Barcode Does not Scan—What Could Be Worse?

Q: What could be worse than a non-scanning barcode? The answer might surprise you!

A: A barcode the scans the wrong information perfectly.

Here is a real train wreck, a case history we personally witnessed.

Jamie Street on Unsplash

When Vanilla is the new Lemon

A few years ago, we bought some vanilla yogurt along with about $150 of other stuff—a typical family grocery shopping trip. Nothing unusual happened at the register but I noticed that the receipt tape called out lemon yogurt. I thought nothing of it at the time but thought of it again when the yogurt was eaten and the empty container was in my hand about to be discarded. Instead I brought it to the test lab and sure enough, it read perfectly but the encoded information was different that the numbers beneath the barcode.

Next trip to the grocery store I questioned the store manager and yes, there had been a major inventory glitch. They kept ordering too much lemon yogurt but were chronically out of vanilla. It took them a while to figure it out; meanwhile, customers were not happy.

There are several ways this could happen:

• Graphics design software does not prevent errors when inserting a barcode file in the FPO. Te technician could simply insert the wrong file.
• Super-aggressive scanners can misread a marginal quality barcode, transposing or substituting a character or two
• Database entry error can cause an erroneous look-up at the register. This could cause a no-read or a look-up to the wrong product

Complete failure is better than a Technical Mistake

Tyler Nix on Unsplash

Whatever the cause, it is a nightmare. That comforting “beep” at the checkout is really no comfort at all: the scanner recognized something—was it the right thing? It would be far better if the barcode failed altogether. That mesmerizing beep lulls us into a warm, comfortable acceptance that all is well, when actually a train wreck is unfolding before our very eyes.

Barcodes that misbehave in this way are difficult to detect. How can these problems be identified? Here are some ideas:

• Guard databases diligently.
• Making central databases accessible to many users makes them vulnerable to mistakes. If this is necessary, designate a super-user to guard data integrity and check it regularly.
• Minimize the use and distribution of multiple database copies. If this is necessary, sync them regularly to a known “golden” copy to avoid inaccuracies
• Validate product barcodes to the ”golden” database. Some barcode verifiers have product lookup capability to perform validation.
• Pay attention to apparent inventory anomalies such as spikes, drop-offs or unexplained supply chain errors. This could signal a barcode mis-identification
• Be aware that the comforting “beep” at checkout only means that something happened. It does not necessarily signal what you think (or hope) happened.

Does this indict barcode technology and predict its demise in favor of some new, better identification strategy such as RFID or invisible barcodes? Actually not. Alternative auto ID technologies have the same reliance on databases, which is where these errors can originate.

Vision-based product recognition could avoid these problems, but that solution is outweighed by the new problems it cannot solve, such as batch or lot based recalls, best-use-by and expiration dating. Barcodes are not perfect but they are still the best, most accurate and least expensive solution for what they do.