Scan Rate as a Basis for Barcode Verification

Recently we were asked to weigh in on an ongoing conversation between a label printer and their customer. The GS1 Databar symbols the printer was producing had an 8 mil X dimension which is necessary to comply with the GS1 guidelines. The customer wanted the printer to use a smaller X dimension to reduce the size and cost of the label. Wanting to be responsive to his customer but also to be GS1 compliant, the printer was exploring the possibility of reducing the label size by truncating the two rows of the Databar symbol, and wanted to know how that would affect scan rate. Familiar with the ISO specification for print quality, the printer wanted to know if there any way to extract or approximate scan rate from the verification report. What a great question.

First we must be clear about what the printer is really asking regarding “scan rate”.  He knew what he meant, but his question has more than one meaning depending upon who you ask: to a scanner manufacturer, “scan rate” means how many times (per second) the scanner attempts to  read the barcode; to a quality person, scan rate means how many times (per second) the scanner successfully captures the data in a barcode. It sounds nit-picky but one relates to the other. Consider the following:

Barcode Scanner Scan Rate

*Unit of measure = scans per second

**Scan rates for area images is not a simple number: it depends on field of view, resolution of the camera, scan lines per image and processing speed

Scanners that make more attempts per second to decode a barcode will achieve higher rates of data capture, but there are other variables that make it difficult if not impossible to predict data capture rate, which is the other meaning of the term “scan rate.” Those variables include:

• Symbol type: 1D or 2D. 1D barcodes will decode faster than 2D symbols
• Symbology:
  • QR Code will decode at a different rate than a Data Matrix Code
  • UPC will decode at a different rate than Code 128
  • A two or three layer stacked GS1 Databar will decode at a different rate than a many-layered PDF417
• Density: smaller X dimensions will decode more slowly than larger X dimensions
• Truncation: it takes more scanning attempts and therefore more time to find a complete path through a less-than-full-height barcode
• Amount of Data: 1000 bytes will take longer to decode than 100 bytes
  • Rate of motion and orientation of barcode on a transport or conveyor, or if the barcode is being moved over a counter-mounted scanner
• Quality of the barcode image: poorly reproduced images take longer to decode—and may fail to decode altogether
• Other hardware considerations: quality and condition (cleanliness, scratches, age) of optics, focusing or fixed focus, rastering or straight-line scan pattern (laser), electronic drift and fatigue, etc.
• Environmental variables: ambient light, dust or other air-borne particulate, distance and angle of scanner to barcode, etc.

Notice how the scan rate information in the table is all approximate. This is because all types of scanners will scan a barcode at least once, and often several times to detect reflectivity and adjust gain before actually decoding the symbol. A scanner with a 100 scan-per-second scan rate will not decode the barcode 100 times. The scan rate of linear and area imagers will vary due to data processing rates, so the theoretical scan rate here will likewise not correlate directly to data capture.

How does all of this aggregate into a concise statement about scan rate in its data capture meaning? It doesn’t.  There are so many variables and considerations it is not possible to predict the rate of data capture in an open system such as barcode scanning. Scan rate is itself a variable. While it is possible to calculate scan rate for a given scanning system, it is pointless to do so if barcodes, such as those produced by our inquiring printer, end up in any number of scanning situations and environments where the variables are, well, variable. This is also why a scanner is a poor substitute for a verifier.

The answer to the printer’s question is to comply with the applicable specifications, both ISO and GS1. The ISO specification covers the technical attributes of the printed symbol; the GS1 guidelines cover the practical attributes of the barcode in an open system such as variable weight produce labels.

Barcodes with smaller X dimensions will have a lower scan rate. Truncated barcodes will also have a lower scan rate—and truncated barcodes with larger X dimensions may have an even lower scan rate.  If circumstances (or customers) force a tradeoff, it is not clear whether a smaller X dimension or truncation is the better choice—but ISO and GS1 compliance is never a bad choice.

Sidebar: Verifier based on Scan Rate

Back in the 1980’s, Symbol Technologies produced a verifier called the LaserCheck, based on a PDT. It pre-dated ISO 15416 verification and was intended to be an improvement on Traditional Verification, grading barcodes based on Scan Rate. Unfortunately it was shown that scan rate did not correlate to actual barcode performance at the scanner and was therefore useless.