An Insight Into 2D Barcodes, Glyphs, OCR and Other Scanning Technologies

By Karl Schumacher, Vice President and General Manager, Document Factory Solutions, Pitney Bowes Production Mail and Software Systems

How Do You Decide What You Need?
It can be difficult for the typical production mailer to sort through all the issues surrounding the best way to create and read the control marks needed on documents. Print or finish suppliers only have finite R&D budgets to invest in developing solutions. They can't cover all the bases, so of course they push the solutions in which they have invested. Meanwhile, the mailer is left to ponder what the best solution really is. Is the ideal solution out there? If not, which solutions are at least readily available and reliable? And looking to the future, what emerging technology is interesting but unproven? Just how long will it be before the solutions that mailers wish and hope for will be on offer?

When it comes to tracking and controlling documents, the questions and the answers reside in the realms of science, technology and physics. A wrong technology decision can really hurt the enterprise, while making the right decision requires paying close attention to all the factors affecting a specific mailing environment and a particular application. The matrix below offers a review of scanning technologies.

"Tried and True" indicates that the technology is widely used and has proven to be reliable for its intended purposes.

"Document Integrity" means that the control symbologies offers the capability for individual piece identification and adequate collation integrity cheeks.

"Advanced Functions" judges the capability of the control code to be used for high-value-added features such as in-line, matched addressing and the use of the code for sophisticated remittance and enterprise processes.

Whatever the technology for encoding and reading, the reality we all face is that placing control symbology on a document requires programming resources. However, because they don't view the mailroom as a "customer" of their services, Information Technology (IT) organisations are relatively uninformed about the financial benefits that increased functionality built into applications can create. Consequently, they tend not to track technologies that improve the print/finish process. The development of new applications makes technology decisions easier.

But whether they are deploying new technology or reengineering legacy code, it is unlikely that internal IT has the knowledge to determine the best scanning technology. The bottom line is that when you choose a supplier for print or finish, it is critical to evaluate their ability to help you with the application engineering, particularly in determining how to implement control codes on the document that are consistent with the application requirements.

What You May Not Know Can Hurt You The majority of applications in the world today use Optical Mark Recognition (OMR) "dash codes". Depending on the complexity of the application requirements, OMR might be perfectly suited for the task. Regardless of which scanning technology you use, the issues that will not go away are these:

• OMR is a very poor guardian against split collations, and

• OMR is a poor way to enable advanced or productivity increasing functions.

While it is true that you can build some layers of collation protection into OMR, the more you do so, the more you defeat the purpose of using OMR in the first place: ease of implementation, moderate symbol size and low start~up costs.

Barcodes: Sometimes Better, But Often Too Big
Barcodes - 3 of 9, 128, 12 of 5 (we'll get to 2D later) - should go into the symbologies hall of fame as the pioneer technology that brought document integrity~ inserter~based postal processing, piece ID, quality checking and other cost-saving functions to the finishing world. But, barcodes get big as the list of functions grows., and the bigger they get, the greater the chance for scanning errors. The more redundancy that is put in to reduce scanning errors, the bigger still they get. And so on. The good news is that barcode symbols have proven their value and so have become more widely known and used. The bad news? Barcodes are more sensitive than OMR marks to print quality issues and this can lead to productivity issues. The rule of thumb on barcodes is: Barcodes can deliver a lot, but the shorter and less data-intensive, the better.

OCR: Up and Coming, But Not Here Yet At the moment, Optical Character Recognition (OCR) is on the doorstep of acceptability for high~speed finishing applications. There are those that will tell you that OCR technology can "read anything, even handwriting." Others will even tell you that it's cheap and easy -just get a digital camera, some off-the-shelf lighting, a video decoding board and some OCR software! Now, ask an engineer. He'll tell you that OCR is a highly sensitive solution that will work for a narrowly and carefully controlled set of parameters. Paper, speed, OCR font, clear zones, code placement (head/foot), and depth of focus are all considerations, to name a few. Two things are for sure, first, OCR is not as reliable or forgiving a scanning technology as OMR or barcodes. Second, when an application requirement arises for OCR, in many instances, the supplier is asked to read an existing OCR string, such as a remittance processing code line, already on the document. The goal, apparently, is to avoid having IT get involved in application changes. OCR most often does not work that way. In time, OCR will become the preferred symbologies in print/finish for many important reasons. It needs more experience.

2D Barcodes: Too Good to Be True?
Two dimensional (2D) barcodes and glyphs are really variations on the same theme - the placement of bytes in a horizontal and vertical plane to increase data capacity and improve error checking. Many claims are being made for 2D barcodes, but few are verifiable as not many production mailers have implemented them. On one hand, supplier A will tout 2D barcodes' ability to carry information and their reliability due to redundant bits carried for error correction. On the other hand, supplier B will tell you how small 2D barcodes are and how they look like greyscale. Each supplier is implying you can have your cake and cat it too! As with everything, there are trade-offs. Some 2D barcodes, when finally rendered for reliability and appearance, are not that small. Some say 2D codes are "just as easy as regular barcodes." And they do have a point. Regular barcodes still rule as the standard symbologies.

File-Based Processing: The Best of All Possible Worlds
File-based processing provides a solution to the one shortcoming of conventional barcodes - the fact that they get larger as more functionality is added - by using a small identifying barcode and placing the controlling information in a data file. If you need more integrity than OMR or want to eliminate the audit checks at the end of the inserter, put the integrity control in the data file and run the job on a high-integrity, mailpiece tracking chassis. If you need postal processing, put the postcode and breaks in the data file and have an automated stacker build the bundles or trays for your operator If you want the security and cost savings of a closed-face envelope, just put the customer's information in the data file and have in-line addressing spray on a delivery address and a personalised marketing message. If you want the assurance that 100% of the mailpieces got mailed, put the reprint retrieval code in the data file and set up an automated damaged mailpiece regeneration capability. If you want to do a "late pull" of a mailpiece that you shouldn't have printed, pull up the data file and set the divert control flag for that mailpiece. Think you need marketing's blessing and IT's application programmers to do all that? Not at all - the barcode hasn't changed. A script from your software supplier is controlling the data file, and the inserters are running Windows® on a LAN.

The bottom line is you need the best of all possible worlds. The legacy code prints OMR marks and it isn't going to change. The standard barcode application you brought online last year isn't broken and when you mention "210 glyphs" to IT, they say Y2K and EMU, and your boss's boss is now the Finance Director.

You don't need technology or buzzwords, you need solutions. You need a finishing platform that will deliver productivity today and tomorrow, that will scan OMR and barcodes, that can deliver cost saving functionality from printed codes, and, to make the Finance Director happy, can support filed-based processing, OCR, 2D barcodes and next year's technology. Look for flexibility, proven by multiple installations.

Back to the Production Mail and Software info Centre