The Claim and Its Caveats
Unimate is very often called the first industrial robot. The statement is reasonable if you mean the first robot sold as a commercial product for factory work, and less reasonable if you mean the first programmable machine of any kind. Earlier inventors had built automata and remote manipulators, and other people in the 1950s were working on programmable machines too. What sets Unimate apart is that it moved from idea to a company and to working installations in industry. This article follows that human and business story. How the arm itself works is the subject of a separate article.
George Devol and the Idea
George Devol was an American inventor who held many patents. In the early 1950s he began to work on the idea of a general-purpose machine that could be taught a motion and then repeat it. He filed a patent application in 1954 for what he called programmed article transfer, and the patent was granted in 1961. The long gap between filing and grant is typical of complex patents, and it means that dates for “the invention” vary by source, depending on whether they count the filing, the prototype or the grant.
Devol’s idea combined a method of recording motions with a manipulator that could repeat them. The concept was ahead of its time in one respect: much of the supporting technology, such as small, low-cost computers, did not yet exist, so the first machines relied on tools that were available then, including hydraulic drives and a magnetic drum for memory.
Joseph Engelberger and the Company
In the mid-1950s, Devol met Joseph Engelberger, an engineer and businessman with an interest in automation and a keen interest in science fiction, including Isaac Asimov’s robot stories. Engelberger saw commercial potential and helped create the company Unimation, short for universal automation, in the late 1950s. Engelberger became its public face and is often called the father of industrial robotics, a title that reflects his work promoting and selling the technology as much as designing it.
Selling the first robots was hard. Factory managers were cautious, the machines were costly, and few companies had experience with this kind of equipment. The new firm needed investors with patience and customers willing to experiment.
The First Installation
A Unimate machine was installed in 1961 at a General Motors plant in Ewing Township, New Jersey. As commonly told, it unloaded hot die-cast metal parts from a casting machine and stacked them, work that exposed people to heat and fumes. The specific part and process are described somewhat differently in different retellings, so it is safest to say that the robot handled die-cast components in a hot, hazardous setting. The significance was less the task itself than the proof that a programmable machine could work reliably in a production line, shift after shift.
General Motors then used the robots in other tasks, notably spot welding of car bodies, and automotive welding became one of the biggest applications through the 1960s. Companies in other countries, including in Japan, began licensing or developing their own designs; Japanese firms later became major producers of industrial robots.
Public Attention
Unimate also became a media figure. It appeared on American television in the 1960s, and there are accounts of demonstrations in which it performed playful tasks. These appearances helped create the public picture of a robot as a mechanical worker, even though the real machine was a stationary arm attached to a factory fixture, nothing like the humanoid of fiction.
What Happened Afterward
Unimation grew through the 1960s and 1970s but faced competition from new entrants, particularly those offering electric arms with better control and the growing use of embedded electronic computers. In the 1980s the company was acquired by Westinghouse, and its robot lines later passed to other firms. The Unimate name faded, but the design lessons stayed in industry.
The story shows how a technical invention becomes an industry. It needed an inventor with a concept, a salesperson with persistence, a large customer in the car business, and years of patience. It also shows why the “first” label should be used carefully, since a pioneering product usually depends on earlier ideas and is soon overtaken by better ones. Modern descendants, from welding cells to service robots in warehouses, trace a line back to those early installations, though no single machine should be given all the credit.