Forty years ago today, a very minor purchase took place. We tend to overlook the very small things that don’t really impact the balance sheet, and this particular purchase was about as trivial as they come: sixty-seven cents for a multipack of Juicy Fruit chewing gum. Clyde Dawson was the customer, and cashier Sharon Buchanan made the sale. Clyde obviously wasn’t a big fan of gum, as the purchase can now be seen in the Smithsonian in Washington.
Machine-readable information was nothing new: punched cards had been controlling textile looms since about 1725… but the technology wasn’t adapted into anything like the modern barcode for more than two centuries – not so much because the technology was unsuitable (although punched cards are a bit too expensive, and fragile) but because it was the supermarket that provided the business need to spur the development – and computers had some catching up to do, as well.
In 1948 a graduate student called Bernard Silver overheard a conversation with a grocery store president who wanted a way to read product information automatically at checkout. He got his friend Norman Woodland involved, and they conducted some experiments. Taking inspiration from the Morse code, Woodland reproduced information expressed in dots and dashes as narrow stripes and wide ones: the world’s first barcode was drawn in sand, on a beach. Woodland decided that a barcode would be more readable if it could be accessed from any angle, and thus wrapped his thin and thick stripes around, into concentric circles. Barcodes became ‘bullseyes’, but this proved to be a dead end. Eventually, a patent (2,612,994) was issued in October 1952, describing both circular and linear barcodes. By then, Woodland was working for IBM, but it was Philco that bought the rights.
What we generally refer to as a barcode is more properly called the Universal Product Code. Figuring out the technology for reliably reading the ‘dots and dashes’ off a label or package was one challenge, but establishing a standard for product identification was also necessary. The ‘number’ that is read from a UPC barcode is actually two separate numbers: a manufacturer’s ID number and an item number. There are common features that ordinary customers such as you and I don’t need to know about: like how a ‘2’ prefix denotes products where the price varies with the weight, while a ‘4’ prefix identifies barcodes reserved for the store’s own use, such as promotional coupons. For this foresight, we have to thank a group of grocery industry trade associations called the Uniform Grocery Product Code Council.
That all this was figured out in time for Clyde Dawson’s purchase back in 1974 isn’t particularly remarkable, perhaps. The Apollo space program had been and gone; Intel’s 8080 microprocessor was on sale… this was the information age. The machinery to get information at the point of sale (and do something useful with it) had arrived. What is far more remarkable is that forty years on, the fundamental concept of the UPC barcode hasn’t changed a bit. It’s still doing the job, beautifully, while a number of ‘successors’ have disappeared. (Microsoft’s ‘Tag’ will be discontinued next year, for example.)
There are many other barcode standards, of course. If you worked in a warehouse environment you probably used the MSI Barcode. If you wanted to send out machine-readable post, you probably know the up-and-down format called RM4SCC (in the UK) or the similar Intelligent Mail/ IM Barcode (in the US). If rolling stock was your thing, you might remember the colourful KarTrak (also discontinued).
Moving on a few years, we find a variety of two-dimensional barcodes. The term is not very informative, as the bars have largely disappeared, but the evolutionary leap is considerable. In place of the humble twelve digits offered by the UPC barcode, these new matrices can hold a vast amount of information… and the smartphone you already own is a perfectly good ‘reader’ device.
Originally developed by a Toyota subsidiary for use within the automotive industry, Quick Response (‘QR’) Codes seem to be the most promising choice among the current options, although some industries still enjoy success with alternative standards. Technically, these baffling patterns are a masterpiece: try damaging a label, and you’ll be impressed when you see what the error correction protocol can achieve. So far, so technically brilliant, but an intriguing aspect of these modern, open barcode standards is the democratisation that results: anybody can create a QR Code, whether they’re an advertising executive or a member of a protest movement: you don’t need to have groceries to sell, or cars to assemble. What could you do with one?
Among all the tracking and point of sale options available to us in 2014, including RFID, Bluetooth 4.1 and so on, the beauty of the barcode remains. Whether it’s the humble UPC version or one of its successors, what we have is an identifier that can be added to a product for next to nothing (most formats use a single colour), and at no cost at all if a generic code is simply included as part of the packaging’s design. This is read quickly and reliably with a low-cost device, and virtually everybody understands how to do it.
Even if all you ever see of the barcode is what happens at the supermarket checkout, it’s a brilliant little piece of common-sense technology. That we seldom notice them is evidence of how elegant the solution is. Happy birthday to the UPC barcode, and thanks to Bernard Silver and Norman Woodland… who only ever made $15,000 on the invention.