Industries / Health / Medical devices / Prosthetics
Limbs
Limbs covers the parts of an artificial leg or arm that work without motors: feet, ankles, knees, hooks and hands.
6
Milestones achieved, from 950–710 BCE to 2020.
2
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
950–710 BCE
A wooden toe that worked
A big toe of wood and leather, in three parts, was found on the foot of a woman's mummy buried near Luxor, Egypt, and dates from between 950 and 710 BCE. In 2012 Jacky Finch of the University of Manchester had replicas of it, and of the cartonnage Greville Chester toe from before 600 BCE, worn by two volunteers who had lost a right big toe; wearing the wood and leather copy in a sandal, one volunteer reached nearly 78% of the bend of their own left toe, and pressure under the foot stayed low. The university says the two toes are likely to be the world's first prosthetic body parts. The Science Museum in London had called a bronze leg from a grave at Capua, Italy, of about 300 BCE, the oldest artificial limb to be discovered; the original was destroyed in an air raid in the Second World War.
Credit
unknown Egyptian makers; tested by Jacky Finch, University of Manchester, in the gait laboratory of the University of Salford
1912
A hook that grips itself
David W. Dorrance of Koler, Oregon, filed a patent on 17 February 1912, granted on 29 October 1912 as US 1,042,413, for an artificial hand made of two hooked fingers, one fixed and one hinged. An elastic band holds the fingers shut, so the hook grips a knife or an umbrella handle on its own; the wearer opens it by bending the elbow. Devices that work this way are called voluntary opening: effort opens them and a spring or band closes them. Hooks closed by rubber bands were still among the devices tested in Delft in 2012 (below).
Credit
David W. Dorrance
1957
Solid ankle, cushion heel
The SACH foot (solid ankle, cushion heel) has no moving ankle: its soft heel absorbs the shock of each step and the foot rolls from heel to toe. J. Foort and C. W. Radcliffe of the Prosthetic Devices Research Project at the University of California, Berkeley, designed its first prototypes in 1957, and production models were approved on 24 May 1957 by the Committee on Prosthetics Research and Development, according to a 2019 history by Rakesh Bhargava. It was accepted almost everywhere in Europe and North America but not in Southeast Asia, South America or Africa, because it needed a shoe and did not wear well in water-filled fields or on rough ground.
Lessona foot designed around shoes fails where people work barefoot.
Credit
J. Foort and Charles Radcliffe, University of California, Berkeley
1984
A carbon spring foot
Van Phillips lost his left leg just above the ankle in a water-skiing accident in 1976. After studying biomedical engineering at Northwestern University he shaped a foot after the curved hind leg of a cheetah and, with the aerospace materials engineer Dale Abildskov, built it from carbon graphite: when weight comes onto the heel the foot bends like a spring and gives the energy back at push-off. He founded Flex-Foot, Inc. in 1984 and sold it to Össur of Iceland in 2000. Lemelson-MIT says the Flex-Foot design or a variation of it is used by some 90% of Paralympic athletes.
Credit
Van Phillips and Dale Abildskov
2012
Hooks unchanged in 24 years
Gerwin Smit, Dick Plettenburg and colleagues at Delft University of Technology tested nine adult voluntary-opening hooks and hands on a bench, measuring the cable force needed to open each one, the pinch force it gave, its mass and its opening. Hooks needed less force and pinched harder than hands; several devices needed very high forces, all the hands pinched too weakly, and nothing had improved since comparable data from 1987. In 2018 Mona Hichert, Alistair Vardy and Plettenburg found that 12 of 23 people with a below-elbow limb difference could not pull hard enough to work 8 of 10 body-powered prostheses without tiring.
Lessonwithout published bench numbers, body-powered hands went decades without measurable progress.
Credit
G. Smit, R. M. Bongers, C. K. van der Sluis and D. H. Plettenburg
2020
Who must prove the advantage
Blake Leeper, a US 400 m runner with both legs amputated below the knee, asked the IAAF (now World Athletics) in July 2019 to let him race on carbon-fibre running blades against athletes without prostheses. Its rule made the athlete prove the blades gave no overall advantage. On 23 October 2020 the Court of Arbitration for Sport (CAS) ruled that burden unlawful, but found that the IAAF had shown his blades, which made him run "unnaturally tall", did give an advantage. In April 2021 a World Athletics panel, now carrying the burden, refused blades that set him about 5 cm lower, using the maximum allowable standing height (MASH) formula from para athletics; Leeper argued the formula had no body data from Black athletes, and scientists who gave evidence for each side still disagree in print.
Lessonwhen the science is disputed, the rule on who must prove an advantage decides the case.
Credit
the CAS panel in Leeper v IAAF; the World Athletics Mechanical Aids Review Panel chaired by David Grace
Watch for
A sole that wraps obstacles
The SoftFoot Pro is a foot whose flexible sole passively shapes itself over bumps and stones, widening the area in contact with the ground. Its makers report that in two people with below-knee amputations it lowered the torque and power needed at the other knee and at both hips while standing on the foot, against a standard carbon-fibre foot, on level and uneven ground.
Achieved whena group outside the makers repeats the result in a larger trial, or the foot goes on sale and is tested independently.
Claimed by
Anna Pace, Manuel G. Catalano and colleagues at the Istituto Italiano di Tecnologia and the University of Pisa, January 2026
Checking feet before reuse
Feet from rich countries are often replaced because of guidelines or expired warranties rather than damage, and many are donated to low- and middle-income countries. Michael Berthaume of King's College London and the UK charity STAND compared 196 donated feet received in Fort Portal, Uganda, before STAND brought in a quality checklist with 170 received after: usable feet rose from 83.16% to 94.12%. The authors propose the checklist as a first step towards international standards for reusing feet.
Achieved whena standards body or the World Health Organization adopts rules for reusing prosthetic feet, or another donor network repeats the result.
Claimed by
Michael A. Berthaume (King's College London) and STAND, May 2026
Open
01
Bench checks for used feet
A bench test that shows whether a used prosthetic foot still behaves like a new one of the same model: a jig that loads the heel and the forefoot at fixed angles and records how far each bends, a photo station that scores sole wear on a grid, and a pass-or-fail rule. Feet arrive with labels removed and nothing recorded about who wore them.
The briefCredit · unclaimed02
Gloves measured across makers
A public table of what cosmetic gloves take from a hand. A cosmetic glove is the skin-coloured cover pulled over a prosthetic hand, and its stiffness works against every closing of the fingers. For each glove, record how stiff it is at each finger joint, how much energy it soaks up in one close-and-open cycle, and how much pinch force is left when it is fitted over two reference hands, one body-powered and one motor-driven. Everything is measured on a bench, with nobody wearing a glove.
The briefCredit · unclaimed
Open 01 · the brief
Bench checks for used feet
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A bench test that shows whether a used prosthetic foot still behaves like a new one of the same model: a jig that loads the heel and the forefoot at fixed angles and records how far each bends, a photo station that scores sole wear on a grid, and a pass-or-fail rule. Feet arrive with labels removed and nothing recorded about who wore them.
Why it is still open
In the 2026 study of STAND's checklist, donated feet were inspected by eye and bent and pressed by hand, "no mechanical tests were run to investigate structural safety", and the authors write that there are no standards or regulatory requirements for used prosthetic parts. A second 2026 study scored sole wear on photos of 62 used SACH feet from Sri Lanka by eye, with fair to excellent agreement between raters, and measured no stiffness.
Built with what exists
a bench press frame or universal test machine with a 2 kN load cell and a displacement sensor; heel and forefoot loading plates; a camera stand and the 20 by 10 wear grid published by Berthaume's group; optionally a vision model trained to score wear from photos of the feet; one new foot of each model as the reference.
First step
Ask a donation programme such as STAND for 20 retired feet of one common model, buy one new foot of the same model, and record heel and forefoot load-deflection curves for all 21.
Done looks like
100 donated feet of at least five models, each with its checklist result, heel and forefoot stiffness against a new foot of the same model, and a wear score; a published table of how often the checklist and the bench disagree; a pass rule, such as stiffness within 15% of new, tested on a second batch of 50 feet.
Open 02 · the brief
Gloves measured across makers
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public table of what cosmetic gloves take from a hand. A cosmetic glove is the skin-coloured cover pulled over a prosthetic hand, and its stiffness works against every closing of the fingers. For each glove, record how stiff it is at each finger joint, how much energy it soaks up in one close-and-open cycle, and how much pinch force is left when it is fitted over two reference hands, one body-powered and one motor-driven. Everything is measured on a bench, with nobody wearing a glove.
Why it is still open
Gerwin Smit's 2013 Delft University of Technology thesis says that data on cosmetic gloves "is scarce" in the literature. His own tests used one PVC set and one silicone set "made by the same manufacturer" in identical size and shape, and showed how much the choice matters: the silicone glove was 2.5 to 4.5 times less stiff and needed 1.8 to 3.8 times less energy to move. A web search found no comparison of gloves across makers.
Built with what exists
The Delft finger-frame method, a one-joint frame placed inside a glove finger and bent through 90 degrees by a cable over a pulley while force and travel are recorded; a linear actuator; an S-beam load cell read through an HX711 board by an Arduino; gloves bought from hand makers and from independent glove makers; a body-powered hand and a myoelectric hand that a prosthetics school already owns, with a pinch-force sensor between thumb and finger.
First step
Build the finger frame from the thesis description and repeat its PVC-against-silicone result on one pair of gloves.
Done looks like
At least 12 glove models from at least four sources, three samples of each, with joint stiffness, energy lost per cycle and pinch force left on both reference hands; repeat measurements of the same glove within 5%; methods and raw data published.