- Protective sleeve: A cotton or nylon stockinette liner sits between the skin and the socket to reduce friction and absorb sweat.
All of these are available from Amazon or medical supply retailers. The total hardware cost for a body-powered hand typically runs $15 to $30.
Assembly Basics
Assembly follows a logical sequence: build the fingers first, then attach them to the palm, then fit the palm to the socket.
Each finger is a chain of two or three printed segments connected by pin joints (the small screws). Thread the tendon line through the channel on the underside of each segment, anchor it at the fingertip with a bead and knot, and run it down through the palm to a common tensioner point at the wrist. Thread elastic cord through a separate channel on the top of the finger, anchored the same way. The elastic holds the finger in an open position at rest. When the user flexes their wrist, the tendon line pulls tight, curling the fingers closed. When they relax, the elastic snaps the fingers back open.
The tensioner screws at the wrist let you adjust how much wrist flexion is needed to close each finger. Tuning these is one of the most important steps. Too loose and the fingers barely move. Too tight and the hand is exhausting to use. Start with light tension and increase gradually while the user tests gripping soft objects like a foam ball.
Adding Electronic Control
If you want to build a myoelectric hand, you’ll layer electronics onto a similar mechanical frame. The core components are muscle sensors, a microcontroller, and servo motors.
Surface EMG sensors attach to the forearm over the muscles that control wrist and finger movement. The two key muscle groups are on the inner forearm: the flexor carpi radialis and flexor carpi ulnaris. A typical sensor uses three small silver electrodes spaced about 1.25 cm apart, two over the target muscle and one reference electrode, usually near the elbow. When the user contracts those muscles, the sensor picks up the electrical signal through the skin.
An Arduino Uno or similar microcontroller reads the sensor output and converts it into motor commands. A PCA9685 servo motor driver board lets a single Arduino control multiple servo motors, one per finger. The code maps the strength of the muscle signal to the angle of each servo: a light contraction might close the fingers partway (a pinch grip), while a strong contraction closes them fully (a power grip). The servo angle range on the PCA9685 driver maps from 150 to 600, corresponding to roughly 0 to 150 degrees of physical rotation. Open-source Arduino libraries like PWMServoMotorDriver.h handle the low-level communication with the driver board.
Calibration is the hardest part. Every user produces different signal strengths, and the same user’s signals vary with fatigue, sweat, and sensor placement. You’ll need to record baseline readings with the sensors in place, then set thresholds that reliably distinguish between “open hand,” “pinch,” and “power grip” without false triggers. A simple proportional control scheme, where signal strength directly scales motor speed, is the most forgiving starting point.
Grip Patterns That Matter
A natural hand performs dozens of grip types, but a prosthetic hand only needs a handful to cover most daily tasks. The functional grips that matter most fall into two categories.
Power grasps wrap all fingers around an object. A cylindrical grip (grabbing a bottle) and a spherical grip (holding a ball) are the two main variations. Pinch grasps use the thumb and one or two fingers for precision. A key grip (turning a key between thumb and the side of the index finger), a pinch grip (thumb to index fingertip), and a three-fingered grip (thumb, index, and middle finger) cover most fine manipulation tasks. Research on grip classification shows that distinguishing between these broad categories, power versus pinch versus open, is much more reliable than trying to classify every individual grip type. For a DIY myoelectric hand, programming two or three distinct grip modes is a practical target.
Cost: DIY vs. Professional
The cost gap between a homemade and professionally fitted prosthetic hand is enormous. A basic 3D-printed body-powered hand costs $15 to $30 in materials, assuming you already have access to a 3D printer. A DIY myoelectric hand with servos, sensors, and an Arduino runs $100 to $400 depending on component quality.
On the professional side, a multi-articulating myoelectric hand from a major manufacturer costs tens of thousands of dollars, and that’s before fitting and rehabilitation. Even lower-cost commercial alternatives, like Unlimited Tomorrow’s TrueLimb at around $7,000, proved too expensive for many consumers when sold directly. The gap isn’t just price: professional prosthetics include custom-molded sockets, clinical fitting by a certified prosthetist, and ongoing adjustments. A DIY hand won’t match that level of fit or durability, but for partial hand loss or as a functional tool for specific tasks, it can be remarkably useful.
Fitting the Socket
The socket is the interface between the device and the body, and it determines whether the hand gets worn daily or abandoned in a drawer. A poorly fitting socket causes skin irritation, pressure sores, and pain within minutes.
Start by taking circumference measurements at multiple points along the residual limb and recording the limb’s length. If you have access to a 3D scanner (even a smartphone-based one like Polycam), scanning the limb gives you a digital model you can use to shape the socket in CAD software. Print the socket in TPU for flexibility, and line the interior with thermoformable plastic sheeting or foam padding. The socket should distribute pressure evenly, with relief areas over bony prominences. A cotton stockinette liner worn underneath reduces friction and makes the socket easier to put on and remove.
Plan to iterate. Print a test socket, have the user wear it for 15 to 20 minutes, then check for red marks or pressure points. Modify the digital model, reprint, and test again. Professional prosthetists go through a similar cycle with plaster molds and manual rectification. Three to five revisions is normal before the fit feels right.