On a Miami Beach in 1948, Norman Woodland drew concentric circles in the sand and extended Morse code into two dimensions. The idea was immediate and obvious to him: dots and dashes arranged radially, readable from any angle by a light source, encoding product information in a form machines could parse faster than any human. He and Bernard Silver filed for a patent the following year. It would take twenty-six years before a scanner read a barcode at a retail checkout.
The 1949 Patent and Its Twenty-Year Sleep
U.S. Patent 2,612,994, granted October 7, 1952, described a "classifying apparatus and method" using circular bull's-eye symbols and a reader based on the same technology used to read optical film soundtracks. The invention was technically sound. The problem was the technology available to implement it: the light sources and photodetectors of 1952 were too slow, too expensive, and too unreliable for retail deployment at scale. Woodland and Silver's employer, Drexel Institute of Technology, showed little interest. IBM hired Woodland in 1951. The patent sat.
In 1962, Philco purchased the patent. In 1971, RCA—which had licensed it from Philco—began serious development of the bull's-eye symbol as a potential retail standard, driven by pressure from grocery chains desperate to automate checkout. The National Association of Food Chains had been studying checkout automation since the late 1960s. Labor costs were rising. Checkout lines were the bottleneck in every store. The industry needed a machine-readable symbol on every item in a supermarket.
The IBM Intervention and the Linear Format
IBM assigned the problem to George Laurer, an engineer in Research Triangle Park, North Carolina. Laurer's analysis concluded that the bull's-eye symbol—Woodland's original concentric-circle design—had a fundamental flaw: print distortion. If a circular symbol smeared slightly in any direction during printing, the ring spacing changed and the reader failed. Linear barcodes, by contrast, could tolerate distortion in one axis entirely: variation along the bar's length didn't matter, only variation in bar width did. This made linear symbols printable on the flexographic presses that packaged goods used.
Laurer developed the Universal Product Code symbol—ten digits encoded in alternating light and dark bars using a specific encoding table (later called UPC-A)—between 1971 and 1973. The symbol was designed to be omnidirectional when scanned with a cross-hatch or raster laser pattern, solving the orientation problem that had motivated Woodland's circular design differently: not by making the symbol orientation-independent, but by making the scanner orientation-independent.
The Ad Hoc Committee on a Uniform Grocery Product Identification Code, convened by the grocery industry, evaluated proposals from multiple companies in 1973. Seven companies submitted symbol proposals. The committee selected Laurer's UPC symbol in April 1973, establishing the standard that would eventually be printed on virtually every packaged retail product in the world.
June 26, 1974: Wrigley's Gum in Troy, Ohio
The first retail barcode scan occurred at a Marsh Supermarket in Troy, Ohio, at 8:01 a.m. on June 26, 1974. The item was a 10-pack of Wrigley's Juicy Fruit chewing gum, purchased by Clyde Dawson, the store's head of research and development. The price was 67 cents. The receipt and the pack of gum are now in the Smithsonian Institution's National Museum of American History.
The scanner was built by NCR. The IBM 3660 Supermarket System, which processed the transaction, occupied a room. The entire installation cost approximately $125,000 in 1974 dollars. There were 26 stores in the United States with barcode scanners at the end of 1974. By 1980, there were 8,000. By 1984, 33,000. By 1987, a third of all U.S. supermarket stores had scanners.
The Downstream Transformation
The consequences of the barcode reached far beyond checkout speed. Point-of-sale scanning generated inventory data that had never existed before: real-time, item-level records of exactly what was selling, where, and when. Retailers could manage inventory without physical counts. Manufacturers could see which products were moving and which were not, by store, by region, by hour. The barcode was the first sensor deployed at scale across the supply chain.
Walmart's use of barcode data in the 1980s and 1990s—feeding scanner sales data directly to suppliers and optimizing replenishment automatically—was not a feature of Walmart's management genius alone. It was built on the barcode's existence as a universal product identifier. Without UPC standardization, the data wouldn't have been comparable across suppliers and stores. The barcode created the data substrate that just-in-time retail logistics required.
Three Observations
Norman Woodland drew the bull's-eye in the sand in 1948. George Laurer's linear format was chosen over it in 1973. The commercial scan happened in 1974. Woodland lived until 2012 and received the National Medal of Technology in 1992. He spent most of his career at IBM working on other problems. The invention that defined modern retail was not the focus of his career—it was a weekend idea he drew in the sand and patented, then waited a generation to see deployed.
The twenty-six-year gap between patent and commercial use was not unusual for enabling technology. The gap was not caused by obscurity or lack of interest—it was caused by the absence of cheap lasers and reliable photodetectors. The barcode required solid-state laser technology to implement at retail scale, and that technology didn't exist in 1952. The invention was correct. The implementation environment wasn't ready.
The battle between the bull's-eye and the linear format illustrates a pattern that recurs in standardization: the technically superior invention doesn't always win. The linear UPC was chosen not because it was a better symbol in the abstract but because it was more compatible with the printing technology that the manufacturing industry already used. The barcode that won was the one that worked with the existing infrastructure, not the one that was theoretically more elegant.
More from Anethoth: anethoth.com · builds.anethoth.com