Monday 30 September 2013

2.5 Million UK Patents? I don't think so.

The UK IPO recently announced in their IPO Facto blog (archive version here) the "2.5 Millionth Patent!", on the occasion of publication of GB2500000. The IPO have also added it to their history section here, and the story has been picked up here, here, here, here, here and here. The immediate impression that any reader might get from this is that 2.5 million patents have now been granted since the Patent Office started. Great, you might think, and a vindication of the inventiveness of UK inventors punching far above their weight. After all, the much bigger USA only passed the 8 million mark a couple of years ago.

This impression would, unfortunately, be completely false. The Patent Office has not granted 2.5 million patents. It has, however, published 2.5 million patent applications (for the time being let's ignore what happened prior to the 1977 Act*). This is a very different thing. Under UK patent law, every patent application in the UK has to first be published before it is granted, unlike in the US where a patent number is only given once a patent issues. UK applications are, once a search has been done, first given a publication number ending in 'A', which is currently what GB2500000A is, and normally published 18 months after the earliest priority date. Only once the application has been through examination can it be granted. The same number is then used for publication of the granted patent, but with a 'B' at the end instead.

Given that not all patent applications proceed to grant, there will certainly be a lot fewer than 2.5 million patents granted by the UK Patent Office. I don't have the figures easily to hand (does anyone know?), but I would hazard a guess that the number will probably be somewhere around half to two thirds of that. There might have been two million patents granted, but I doubt it has got quite that far yet.

Currently, patent applications cannot be granted until at least 3 months from the date of publication. Therefore, even if GB2500000 is viewed by the examiner as being allowable (and it isn't at the moment), it cannot be granted until after 11 December 2013.

I should say, in case anyone is thinking I am being overly negative, that I think the invention defined in GB2500000 seems pretty neat. The application describes a monitoring method using microwaves, where electrically conductive fibres embedded in fabric are used to generate signals by reflection of microwave radiation. This could, the application says, be used for remote monitoring of personnel in hazardous environments, for patient monitoring or for monitoring athletes during training or performance. Very clever (if it actually works, of course). The search results look pretty promising too, with only citations of background relevance. If the applicant, Liverpool John Moores University, can overcome the various clarity objections raised by the examiner, it looks like they would be able to get themselves a granted patent. This will, however, not happen until after 11 December 2013, and it certainly won't be the 2.5 millionth UK patent.

*UPDATE 4/10/13: There is an interesting quirk here, for those interested in what happened before the 1977 Act came in. Since numbering restarted at GB2000001 under the 1977 Act (as mentioned in the IPO's history section), there are quite a lot of numbers that were not used, leaving a hole of nearly 400,000 that has not been filled. The total number of GB applications published is therefore at most about 2,100,000. 

Thursday 12 September 2013

Lantana: Mind your language!

A definition of insanity, at least according to some (and credited to various famous figures including Benjamin Franklin, Mark Twain and Albert Einstein), is doing the same thing over and over again but expecting a different result. This is the impression I sometimes get from reading decisions on patentability from the UK Intellectual Property Office. There are, of course, some cases where the applicant manages to pull a patentable rabbit out of their application hat when faced with a hearing, but in most cases the applicant is on a hiding to nothing. This is certainly borne out by the numbers: according to the IPO's decisions database, only 37 decisions in the area of excluded matter over the past 15 years have resulted in an application being allowed or remitted to the examiner, while 371 have resulted in refusal. A hit rate of just over 9% does not sound particularly promising. I suspect the number of applications that have simply been abandoned rather than pursued further at a hearing is much greater.

[Right (Figure 21 from the application): does this look technical enough?]

One tactic often used in such cases that has always struck me as being strange, not least because it never seems to work, is where the applicant (or more probably their patent attorney) rewrites the claims to sound more 'technical' without apparently changing the substance of the invention at all. The reasons for doing this might be more to do with wanting to be seen to be doing something than a belief that such changes might give the application more of a chance. A judgment from the newly-appointed Patents Court judge Mr Justice Birss, Lantana's Application [2013] EWHC 2673 (Pat), illustrates this point quite well. Claim 1 of the application (which admittedly from the outset was quite technical-sounding anyway) was rewritten during prosecution to read as follows:
An electronic data retrieval system comprising a local station, a remote station, a packet switched network to provide a transmission path between the local station and the remote station, and a machine-readable data storage device storing retrievable data files including machine-readable data representing at least one of a visual product and an audio product,
wherein said local station includes:
a data store storing a plurality of machine-readable data retrieval criteria identifying data files among said retrievable data files stored at said machine-readable data storage device to be retrieved;
a packet switched network interface connected to said packet switched network;
a user interface co-operable with said data store and interactable with a user, to enable selection by the user of one or more machine-readable data retrieval criteria; and
an electronic processor configured to produce, in response to the selection by the user of the one or more machine-readable data retrieval criteria, a first e-mail message including the selected one or more machine-readable data retrieval criteria together with a machine-readable instruction for retrieving data files, among said retrievable data files stored at said machine-readable data storage device, using the selected machine-readable data retrieval criteria, and to send the first email message to the remote station via said packet switched network interface and said packet switched network;
wherein said remote station includes:
a packet switched network interface connected to said packet switched network to receive the first e-mail message from the packet switched network;
a filter adapted to parse the first e-mail message to determine whether the first e-mail message includes any machine-readable instruction and any data retrieval criteria; and
an electronic processor to execute the first machine-readable instruction, and upon execution of the machine-readable instruction and in accordance with the selected machine-readable data retrieval criterion, retrieve the one or more required data files among said retrievable data files stored at said machine-readable data storage device from the machine-readable data storage device, produce one or more second e-mail messages, the one or more second e-mail messages including the retrieved one or more data files as one or more attachments, and send to said local station, via the packet switched network interface of the remote station, and the packet switched network, the one or more e-mail messages and one or more attachments.
A total of 378 words seems like a lot to define what is in essence a system with two computers arranged so that one computer could retrieve files by sending an email asking for a file stored on the other computer, which would then send an email back with the file attached. Justice Birss seemed to think so too (see paragraph 8). The rest of the decision is fairly standard, at least to anyone familiar with the UK case law on patentable subject matter, leading up to the Court of Appeal judgment in HTC v Apple earlier this year (see here for my comments). Unsurprisingly, the appeal was dismissed on the basis that Birss J could not see that the contribution made by the claimed invention had any kind of technical effect.

The question I have from this case is not whether the decision was correct (I think it probably was), but whether there are any reasons why the type of technical obfuscatory approach to claim drafting used in the case can be in any way useful. The only one I can think of is that it might sometimes sufficiently bamboozle an examiner into thinking there is more to the application than a more concise claim might suggest. Can anyone think of any other reasons?