By Scott Blau, CEO Datacap
In 1972, the technical publisher Auerbach released a book on automation technologies. Toward the end of chapter three, there is a discussion on the arcane topic of handwriting - also called hand print - recognition. The author observed that this technology showed some promise in reading handwritten numbers, but also registered a caveat: whereas Germany and Japan had observed success using these techniques, in the U.S., not much progress had been made.
Why not? The author surmised: "In societies where individuals are more willing to follow instructions, this technology may have greater long term success."
Fast forward 30 years to the present. Intelligent Character Recognition ("ICR") hand print recognition is widely used in the "unruly" United States, and pretty much everywhere else. Chances are, ICR technology is being used to read your state income tax forms, mail-in catalog orders, and health insurance claims.
What was wrong with the Auerbach observation? Nothing, actually. It turns out that, in an unexpected way, they got it right.
Starting in the 1970s, something happened in Germany that changed the face of the ICR market - Germany's vaunted engineering skills were applied to software problems.
As a result, SAP was born in a sleepy little town south of Frankford. Further south, on lovely Lake Constance, across from Switzerland, some very serious engineers got down to the business of breaking the human code. Not the genetic code, but the code we use to communicate with one another on paper: handwriting, or rather, hand printed characters. At AEG, and then later at a splinter group that became CGK, some of the most powerful recognition technologies of the day were developed.
Indeed, throughout the 1980s and early 1990s, as ICR began to make slow headway in the U.S., banking documents were already successfully being processed throughout Europe with European-developed ICR technologies.
The first ICR technologies to be used in the U.S. were "native" technologies, developed by U.S. developers for the U.S. market. It turns out that Americans don't write so neatly and learn different ways of forming letters than they do in Europe. Because the U.S. is such a massive market, the expectation was that, as in so many other areas, once the kinks were worked out, U.S. technology would overwhelm the Continent.
However, exactly the opposite has happened. As U.S. developers struggled to extend to the fragmented European market, the Germans have swept in to North America. Constance spawned at least one more ICR developer, re Recognition, while CGK and AEG were reunited and then bought by OCE. Other European developers, from France (A2IA), U.K. (Neuroscript), and even Russia (ABBYY) have fielded new, ever more powerful recognition "engines." Meanwhile, U.S. development has "consolidated" - and today, there are fewer ICR developers than even three years ago.
As part of a broader trend of European, particularly German, investment in U.S. software companies, German companies have purchased or placed significant stakes in at least two low-end forms processing companies, Kofax and Cardiff. Almost all the forms processing vendors, including the Swedish-owned Readsoft, use German or other European recognition engines.
The fact is that ICR is now a global technology. My company, Datacap, provides forms processing and capture solutions tailored for the U.S. market. Our software uses the best recognition components available, and, not surprisingly, some of them are from abroad. The ongoing challenge is to provide solutions that serve the complex requirements of the U.S. market, without falling behind because of technology that didn't stay in tune with the realities of this market. Clearly, the best way to do that today is to maintain a global perspective.
back to top | request more information | contact Datacap | site map