Orthopedic Implant Manufacturing: Charnley to ABB Robots

Every artificial hip or knee starts as a bar of titanium, a cobalt chrome casting or a block of plastic, and ends as a part polished to a mirror finish. Orthopedic implant manufacturing has moved from a surgeon's home lathe to CNC machining centres, metal 3D printers and robot finishing cells. This post covers how the first hip implants were made by hand, when robots first took over the grinding and polishing, and where a new compact ABB robot such as the IRB 1700 could fit. Our posts on hip replacement surgery and knee surgery cover the operation itself.

Titanium alloy hip stems, a metal knee component and a plastic insert made by Sulzer Orthopedics in the United States, 1998. Photo: Wellcome Collection via Wikimedia Commons (CC BY 4.0)

How the first hip and knee implants were made by hand

Early joint replacements were made in tiny numbers, often by the surgeon or a local workshop. In 1949 the French surgeon Jean Gosset designed a hip ball made of Perspex, an acrylic plastic, and a Paris maker was still producing modified versions in the late 1950s. Acrylic and early metal parts often wore out or loosened. Each part was shaped on lathes and by hand, then checked by eye and with simple gauges. Nobody yet had good data on how long an implant would last inside a patient, and designs varied from one surgeon to the next.

A Perspex hip ball made in Paris around 1958 to 1960, a modified version of Jean Gosset's 1949 design. Photo: Wellcome Collection via Wikimedia Commons (CC BY 4.0)

The English surgeon John Charnley changed that at Wrightington Hospital, where his biomechanical laboratory opened on 23 June 1961. His first sockets were made of PTFE, which wore quickly and caused tissue reactions. In November 1962 he implanted his first socket of ultra high molecular weight polyethylene, a plastic still widely used for hip sockets. Chas. F. Thackray Ltd of Leeds made his stainless steel stems, while Charnley turned the plastic sockets himself on a lathe in his home workshop. That lathe is now on display at the Thackray Museum of Medicine in Leeds.

The Thackray Museum of Medicine in Leeds, which displays a lathe Charnley used to make hip sockets. Photo: Creamycanarie98 / Wikimedia Commons (CC0)

As hip and knee replacement grew into routine surgery, companies moved to cast and forged cobalt chrome, titanium alloys and machined polyethylene. Finishing remained the slowest step, because the bearing surfaces must be very smooth. In 1998 the journal Industrial Robot described one of the early robot cells for this job. Built for a large implant maker, it used a FANUC S10 robot to pass cobalt chrome knee joints over eight abrasive belts, from 240 grit up to a greased 600 grit belt. Finishing time per knee dropped from about 30 minutes to 10.

NASA researchers Robert Johnson and Donald Buckley at an artificial hip simulator testing bearing alloys, 1966. Photo: NASA Glenn Research Center via Wikimedia Commons (public domain)

CNC machining, 3D printing and robot finishing today

Most implants today are machined on multi axis CNC centres, either from bar stock or from near net castings and forgings. Metal 3D printing is used as well. Laser and electron beam powder bed printers build titanium parts with porous surfaces that bone can grow into, which would be very hard to machine. After printing or machining, parts are still ground, polished, cleaned and inspected. Skilled hand finishers still do many of those steps, and robots are now moving into that work, for example at a contract maker in Limerick, Ireland.

A DMG Mori DMU 50 eVolution five axis CNC machining centre, the type of machine used for precision metal parts, 2005. Photo: Glenn McKechnie / Wikimedia Commons (CC BY-SA 2.0)

Croom Medical, founded in 1984, makes cobalt chrome knee implants for other companies. According to FerRobotics, hand grinding and polishing there was slow, specialised and exposed workers to vibration, dust and repetitive postures. Earlier attempts with other suppliers' systems did not reach the required finish. The working solution uses a robot arm to guide each implant against FerRobotics' Active Belt Grinder, which keeps the contact force constant. The company reports higher throughput and repeatable results, though it gives no figures, and the case study does not name the robot brand.

A Renishaw AM 400 metal powder bed printer at the opening of the 3DMetal lab at the University of Pavia, 2018. Photo: Università di Pavia / Wikimedia Commons (CC BY 2.0)

On 12 October 2026 the Association for Advancing Automation (A3) reported that ABB Robotics had launched the IRB 1700 and IRB 1700ID. The family covers payloads from 6 to 12 kg in five variants and runs on the OmniCore controller. ABB quotes path repeatability down to 0.03 mm, up to 20 percent shorter cycle times and up to 15 percent lower power use than the IRB 1600, and lists deburring, sanding and machine tending among its tasks. ABB does not mention implants, so this use is our suggestion. A small, accurate arm that can tend a CNC machine and then sand or deburr is the kind of robot an implant line needs.

The EasyGrinder robot at Ancofer picks steel parts from pallets, grinds them and stacks them again. Photo: ABB

Where implant factories go next

Implants are moving toward more sizes and more patient specific parts, especially for complex cases, and 3D printing makes small batches practical. That puts pressure on finishing, because a robot cell has to handle many shapes without long reprogramming. Force controlled grinding, like the setup at Croom Medical, helps here, since the robot keeps steady pressure on a curved surface instead of following a fixed path. Programs for new part families are usually built and tested first in RobotStudio, where engineers can check reach and paths before cutting metal.

The new generation ABB IRB 1200, launched in 2025, a compact arm for machine tending and small parts. Photo: ABB

Cobots may take on finishing in smaller shops. ABB's GoFa CRB 15000 already sands curved panels with a Mirka sander in ABB's OmniVance finishing cell, and the same idea could carry over to polishing smaller implant parts. We found no reported project of that kind yet. Robots may also link machining, finishing and inspection into one flow, so parts move from the CNC machine to the polishing station and then to measurement without waiting in trays. For patients, the hoped for result is implants that are more consistent from one part to the next.

Different types of hip prosthesis on a wall display in the Wellcome Medicine Gallery at the Science Museum in London, 2022. Photo: FE2656BA / Wikimedia Commons (CC BY-SA 4.0)

Medical device rules mean every process change has to be validated, which makes factories slow to swap a proven manual step for a robot. Robot cells for polishing are expensive and need dust extraction and safe guarding. The hardest part may be skills. Experienced finishers judge a surface by eye and touch, and their knowledge has to be turned into force settings, paths and inspection limits before they retire. Patients waiting for a knee replacement or a new hip never see this side of the work, yet the finish on the bearing surface affects how long the implant lasts.

A GoFa cobot sands a curved wooden panel with a Mirka sander inside ABB's OmniVance finishing cell. Photo: ABB

Innovation

AI solutions for effortless ABB robot control.

Automation

Robotics

ceojohntran@chatwithrobot.net

+84905311611

© 2025. All rights reserved.

qtran1215@gmail.com