In April 2007, as part of our Patent Busting Project, we asked the U.S. Patent and Trademark Office (PTO) to revisit its decision to grant NeoMedia a patent that broadly claimed to cover database lookups using things like barcodes. In October 2007, the PTO agreed to take another look, and last July, it issued an initial opinion that all 95 claims of the NeoMedia patent were invalid.

Earlier this week, the PTO announced that it intends to reissue the patent, but the reissued patent will have 89 rather than 95 claims, and those 89 claims are all narrower than the original claims. That's a win for everyone, though as you'll see if you read on, it's not at all clear that the reissued patent claims are any more novel than the original ones. At a minimum, though, the scope of the patent has been pared back, and that opens the door to new ways to "design around" the patent (that is, make design decisions that will create products or services that fall outside the scope of the patent). If you want to look at the documents from the reexamination proceedings, we've posted many of the key ones on our NeoMedia patent page.)

The rest of this post is rather detailed. It's difficult to explain what happened without a lot of details. But for those interested, let's plow ahead.

After the PTO initially rejected the 95 original patent claims, NeoMedia responded by meeting with the patent examiner, and then formally amending the claims. The amendments narrowed the claims by adding several additional requirements. The amended claims require:

  1. use of a database on a "remote server computer";
  2. that the remote database contains "predetermined relationships" between the barcodes and "pointers" (e.g. URLs); and
  3. that a "user computing device" (i.e. a local client) receives the pointer from the remote server and uses it to access the referenced information (i.e. client-side).

The second of those points wasn't mentioned in the original patent claims. And the first and last points were referenced only in a few specific claims in the original patent:

16. The method of claim 1 wherein the steps of accessing a database and extracting a pointer therefrom are carried out on a server computer located remotely from the user computing device.

51. The system of claim 36 wherein the means for storing a database is located on a server computer located remotely from the user computing device.

86. The user computing device of claim 71 wherein the software program is adapted to utilize the index to access a database located on a server computer remote from the user computing device.

So while previously only claims 16, 51 and 86 required that the barcode-to-URL mapping database be on a remote server (as opposed to being on the device used to read the barcode), that requirement now extends to each and every claim in the patent. Because claims 15, 51 and 86 would be redundant now that the rest of the claims have been narrowed by amendment, NeoMedia canceled those claims.

Moreover, the original patent expressly sought to claim variations in which the database had to be on the user computing device:

15. The method of claim 1 wherein the steps of accessing a database and extracting a pointer therefrom are carried out on the user computing device.

50. The system of claim 36 wherein the means for storing a database is located on the user computing device.

85. The user computing device of claim 71 wherein the software program is adapted to utilize the index to access a database located on the user computing device.

Because all of the claims have been amended to require exactly the opposite, namely that the database be located remotely, claims 15, 50 and 85 would, post-amendment, be internally inconsistent. So NeoMedia canceled those claims, too.

Thus, when NeoMedia two days ago issued a press release saying, "All ninety-five claims of this Barcode Lookup Patent have been confirmed as being patentable by the PTO," that was just plain wrong. First, NeoMedia itself canceled six of the claims. Second, and more importantly, although the PTO has said that it will be reissuing the NeoMedia patent with the remaining 89 claims, each and every one of those claims is now significantly narrower than in the original patent.

What does this mean for those seeking to make use of barcode technologies? Probably the most notable change in the NeoMedia patent claims is that the remote lookup database now must be a database of "predetermined relationships" between barcodes and things like URLs. NeoMedia argued to the patent examiner that the prior art disclosed the use of what it called a "direct encode" system—one in which the barcode directly encoded the pointer to further information. A court would therefore likely conclude that a barcoding system that uses an algorithm that directly converts barcodes into URLs (rather than converting the barcode into an arbitrary code that then is used to query a remote database that holds the URLs) does not infringe the patent.

The requirement that the database be on a remote computer is also an important new limitation. If the database is on the user computing device itself, that is outside the scope of the Neomedia patent. It seems likely that even if that client-side database is periodically synced with a remote database to keep it up-to-date, that would not infringe.

Finally, the requirement that the URL (or pointer) be used by the user computing device seems to require at least the following three steps:

  1. the user device sends the decoded barcode to the remote server;
  2. the remote server queries a database of predetermined relationships and sends the user device a pointer; and
  3. the user device uses that pointer.

That would seem to exclude the following scenario:

  1. the user device sends the decoded barcode to the remote server via, e.g., a web browser; and
  2. the remote server queries a database of predetermined relationships and redirects the browser to the appropriate URL based on that lookup.

NeoMedia might well have a different opinion on the foregoing issues, which a court will need to address in any lawsuit over the reissued NeoMedia patent. And because patent law can be complicated and sometimes counter-intuitive, if you're trying to decide what course of action to take in light of the reissued patent, you may want to hire a patent attorney to discuss with you the specifics of your plan.

All that said, and with due respect to the PTO, even the narrowed claims are of suspect validity.

Is it really novel to add the feature of having a remote database of predetermined relationships? (You'd think that UPC barcode systems would have done that years before.) Or the features of reading data client-side and doing a lookup based on that data server-side? (Standard fare in millions of client-server systems.) Or using the resulting information client-side? (Again, common practice in all sorts of client-server systems.)

Given that those features are the basis for the PTO's decision to reissue the amended patent claims, and given that those features were well known long before before the NeoMedia patent application was filed, the amended patent claims are obvious and thus invalid unless the combination yields novel and unexpected results. Where are the novel and unexpected results here?