The reader is a separate problem
A barcode is only a pattern. To be useful, something has to find the pattern, convert the dark and light bars into an electrical signal, work out where each bar begins and ends, and turn the result into digits. The history of that equipment overlaps with the code’s history but is not the same story, and the machines took a long road before they became everyday furniture at the till.
Early attempts
The earliest scanning ideas date from the late 1940s and 1950s, when an electric eye, a light source and a photocell were the available tools. A reader of that kind could, in principle, tell dark from light, but it needed bright illumination, and the signal processing was done with vacuum tubes. Those parts were large and gave off heat, so the equipment looked more like a laboratory rig than something a cashier could use.
Through the 1960s, several companies experimented with machine-readable labels, including colored stripes for tracking railroad cars, and the same period saw growing interest in image scanners and other optical readers. Reports from that era suggest such systems worked best in controlled conditions and struggled with grime, glare and damaged labels. A supermarket checkout is a rougher environment than it appears: packages are curved, wet, crumpled and pushed past at odd angles.
Lasers change the picture
The ruby laser, first demonstrated in 1960, and the gas lasers that followed it offered a narrow, bright beam that stayed in focus over a useful distance. A small gas laser, paired with a spinning mirror or similar scanning element, could sweep a spot of light across a label many times a second. A photodetector measured how much light bounced back: bars absorbed it, spaces reflected it. This turned the problem into one of timing, because the width of each bar and space could be inferred from how long the returning signal stayed low or high.
Two things made that practical. One was the arrival of integrated circuits, which allowed the decoding electronics to shrink. The other was the growth of computing power, which let a store register look up a price instantly. Without both, a laser reader would have been a clever curiosity.
From trial to first checkout
By the early 1970s, retail trials were under way. Companies including RCA and others developed scanning systems for stores, and some used the bullseye pattern from the original patent. After the grocery industry agreed on a linear code, scanning equipment makers rushed to support it. NCR is usually named as the maker of the scanner used in the widely cited first retail scan in Troy, Ohio, in June 1974, though accounts of exactly which company supplied what vary, and several firms were involved in the broader rollout.
An important design choice shaped these early machines. A person scanning a label by hand is imperfect, so designers wanted a reader that did not require careful aim. The solution was a scan pattern made of multiple lines crossing in different directions, produced by rotating mirrors. A product passed over a window would have its code crossed by at least one beam, even if it was turned the wrong way.
Types of scanner
Over the next decades, several formats appeared:
- Slot or flatbed scanners are built into the counter and read items as they are dragged across a window. Later versions added a vertical window as well, sometimes described as bioptic, so items could be read from two sides.
- Wand or pen readers are touched to the label and moved across it. They are inexpensive, and they require a steady hand.
- Handheld laser guns let the user point at a label from a distance, which suited large or heavy items.
- CCD and imaging readers use a row of light sensors, or a full image sensor, in place of a moving laser spot. They have no moving parts, and the camera-based versions can read two-dimensional codes as well.
Light emitting diodes were also used as the light source in some readers, offering lower cost at the expense of range.
What changed afterward
Scanners spread beyond checkouts. Warehouses, delivery services, libraries and hospitals adopted them to track objects and to reduce hand-keyed entry, which is a common source of errors. Because the computing underneath became cheaper and smaller, the same scanning ability eventually moved into smartphone cameras, with software doing the decoding that custom electronics had once done.
It is tempting to summarize this as a straight line from laboratory to store, but the path was uneven. Early trials sometimes disappointed, labels were misprinted, stores had to be wired, and shoppers and staff had to adjust. The commercial result in the 1970s depended on a standard code, affordable hardware and employers willing to invest in both.
Credit and uncertainty
As with many inventions, no single person built the checkout scanner. Engineers at several companies worked on optics, mirrors, lasers and decoders at roughly the same time, and later retellings often compress this into a single heroic moment. The most defensible summary is that a combination of the laser, solid-state electronics and a shared product code arrived in the early 1970s, and that the first supermarket scans in 1974 marked the point where the equipment became commercial rather than experimental.