EP 3 543 928 A1 relates to a method for opening a door vehicle, more precisely to the localization of objects in the hold of a transport vehicle, the loading space being accessible via a plurality of doors. The position of an object determined during loading is tracked by means of optical position detection, so that changes in position (e.g. due to slippage) are known and the most effective door for accessing the object is indicated during delivery.
Brief outline of the case
The application was refused for lack of sufficiency and the applicant appealed.
Next to the lack of sufficiency, the board also dealt with the choice of the CPA.
The ED chose D1 as CPA as in [0035] it discloses a plurality of doors.
The board’s decision on the CPA
The aim of the invention is to select from a plurality of doors in a delivery vehicle the one that provides the most direct access to an object in the hold.
However, this also means that this problem and the claimed solution would not come into play at all without a plurality of doors..
The board noted that, at no point in D1 there is a concrete example disclosed in which several doors are used or a decision is to be made as to which of several doors is to be used. Figures 3A-3E in D1 each show only one door.
The task of opening the nearest door according to the invention does not arise in D1 and D1 is therefore unsuitable as CPA in order to assess IS.
Furthermore, D1 does not relate to determining the position of objects in the hold, but to determining the capacity of the hold. This represents a further significant difference to determining the storage location of individual objects.
The board held that, for the assessment of IS, the CPA should generally be a prior art document disclosing an object developed for the same purpose or with the same objective. The formulation of the original problem, the intended use, as well as the effects to be achieved should generally be given more weight than a maximum number of identical technical features.
For this reason, the board dismissed D1 as CPA and chose D3 as CPA.
D3 relates to a delivery van with a localization device, whereby the storage location is recorded during loading and displayed during delivery.
D3 comes much closer to the problem of the invention and is therefore a realistic springboard for the assessment of IS.
Comments
The definition of the CPA by the board is certainly not a scoop, as it is supported by a long line of case law.
It is surprising that an ED choose as CPA a document which has no link with the object of the claimed invention, even if it had apparently lots of features in common.
If a candidate to the EQE wants to be successful, choosing the document with the most features in common as CPA is the first step to the ticket for the next year, as in general, the argumentation is then as poor as the choice of the CPA.
Comments
9 replies on “T 553/23 – Choice of the CPA – The most features in common is not the best criteria for the choice”
Given that the UPC is distancing itself from the EPO-PSA obviousness enquiry, I wonder if this case (in German) is useful as education, and reminder, on the issue of hindsight bias (whether real or imagined) and what is a “realistic” starting point for the obviousness enquiry.
The applicant for patent can be assumed to know the prior art and to sculpt the specification in the light of that prior art. It falls to the drafter of that specification, with all that hindsight knowledge of the invention, to disclose in the specification what was the technical problem and what was the inventor’s solution.
For the EPO to apply that applicant-defined problem against the applicant seems to me nothing more than 1) giving full faith and credit to that which the applicant has written, and 2) “equality of arms” between the public and the inventor, with no more hindsight being applied by the EPO than by the application drafter.
I have yet to see any convincing reason why correct application of EPO-PSA to the obviousness enquiry, with competent representatives, adequately-financed adversaries, and a competent tribunal, will ever deliver a wrong result.
The Board’s decision is to be approved. The CPA has to relate to the « art » of the skilled person where the technical problem arises, as defined by the claimed subject matter. If the claims are amended to include a specific field of use, the « art » of the skilled person is shifted to become that field of use and a relevant CPA has to relate to that field.
As to the applicant’s knowledge of the invention, I do not think the notion of hindsight has any relevance. It makes sense for the applicant to describe the prior art and the problem to be overcome in a way consistent with the claimed subject matter.
As to the choice of the CPA, if different from the prior art presented in the original application, one must be wary of avoiding hindsight in the implementation of the PSA. The EPO has correctly acknowledged that the search of prior art inevitably implies hindsight, since it can but be based on the foreknowledge of the invention. An objective of the PSA is precisely to mitigate hindsight-guided use of search results in the assessment of IS. The risk arises if the CPA and the technical problem are redefined in view of prior art found after filing, on the basis of the foreknowledge of the invention.
What a good comment, FH. The caution not to formulate the OTP as one that contains elements of the solution is in itself trite. But it had not occurred to me before that the collection of prior publications gathered by a search based on the claimed solution is the thing that might render seemingly legitimate what would in truth be an illegitimate formulation of the OTP.
But if the applicant does a good search before drafting the patent application, and the drafter is experienced in EPO-PSA, that ought to protect the inventor against unjustified revocation of the patent.
And if the EPO tribunal also holds tight to the imperative of deriving the OTP out of the disclosure in the application as filed, that should protect against wrong decisions on obviousness (either way) by an EPO tribunal, oder?
@MaxDrei
Thank you for your appreciation. The principe as laid out by the EPO in the case law I.D.4.2.1, also and Guideline G-VII 5.2 EPO is that the definition of the OTP must not be such that it provides a pointer to the solution. This is an important requirement.
EDs should bear this in mind when they require that the technical problem be reformulated. As a matter of fact, EDs do not always bother to explain by adequate reasons why a reformulation of the problem is necessary. In this respect, a significant element should be for the ED to show that the proposed reformulation is no pointer to the invention.
As there is a large number of German judges acting at the UPC, there is no surprise that they want to distance themselves from the EPO-PSA. As it is not possible to say that it was “not invented here”, it is much easier to claim the danger of hindsight alleged to be present in the EPO-PSA.
I would claim that any form of assessing IS can be suspected of hindsight, as the problem and its solution are presented in the application. It is a bit the same as when one sees old exam papers and their solution. It all looks pretty obvious. The hindsight objection does not resist a careful look.
If the search allows to find a document closer to the claimed invention, the prior art disclosed in the application as filed should not be kept. Then it is easy to allege IS, whichever way IS is assessed.
The inventor has certainly started from a prior art known to him, as inventions out of the blue are rather rare those days.
I agree with you, that properly applied, with all its steps carefully respected, the EPO-PSA cannot deliver a wrong result.
Here the realistic stating point was the delivery van in which the position of the goods is determined when loading and certainly not a container with allegedly a plurality of doors, but actually having only one, and in which the problem is to determine its volume.
perhaps we’d be OK if this we were to drop the obsession with the term “closest” (and the often-accompanying assertion there can only one), and replace with “realistic” (as you both have indeed done). CL II D 3.1 (especially the bit about T405/14) rules this out but that does not stop many from persisting with creative arguments on what is “closest”.
I cannot find a clear distinction between cases in which several prior art documents are considered as CPA, and cases in which there is definitely only one CPA. For example, as you have written for T 364/22 here on the blog:
“The fact that at least one other prior art document could be used as a starting point for analysing IS, cf. D5′ and D18 considered in the contested decision and/or documents D1 or D6 considered during OP before the OD, is not a sufficient condition for not considering D3 as a possible alternative starting point. Thus, according to established case law, when two documents are both adequate to serve as the closest prior art, an IS must necessarily be demonstrated for both of them.”
So in this case, ED has chosen a prior art (D1), which is not even considered as an adequate starting point? Then how far away from the invention must the prior art document be for ruling it out as an “adequate” starting point. As not even ED seems to be able to figure this out, it is obviously not always a straightforward task.
@Oskar
As stated in T 364/22 you quote, an opponent’s right to be heard implies that each IS attack must be assessed on its merits. If an opponent submits several attacks based on different starting points, this should not be dismissed for the sole reason that this does not follow the PSA (or rather, a rigid application of the PSA). After all, the PSA is not binding on an opponent. On the other hand, a multiplicity of attacks may undermine the value of an opponent’s argument, as it may be viewed per se as an indication of non-obviousness.
That said, the requirement for the choice of a prior art as starting point for the IS assesment to be « realistic » makes sense. The prior art should be one which the person skilled in the art will likely consider for solving the technical problem as defined in the claims and the original application. This matches the settled case law which insists on a similarity of purpose or use. I think recent UPC decisions, while stating that the UPC is not bound by the PSA, have a similar approach.
@Oskar
See the twin decisions of the UPC Central Division (Munich) of 16 July, 1/2023 (Sanofi v Amgen) and 14/2023 (Regeneron v Amgen, Headnote 3.