Smart invalidation

Today’s post will be relatively short, I think, as it will merely confirm a trend already noted in earlier posts for instance here, here and there.

The trend is that, yes, it is incredibly difficult to win a patent nullity case based on public prior use.

On April 4, 2011, Grenoble-based company Smart & Green (formerly IPW Europe) filed a nullity suit against Mr. Christophe Frilley, owner of French patent No FR 2944858. It took 6 years for the Paris Tribunal de grande instance (TGI) to issue a decision. But before a rumble of discontent starts spreading in the readership, I should add that the TGI judges cannot be blamed for this long period of time, based on what I could gather from the public inspection of the patent file.

In fact, at the time the nullity suit was filed, the patent had not been granted yet. The judgment does not specify whether a formal stay was pronounced, but at any rate the case could certainly not progress until the grant of the patent.

In addition, the plaintiff apparently also filed a claim for ownership of the patent application. This led the French patent office to stay the grant proceedings, from May 2011 to September 2014, at which point the ownership claim was withdrawn, which led the French patent office to resume the grant proceedings.

The patent finally issued in August 2015. A few days later, the patent proprietor filed a request for limitation of the patent, which was granted a few months afterwards. 

Foreign readers may wonder why the applicant did not simply amend the claims before issuance of the patent, as opposed to afterwards. The explanation is simple: there are only limited time frames within which an applicant can amend claims, in French grant proceedings. In this case, the relevant time frame had closed after the applicant responded to the preliminary search report back in 2010. Therefore, the applicant was unable to react to the plaintiff’s nullity arguments by way of a claim amendment in the grant proceedings. He had to take advantage of limitation proceedings after grant.

Let’s now move on to the nullity arguments presented by Smart & Green.

Looking for something smart and green? How about this 1895 “mechanical flower for advertising purposes“?

The patent at stake relates to a lighting assembly. Smart & Green argued that the claimed invention had been disclosed in its so-called Rock lamp (for swimming pools) which was allegedly put on the market in 2008. 

The court was convinced that the Rock lamp was indeed made available to the public before the filing date of the patent, since at least samples of the lamp were sent to third parties for future marketing.

However, there was not enough evidence that these lamp samples were according to the claims of the patent. In particular, claim 1 of the patent as limited mentions a plurality of diffusers which can interchangeably accommodate a lighting module, in a waterproof manner.

But the documentation from 2008-2009 provided by the claimant did not clearly show whether this feature was present. 

By way of example, one email mentioned that any diffuser could be adapted on the lamp, but the court noted that it was unclear whether this meant that the Rock lamp was modular, as required by claim 1; or whether it simply meant that manufacturing could be adapted to include any diffuser.

Smart & Green had also filed two Soleau envelopes in 2008 on a so-called Little Rock lamp (for a presentation of what is a Soleau envelope, see this previous post). But again, the product described in these documents did not clearly comprise interchangeable diffusers.

The court was apparently more convinced that the feature at stake was indeed shown in internal development documents dating back to the summer of 2008. But these documents were not public and related to a so-called “IPW Lamp Project“. It seems that there was no evidence that the IPW Lamp Project was implemented in the Rock lamp which was made available to the public – or that it was publicly disclosed in any other way.

This led the court to reject the public prior use challenge.

Fortunately for the plaintiff, they also had good old traditional prior art in store as well, in the form of a Japanese patent application and a European patent application. 

The court held that each of these documents disclosed most of the features of claim 1, except the two which were added in the limitation proceedings, per which the diffuser can interchangeably be fixed to the lighting module owing to clicking or screwing, and a sealing ring is provided between the lighting module and the diffuser.

But these features were found to be obvious for the skilled person, even without referring to any specific secondary reference: the clicking or screwing fixation would be achieved simply based on common general knowledge, whereas the sealing ring was obvious to provide in view of the objective of supplying a waterproof lighting assembly.

Claim 1 was thus canceled for lack of inventive step. The dependent claims were quickly reviewed one by one, and as we like to say in French, the baby was thrown out with the bathwater.

I promised a short post, so now seems like a good time to stop.

But the bottom line is that I am still looking forward to reading a successful public prior use argument these days.


CASE REFERENCE: Tribunal de grande instance de Paris, 3ème chambre 2ème section, March 24, 2017, Smart & Green v. Christophe Frilley, RG No. 11/06949.

Shooting gallery

We are all familiar with the quote “Football is a simple game. Twenty-two men chase a ball for 90 minutes and at the end, the Germans win. Turns out the quote has an author, by the way, if Wikipedia can be trusted on this one.

Well, the same could be said of public prior use as a patent invalidity defense. It is a simple game. The alleged infringer (or the opponent) collects bits and pieces of evidence for months, and at the end the court (or the EPO) says the patentee wins.

This is because no matter how thorough and careful you are when documenting a public prior use story, it seems that you are almost never thorough and careful enough.

OK, this is oversimplified of course. But I have already reported on how severe judges can be with evidence of public prior use. And the Filmop decision that I talked about last week is another telling example of how a public prior use story can be wiped out, piece by piece, like in a shooting gallery.

This shooting gallery was patented in 1914 – a year sadly famous for many other shootings.

In my previous post, I explained that the FR’204 patent to Concept Microfibre was revoked for lack of inventive step on appeal. But before getting there, the court first discarded allegations of lack of novelty in view (among others) of two public prior uses.

The first public prior use was a so-called Puli-Srub washcloth, which was precisely the allegedly infringing product. According to the defendants, this product was already marketed by Korean companies B&H and Daego, starting from the end of 2006.

The second public prior use was a so-called MN Heavy washcloth, supposedly marketed by Swedish company MN Produkter Ab since 2006.

The FR’204 patent was filed in January 2008.

Regarding the Puli-Srub product, we can gather from the decision that the following evidence was offered:

  • one or more affidavits from the CEO of B&H, with a sample of washcloth attached, concerning sales to a company called Lilleborg in 2006;
  • an affidavit from the CEO of B&H, with a sample of washcloth attached, concerning sales to another (Danish) company called Stadsing, together with an affidavit from Stadsing itself;
  • screenshots from the 2006/2007 Stadsing catalogue;
  • invoices from Daego to Stadsing dated 2006-early 2007;
  • emails dated 2007 between B&H and Lilleborg;
  • emails dated 2005 between Daego and Filmop (one of the alleged infringers in the lawsuit);
  • invoices from Daego to Filmop as well as travel documents dated 2006;
  • affidavits from two employees of DME (the other alleged infringer in the lawsuit, a subsidiary of Filmop).

Quite an impressive list, right? Uh, no, in fact the court was not impressed at all.

First, the court remarked that the B&H affidavits, “drafted more than six years after the facts, are not confirmed by contemporaneous objective evidence”. Further down, they noted that some of the other affidavits were drafted by “employees of DME, a party to the present proceedings”.

I think the bottom-line is that the court does not really trust affidavits. Those signed by employees of a litigant are inherently suspect, while those signed by third parties are also deemed untrustworthy if they relate to facts which took place in the distant past. Such mistrust is understandable but, on the face of it, somewhat unfortunate. Testimonies should be given serious consideration – even though they should of course be carefully examined. Also, they are necessarily prepared only when litigation arises, as there is no need for them beforehand. But as will appear below, in this case there were probably a number of objective reasons for not simply trusting the affidavits.

So, it seems that affidavits are only really taken into account if they confirm non-testimonial evidence on the record.

Here, the non-testimonial evidence was deemed inconclusive.

For instance there was no clear evidence that the washcloth samples were the same as those marketed from 2006. Besides, the packing list which was provided regarding B&H’s sales to Lilliborg was only dated 2008.

The Stadsing affidavit was altogether ignored as it was in English and untranslated. The screenshots of the Stadsing catalogue were said to be less conclusive than an original version or even a photocopy, and were also said to be illegible – which certainly does not help.

The Daego invoices contained product references which could not be directly linked to the Puli Srub product. And the various emails on file concerned various projects, labeling or technical issues, but not actual sales.

Regarding the alleged sales to DME / Filmop, the court primarily relied on the patentee’s evidence, namely the infringement seizure dated July 2012. Based on the bailiff’s report and on the Filmop catalogues, the court held that the Puli Scrub was placed on the market only in 2010. DME and Filmop’s argument was that the same product was previously marketed under the name Micro Activa. But the court concluded that it was a different product, as it was still advertised in the catalogue in 2010 and 2011 beside Puli Scrub. Furthermore, based on the 2007 catalogue, it appeared that the Micro Activa washcloth was made of microfibers only – and therefore was not according to claim 1 of the patent (as reproduced in the previous post).

Next, regarding the MN Heavy prior use, we can gather from the decision that the following evidence was offered:

  • printouts of the MN Produkter website made by a bailiff in 2015;
  • a letter from the CEO of Decitex dated 2015, further to a judicial request made by the appellants, regarding the provision of a sample of MN Heavy from MN Produkter to Decitex;
  • emails dated 2008 and 2016 between the CEO of Decitex and MN Produkter’s commercial manager.

Starting with the website printouts made by the bailiff, the nullity claimants apparently relied on the mention “copyright 2006” on the website to argue that the content of the 2015 website was the same as that of the 2006 website. The court replied that this mention was no evidence at all. In addition, an interesting argument was made based on the photograph dates of the MN Heavy product on the website, which were apparently earlier than the patent filing date.

In this respect, the court relied on a counter-report filed by the patent proprietor, which showed that an online photograph date can easily be modified by editing the properties of the file. Therefore, the photograph dates found by the bailiff in 2015 were not necessarily the exact dates at which the photographs were taken. And, most importantly, what mattered was the date at which the photographs were posted online – which remained unknown anyhow.

Turning now to the CEO of Decitex, he stated that he received a sample of MN Heavy when he visited MN Produkter in the summer 2008. This was after the filing date of the patent, but he added that the product was on the market since at least January 2007. The court noted that the circumstances of the supply of the sample were unclear, and doubted how the CEO could possibly be certain that the 2008 sample corresponded to a product on the market since 2007. On a more formal (if not formalistic) note, this person did not testify as an individual but rather as the CEO of Decitex (since the judicial request to which he replied was addressed to the company itself). And he joined Decitex only in 2010, namely two years after the reported supply of sample by MN Produkter.

The 2008 emails between Decitex and MN Produkter were not persuasive as it can be derived from them that MN Produkter ordered washcloths from Decitex, and not the other way around.

Finally, the 2016 emails were also discarded. Although we do not know their detailed contents based on the decision, it seems that they made reference to an old page of the MN Produkter website, available on the web.archive.org website. The court stated that this is:

an archive service operated by a private entity without any legal authority and the operation conditions of which are unknown, which does not make it possible to consider that the date or content of the archived webpage are undeniably proven.

It is interesting to make a parallel with the EPO’s policy in this respect. Although the archive.org website was originally considered in some board of appeal decisions as an insufficiently reliable archive source (see e.g. T1134/06), more recent decisions seem to consider the website as trustworthy (see e.g. T286/10) – although this is also a matter of which standard of evidence the Board decides to apply in each case.

In section G-IV, 7.5.4 of the Guidelines for examination, it is now stated that the archives on this website “are admittedly incomplete, but this does not prejudice the reliability of the data that they contain”.

It is likely that the court, in the present case, was not fully briefed on how archive.org works and why it can be considered as a reliable source of information. If they had been, maybe their approach would have been more similar to the EPO’s.

Apart from that, the court’s way of dealing with public prior use allegations does not strike me as markedly different to the EPO’s. According to the latter, the date, subject-matter and circumstances of the public prior use are to be established in a precise manner (Guidelines for examination, G-IV, 7.1). Again, if there is any loophole, inconsistency, or lack of reliability in the evidential record, the conclusion is generally that the public prior use is not adequately proven.

This is harsh on opponents or litigants who happen to know (or believe) for a fact that the public prior use allegation is true and relevant but have a hard time collecting evidence. But this high standard is the price to pay to prevent patents from being revoked due to unreal or made up prior art.


CASE REFERENCE: Cour d’appel de Paris, pôle 5, chambre 1, September 27, 2016, Filmop Srl & SASU Distribution de Matériel Européen v. SAS Concept Microfibre, RG No. 14/18000.