This month we talk about Sunrise 2027, matrix codes, and KarTrak.
We also discuss IBM, Whiskas, and QR codes.
—
Transcript
Barcodes were patented back in the early 1950s by a couple of inventors, Norman Woodland and Bernard Silver, who wanted to create a visual means of storing data that could be quickly read by a scanner.
The first functional version they came up with was based on Morse code, using thick and thin lines, arranged in a horizontal line, though they also patented a bullseye-shaped version, alongside scanner technology for reading both linear and circular options.
The patented scanner was based on technology already being used in movie theaters to parse information displayed in film that communicated the soundtrack for the movie being played; this technology essentially allowed photons to be converted into electrical signals, and it used a light bulb instead of a projector bulb to convert the barcode lines into the same.
One of the investors took a job at IBM while the patent application was being assessed and eventually convinced the company to consider developing the system—they ultimately declined, saying the tech required to process the resulting electronic information from these scanners was too far off in the future to invest in quite yet, but they did offer to buy the patent, and had that offer rejected by the inventor duo; they thought the patent was worth more than was offered.
A decade after the patent was issued, Philadelphia Battery Company, usually called Philco, which was a big name in radio technology at the time, and it has since been purchased by Ford, then GTE, then Philips—though outside North America Philco is owned by Electrolux—Philco scooped up the patent on these barcode scanning technologies, before then selling it to RCA several years later.
This technology was originally developed by Woodland and Silver for use by a local food chain called Food Fair, the owner of which was keen to speed things up during checkout by using such a system, rather than doing everything by hand.
Interestingly, though, considering that at the grocery store and other points of sale is where most of us are most likely to encounter barcodes and scanners today, the first successful application of the technology was actually in the railroad industry, not at the supermarket.
A man working at GTE Sylvania, which made all sorts of electrical components, David Collins, started focusing on a problem he noted while working an earlier job at the Pennsylvania Railroad: you have to quickly and automatically identify railroad cars as they pass, and the existing solutions were all flawed in many fundamental and costly ways.
He came up with a system called KarTrak which used red, white, and blue reflective stripes on the side of each car to encode a car number and company identification number, and those stripes reflected light that was captured by the same sort of film-industry device utilized by Woodland and Silver for their scanning technology, called a photomultiplier.
It took nearly a decade from when he started working on the problem, but eventually this KarTrak system was selected by the Association of American Railroads as their new standard for such things, and they began to install it across their entire continental fleet, but that installation wasn't 95% complete until a decade later because of an economic downturn that sparked a series of bankruptcies and consolidations in the industry.
A toll bridge in New Jersey ordered a similar system so it could scan cars that purchased a monthly pass, and the US Post Office wanted a version that would allow them to track their trucks, and the KarTrack folks obliged in both cases.
The owner of Whiskas-brand cat food then asked the folks behind KarTrack for a version of this system that could be affixed to crates of pet food to help them track inventory and shipping.
This led Collins to start looking into making a black and white version, rather than a colorful, reflective version, of the system, and he asked his bosses permission to do so, but they said no, they've got their hands full with the railway and other vehicle side of things, and didn't really see the potential of the barcodes on shipped consumables concept.
So he quit and started up a company called Computer Identics Corporation, jazzed-up the scanning technology so it used lasers and mirrors instead of lightbulbs, which among other things made the scanners easier and more reliable to use, even if a barcode is damaged, and the first of these snazzy new scanning systems was installed at a General Motors factory in Flint, Michigan, in 1969.
Three years before that, the National Association of Food Chains, on the lookout for automated checkout systems, was pitched a system based on the Woodland and Silver patent by RCA, which owned the patent at that point, and the Kroger grocery store chain volunteer to serve as a testing ground for the bullseye-shaped option.
By the mid-1970s, that testing had led to the formation of a committee to come up with a Uniform Grocery Product Code, which among other things would choose a symbol and figure out what information that symbol would encode.
They opted for an 11-digit code that would identify products, and a bunch of styles, pitched by companies ranging from Singer to IBM to RCA submitted their ideas, but while RCA was able to conduct a larger test at a big Kroger store for a year and a half, it was found that their bullseye option would sometimes smear when printed out, which often left the codes unreadable by scanners.
IBM was keen to compete with RCA, and they realized, oh yeah, one of the guys who filed that patent RCA owns still works for us, so they reconnected with him and had him come up with a new style of the linear bar-code that he'd invented all those years previous—the bars being preferable in real life because smears usually just lengthened the lines or messed up a small part of a few lines, which left the rest of the bar readable by the scanners.
In 1973, this IBM barcode was selected as the standard for the National Association of Food Chains, and the first-ever product purchased in a store that was using this standard in the wild was a ten-pack of Wrigley's Juicy Fruit gum in June 1974—that pack of gum and the receipt for it are on display at the Smithsonian Institution, today.
Early on, it was evident that people bought more stuff and went to the grocery store more frequently after scanners were installed, leading to a 10-12% increase in sales. There was a 1-2% decrease in operational costs for early adopters of the technology, and it was estimated that stores that invested in them in those early days enjoyed a return on investment of something like 41.5%; those glowing figures led to a lot more adoption of this technology across the US, then around the world.
Variations of the barcode concept have since been adopted across other industries, as well, with pretty much every consumer-oriented industry using some kind of barcode system for sales and inventory purposes, but even militaries use these sorts of automated scanning systems to keep track of what's where, what should be treated in what way, and what should be sent to which base.
What I'd like to talk about today is the next step in barcode evolution, what's driving this change, and what it might mean for those of us who encounter these things so often in the modern world.
—
Sunrise 2027 is the name being used for a moment in which point our currently most common type of bar code reader used at point-of-sale locations—where you buy things in stores, basically, like the check out line at a grocery store—will largely be swapped out for a different type of barcode reader, capable of scanning what's called a 2D barcode or a matrix code.
If you've ever seen or used a QR code, short for Quick Response code, that's a type of 2d or matrix barcode.
Instead of containing readable, scannable information using lines arranged in just one direction, length, let's say, 2D barcodes are represented by information spanning both length and height, often using squares or dots.
There's a nonprofit standards organization called GS1 US that manages the UPC barcode standard, and that's the organization that's managing this transition, as well.
And it'll be quite the transition—there are a lot of retail entities in the world, and the idea is by 2027 all those entities should be switched over to 2D barcodes, and should have all their products and such updated to have these codes printed on them alongside or instead of today's 1-dimensional versions, as well—but despite all that work, it'll theoretically be worth it.
This transitions isn't being done just for kicks: a standard UPC code only contains 12 digits, those digits encoding information about the product—so the first digit might say something about the type of product, the second digit the color, and so on; these digits can be used for just about anything, but there are only so many of them, so the use-cases are quite finite.
In contrast, a GS1 DataMatrix 2D barcode, for instance, has a capacity of 3,116 numbers or 2,335 alphanumeric figures—letters and such, all of that contained in a small little square barcode.
The GS1 QR code, the one many of us have become familiar with over the course of the COVID pandemic, as they became common in touchless environments, like those that wanted us to look at restaurant menus online instead of everyone touching the same potentially germ-carrying physical menu, those QR codes can contain 7,089 numbers or 4,296 alphanumeric characters; and both of these types of 2D barcode can also include thousands of binary and kanji, pictographic Japanese, Chinese, etc, characters.
Again, that's compared to 12 numbers in the current UPC code; quite a change.
That change is meant to allow the producers of these products to encode just a boggling amount of additional information and resources in these little squares.
So you bring a box of Mac and cheese up to the checkout, and the code can be scanned by the checkout person, or by you, if you're self-checking-out, but you can also scan it with your phone or an in-store device to pay for it, or to see where it was made, if there are any recalls for that brand or flavor, how close it is to the sell-by date, information about ingredients, allergens, additional recipes for which you might use the contents of the box—things like that.
You might also slap a little 2D barcode sticker on a piece of fruit to see that all-important sell-by information and where it was grown, maybe something about the chain of custody for it, where it was shipped and then shipped again, information about pesticides used, and maybe information about how to tell when that sort of fruit is ripe and ideal for eating.
You'll be able to scan these codes like we scan them today to go online: embedding a website and code to open a web browser uses relatively few of those thousands of characters you can embed in these 2D barcodes.
Similarly, you can give folks access to limited edition, special content if they buy something and then scan the QR code printed on the box or affixed to the product itself—maybe additional in-game items that you add to your account by scanning a code on a game box, maybe a whole mini-game that comes with a pair of shoes.
Puma is the first major brand to have made this transition, so they might include Puma-brand digital shoes for your online avatar or the ability to register or resell or repair your products, if you scan this little code after making a purchase. They're already including information about the company's sustainability efforts via these codes, and info about the materials used in the shoes you've just bought.
If you're not keen on any of this extra info and content, nothing will really change for you: the codes on the stuff you buy will slowly, over the next few years, be replaced with little code squares instead of bars, but you don't have to do anything different: the Sunrise 2027 effort is meant to make this transition as smooth as possible, serving as a timeline for retailers, not for you—all those potential bonuses are just interesting ways companies might use that additional character space in these codes, and they'll operate the same as the bar codes in terms of checking out and buying things.
We're at a turning point for this transition, though, because GS1 US has released a capabilities test kit for retailers that will allow them to begin making changes to their products and packaging, helping them segue into this new reality and test their adjustments. And in most cases, most companies will include multiple codes on their packaging for a while—and this test kit shows them how to do that, alongside advice on how to use the large amount of new character space to which they now have access.
This runway to 2027 is also meant to allow time for producers and retailers to rethink their shipping and inventory systems, considering this new data-rich reality.
The hope is that things will become more trackable and legible, with more information about where something began, where it arrived and when, and what happened to it in between: creating a smarter supply chain, and then presenting these goods in such a way that there will be less waste, less contamination, and so that it will be less expensive when something goes wrong and requires some kind of recall—it'll be a lot simpler to pull impacted or unsafe products, because things will be easier to casually parse and process.
That's a big ask, though, and 2027, though it seems far off, is just around the corner in terms of retail world timelines.
There's a lot left to do, and there's a nonzero chance we'll see these things all over the place in the next year or two, but it'll be a while beyond 2027 before the full potential of this standard is discovered and fully realized, and in the meantime, there's some evidence that existing QR codes—the ones we often have to scan to read restaurant menus, these days—are becoming less popular, the ones that are affixed to every conceivable surface seeing less use.
So there might be a second uptake-period sometime in the next year or two, meant to get people used to using these little code-squares again, in addition to seeing them all over the place.
—
Show Notes
https://hackaday.com/2023/05/03/barcodes-enter-the-matrix-in-2027/
https://www.axios.com/2023/04/17/2d-barcode-transition-2027
https://www.gs1us.org/industries-and-insights/by-topic/sunrise-2027
https://www.gs1.org/standards/barcodes/2d
https://www.gs1.org/standards/barcodes-epcrfid-id-keys/gs1-general-specifications
https://www.gtin.info/barcode-101/
https://www.nytimes.com/2023/05/22/dining/restaurant-qr-code-menu.html
https://web.archive.org/web/20161016084435/http://www.bar-code.com/upc/bar_code_history.php
