Wednesday 10 April 2013

T 2086/11 – More Mean Misery


This is an appeal against the refusal of the application under consideration by the Examining Division, which had found the claims on file to lack clarity.

Claim 1 before the Board read:
1. An aluminum flake pigment characterized in that the mean particle size of aluminum flake particles contained in said aluminum flake pigment is in the range of 3 to 20 µm and among the aluminum flake particles contained in said aluminum flake pigment, the aluminum flake particles having major axes (D) of not more than 10 µm have a mean aspect ratio in the range of 8 to 20, calculated from individual aspect ratios (D/thickness of each particle) of at least 50 aluminum flake particles having major axes (D) of not more than 10 µm. (my emphasis)
The Board confirmed that this claim was not clear:

[1.1.1] It is established jurisprudence of the Boards of Appeal of the EPO that, in order to ensure legal certainty, a claim must clearly define the matter for which protection is sought in terms of the technical features of the invention; the clarity of the technical features contained in the claim thus serves the purpose of ensuring that the public is not left in any doubt as to which subject-matter is covered by a particular claim and which is not (see Case Law, 6th edition, 2010, II.B.1.1.1, first two full paragraphs on page 253 as well as T 728/98 [3.1]).

[1.1.2] The parameter “mean aspect ratio”, contained in the wording of claim 1, represents, in the present case, the mean of the individual aspect ratios, defined as the ratio of the major axis to thickness, of 50 aluminium flake particles having major axes of not more than 10 µm The method for measuring said mean aspect ratio is described in paragraphs 61 to 63 of the application.

According to this method of measurement a paint plate containing aluminium flake pigments, obtained as described in paragraphs 54 to 56, is cut into 1.5 cm square in order to prepare an observation sample. The aluminium flake particles are then observed in a section of the paint film of the sample by means of a digital HD microscope, the individual aspect ratios of 50 flake particles observed on the screen are measured and the mean of these 50 individual aspect ratios is calculated.

It is undisputed that the observed sample, which contains flake particles with a mean diameter of 3 to 20 µm, can contain thousands of aluminium flake particles having a major axis of not more than 10 µm.

It is also undisputed that the measurement contains necessarily an error (as remarked in paragraph 63 of the application itself) since the flake particles are observed on a screen and the observation of the major axis and thickness of the two-dimensional image of the individual particles depends on their orientation in the observed plate; therefore, it is not possible for an operator to assess with certainty if the observed particle is really showing its major axis and thickness or not. As explained by the Appellant and in the application (paragraph 63), the measurement has been repeated 50 times by taking 50 different particles in order to reduce the experimental error.

However, since the investigated sample contains thousands of particles having a major axis of not more than 10 µm and the method described in the application does not give any precise indication upon how to select such 50 particles, one operator, by selecting at random one group of 50 particles, would measure perhaps a mean aspect ratio according to claim 1 whilst another operator, by selecting for the same sample a different group of 50 particles, could find a value of mean aspect ratio outside the claim.

The Board remarks also that claim 1 requires a mean diameter for all aluminium flake particles of from 3 to 20 µm, i.e. that 50% of the particles are above the mean diameter and 50% below, but it does not contain any further requirement with regard to the homogeneity of the overall particle size distribution or of the particle thickness. Hence, an observed sample could contain, for example, 1000 particles with a size below 10 µm but with very different thickness.

Therefore, the average aspect ratio measured on a random selected group of only 50 particles, which would represent 2% of 1000 particles, cannot represent anyway a statistically significant indication of the real mean aspect ratio of the total group of particles contained in the tested flake pigment.

Therefore, the parameter chosen in claim 1 is by itself insufficient to characterize the claimed product in a reliable way so that a skilled person would know if the product tested falls or not within the extent of the claim.

The Board thus concludes that the parameter “mean aspect ratio” is unclear and that claim 1 does not comply with the requirements of A 84 EPC 1973.

Should you wish to download the whole decision, just click here.

The file wrapper can be found here.

4 comments:

Mandataire en colère said...

"The Board remarks also that claim 1 requires a mean diameter ..., i.e. that 50% of the particles are above the mean diameter and 50% below"

The board is making a confusion between MEAN and MEDIAN. This is rather shocking.

DrZ said...

It is undisputed that the observed sample, which contains flake particles with a mean diameter of 3 to 20 µm, can contain thousands of aluminium flake particles having a major axis of not more than 10 µm.....

However, since the investigated sample contains thousands of particles having a major axis of not more than 10 µm and the method described in the application does not give any precise indication upon how to select such 50 particles, one operator, by selecting at random one group of 50 particles, would measure perhaps a mean aspect ratio according to claim 1 whilst another operator, by selecting for the same sample a different group of 50 particles, could find a value of mean aspect ratio outside the claim.


Shocking statement. It's like the TBA had never heard of statistical analysis. Simple population stats would show this wouldn't be the case for the majority of cases (I can't be bothered to work out the numbers but I'm guessing at less than 2 sigma)

DrZ said...

Hahaha. Hoisted by my own petard

That should read the statement holds for greater than 2 sigma of cases

Anonymous said...

The term "mean particle size" for a flake is a meaningless concept though. Which dimension should be considered to be the "size"? Also, the measurement techniques used will influence the statistics and there is no indication in the claim how to measure the size.