Upon closer look, Micah Forstein, an assistant manager at the company, realized that the arm was a remnant—a prototype of an invention that had changed joint replacement surgery forever.

Called the Robodoc, the innovative robotic system allows surgeons to perform complicated hip and knee surgeries with greater precision using CT scans converted into three-dimensional virtual images for preoperative planning and computer-guided drilling. The tool has been used in more than 28,000 procedures worldwide.

Now, the fully recovered 1989 prototype will be forever memorialized in the collections of the Smithsonian’s National Museum of American History.

“It’s important for us to remember milestones in medical technology,” says Forstein.

The robot is the brainchild of the late veterinarian Howard “Hap” A. Paul and engineer-turned-orthopedic surgeon William Bargar, who were both working at the University of California, Davis, in the 1980s when Bargar recognized what he calls a dilemma in total hip arthroplasty, or hip replacement surgery.

In that era, implants were attached to the patient’s body with acrylic cement, an impermanent material that would eventually break down, sending the patient back under the knife.

Researchers had already attempted to eliminate the need for the faulty cement by using porous implants in which the bone could actually grow. This development addressed the problem of the deteriorating cement, but the implants were still imperfect because they were only manufactured in a few different sizes; they didn’t fit every patient’s body.

“You’d try to put them in and some would fit too tight,” says Bargar, “or you’d break the bone putting it in, or some would fit too loose and it would wiggle, so it was hard to get the right size for every patient. So I had the idea to custom make these things.”

Using a patient’s CT scan data along with computer-assisted design/computer-assisted manufacturing (CAD/CAM) technology, Bargar could design an implant to fit a specific individual’s body. He could then transfer the design to a CAD/CAM machine that would cut the implant out of metal.

Meanwhile, on a different part of the UC Davis campus, Paul was studying joint replacement surgeries in dogs as a resident at the School of Veterinary Medicine. He couldn’t bear to put a dog down for joint issues if there were alternatives, such as hip replacements. The two researchers joined forces, pursuing custom implant research in dogs.

But even with the custom implants, joint replacement surgery was flawed. While a machine made the custom implants, surgeons were still digging the cavities in patients’ bones by hand, often crudely, presenting obstacles for the insertion of the implants and paving the way for harmful consequences, such as bone splintering.

On a flight home from Nice, France, where they had presented their research on the custom implants, Bargar and Paul came up with their next idea: to use a robot to cut the inverse shape of the implant in the patient for a perfect fit.

But the research fields of robotics and computers had developed independently of each other, and teaching a robot to act like a CAD/CAM machine was a new concept. After calls to many robotics manufacturers left the researchers at a dead end, Bargar’s father, a former IBM employee, put him in touch with a group at IBM’s Thomas J. Watson Research Lab. There, researchers had developed an automated machine language but had yet to apply it in the real world.