The Black Art Of Getting A Bogus Patent Approved
from the just-keep-trying dept
Joe Mullin has an excellent and detailed blog post about how Scott Harris — no stranger to patent-related controversy — got some bad news in his patent lawsuit against FedEx over its barcode scanning system, as the court’s claim construction greatly limited the patent. The patent is no longer technically held by Harris, but he may still have some economic interest in it. However, the shell company that holds the patent, BarTex, is still pushing forward and appealing the claim construction, and saying that every FedEx package that ships violates its patent on a two-part barcode. The big problem? It appears that similar technology had been in use long before Harris ever filed his patent — including by FedEx. In fact, there’s prior art in a national standard setting group.
Where the post gets interesting is that Mullin went on to explore just how Harris was able to get such a patent with so much prior art. The story highlights how so many terrible patents get granted:
Unfortunately for Harris, the two-part bar code had already been invented, according to the Patent and Trademark Office examiner who rejected all of the lawyer-inventor’s claims. The inspector based his decision mainly on the existence of an earlier patent, No. 5,920,062. That patent–filed in 1997 and describing a bar code with multiple “modules”–belongs to Uniform Code Council, Inc. (also known as the folks who actually invented standardized bar codes; in case you’re curious, such codes have been in use since since 1974.) See Harris’ rejected original application [PDF].
Harris’ bar-code inventing process didn’t stop there. Like all patent applicants, he was allowed an unlimited number of chances to get it right. He modified his idea so that the bar code’s second part would include its own unique information, but would also contain data embedded in the first part of the code. It was an idea, Harris explained in a four-page letter to the patent examiner dated April 15, 2003, that nobody had ever thought of before: “Why would two bar codes be provided that included the same information? This is quite simply not obvious based on the prior art. In fact, it is the present inventors [sic] recognition that using two bar codes with overlapping information may produce a significant advantage…” See Harris’ amended application [PDF].
Though the examiner found many examples of multiple-part bar codes, he didn’t find any examples of ones that had multiple portions with redundant information similar to what Harris described. As a result, Harris was notified in July 2003 that his patent would be issued.
In a nutshell, Harris tried to get a patent on a certain type of bar code, and was turned down. Then he added what was essentially redundant information to his original invention, and, voila, the nation was blessed with a new invention.
Mullin goes on to explain how this sorta thing happens all the time, and the inevitable dangerous impact on companies that actually do innovate and do produce products. The “cost” to “invent” is incredibly low — but the cost to “uninvent” is quite high:
The prosecution history of ‘377 patent is an object lesson in why defendants in infringement suits often express frustration over having to fend off patents that they consider vague, unreasonably broad, or just plain senseless given the state of the art in their industries. But the rules of acquiring patents strongly favor a determined and clever prosecutor.
Consider the examiner’s position: He or she is faced with limited time and resources–and an applicant free to ask for as many “do-overs” as he can afford. The examiner can’t say “no” without a reason to do so, and if that reason is prior art, it has to be more or less spot-on. It’s also worth noting that examiners work in a system that rewards the granting of patents, but not endless arguments with an applicant.
Because Harris prosecuted the patent himself, it has only cost him about $2,000, according to PTO documents–$861 when he filed the application, $690 upon issuance, and a $450 maintenance fee paid in 2007. Contrast that with the hundreds of thousands of dollars–at going rates for IP litigation FedEx’s costs–that FedEx has probably spent in legal fees to “uninvent” the ‘377 patent. Had the company not won at the Markman stage, and been forced to go to trial, its costs could have soared into the millions, even to secure a clean victory.
It’s a story we’ve heard many times before, but the point is made pretty clearly here. As per usual, we will ask our regular patent system defenders to explain how the system “worked” in this case, and how a random lawyer with no intention of ever producing any product helped “promote the progress” by getting a patent on something that was already in widespread use, and then suing a large company for doing what it had already been doing all along?