Key points
- "As to the objective technical problem solved by present claim 1, the proprietor contended that the combination of features (g) to (m) provided "a robust mechanism for improve[sic] security of an authentication message by a client device" and referred essentially to the conclusions drawn in T 495/91, Reasons 4.2, according to which an objective definition of the problem to be solved by the invention should normally start from the problem described in the application "
- "However, according to the problem-solution approach as defined in T 1/80 (cf. headnote I) (see e.g. also G 1/19, Reasons 26; R 9/14, Reasons 2.1.1) and in the Guidelines for Examination in the EPO (see part G, chapter VII, section 5, items (i) and (ii)), the formulation of the "objective technical problem" should always be done after having identified the closest prior art (i.e. the suitable starting point) for the assessment of inventive step (see e.g. T 1861/17, Reasons 3.4). "
- "As a consequence, the conclusions of T 495/91 do not appear to be reconcilable with the well-established problem-solution approach. To avoid any misunderstandings: of course, by coincidence, the problem described in the application itself, i.e. also called the "subjective problem", may well correspond to the objective problem formulated later on the basis of the selected closest prior art; for example, if the suitable starting point is already cited in the application itself. But, the established problem-solution approach does not imply that, for the purposes of determining the objective technical problem, one should "normally start from the problem described in the application"."
- See also CLBA I.D.4.2.2: "According to the established case law, an objective definition of the problem to be solved by the invention should normally start from the problem described in the application/contested patent. Only if examination shows that the problem disclosed was not solved or if inappropriate prior art was used to define the problem, is it necessary to investigate which other problem objectively exists (see e.g. T 495/91, T 881/92, T 419/93, T 606/99, T 728/01, T 1708/06, T 1146/07, T 1060/11, T 204/16)"
- "Thus, claim 1 encompasses technically sensible scenarios where the "based on" relationship is implemented by merely concatenating the respective keys, i.e. "gluing" them together. Consequently, the subject-matter of claim 1 comprises embodiments where the respective elements are generated in ways which are entirely unsuitable to enhance the security of the underlying hearing-aid system. Therefore, the features of claim 1 which are not disclosed in document D1 do not cause a technical effect which is credibly achieved over the entire breadth claimed (see e.g. G 1/19, cited by the proprietor, Reasons 49, 82 and 124). The board understands that established authentication procedures and the respective messages involved, as such, were commonly known to the skilled person in the field of data communications at the relevant date."
EPO
The link to the decision can be found after the jump.
Peter, I am confused. Not by your write-up but by the text of the decision. Perhaps you can help.
ReplyDeleteI know the expression "closest prior art" and "realistic starting point" but want to know the answer to the questions "closest to what?" and "realistic based on what?".
The Board declares it wrong to "start from" the subjective problem. The correct approach, it declares, is to start from "the suitable starting point". It fails to tell us what starting point is or is not "suitable". But is it not so, that the enquiry into what is the suitable starting point starts with an enquiry into what Applicant declares in the specification as filed to be the technical field and the technical problem?
I think that the phrase "for the purposes of determining the objective technical problem, one should "normally start from the problem described in the application" does not pertain to the choice of the starting point but the next step: given a starting point document, what is the technical effect provided by the distinguishing features?
DeleteThank you, Peter, but that comment/answer leaves me none the wiser, how to decide whether any "starting point" prior art disclosure (D1) advanced by the Opponent is or is not "suitable". Perhaps the trite answer is that every prior art publication is "suitable" so long as it can provide the starting point of a credible line of argument that the claimed subject matter was obvious under the problem-solution approach.
DeleteFor example, if D1 has the most technical features in common with the claim, and yet is not "realistic" as a starting point for the notional skilled person, then it can be waved away as failing to provide a "suitable" starting point. But with such an argument we are going round in circles, chasing our own tail, aren't we?
For me, realistic is a more concrete and satisfying quality to debate than the undefined quality of "suitability". Suitable for what (other than as a starting point). It would be nice if the EPO could retire the terms "suitable" and "closest" and instead talk about what for the skilled person would have been "realistic".
But even then, we come back to the necessity of first fixing the field of interest of the notional skilled person. Can we do that without laying on the shoulders of that notional being a technical problem which that person is seeking to solve?
And if so, how do we identify that particular problem, in order to decide the preliminary question whether D1 is a "realistic" starting point.
Of course, it goes almost without saying, that as soon as Applicant amends the claim, a new game starts, with new opportunities to elect different prior art publications as "suitable" starts.
T 495/91, in translation: " In assessing inventive step, the Opposition Division did not base its assessment on the problem stated in the patent in suit, but on a modified problem, which it considered to be the provision of a coating process with a web throughput speed increased compared to document (4). It did not provide any reasons why the problem explicitly described in the patent in suit should not be considered the objectively solved problem."
ReplyDeleteThis Board implies that the OD in T 495/91 was right.
In other words, the Board is dealing with the 'bonus effect' case law line, not with the choice of the CPA (it seems to me).
Hello Peter and MaxDrei,
ReplyDeleteI think there is way too much emphasis on the OTP, which is an artefact linked to the PSA, to the detriment of the EPC. Article 56 refers to the person skilled in the art. This should imply in my view that the « art » in question is given full consideration in the obviousness assessment, esp. when it relies on the PSA, since the « problem » to be solved pertains to that art. Starting with the definition of a CPA without having defined the art is like putting the cart before the horse.
Turning to the discussion of how to define the problem, the logic is to look at the claimed subject matter and to the problem as presented in the application. The theory of the OTP is in itself disputable, because it relies on a piece of prior art which has been found by a search carried out after filing, which is inherently and inevitably hindsight-guided. As we know, this entails inter alia the risk of a definition of the problem tainted with hindsight.
The artificial nature of the OTP is also apparent when its implementation involves odd definitions, as when the OTP is no more than the search for « alternative » solutions. It may even be unfair to the applicant, when technical effects submitted by the applicant to distinguish over a CPA found only after filing by a hindsight-guided search are considered inadmissible if they cannot be derived from the original application.
The Board’s square statement that the OTP defined on such a basis should supersede the problem as defined in the application is in addition dogmatic and unrealistic. In real life, the problem and the explanations offered in the application provide valuable insights on the technical constraints and the advantageous effects provided by the claimed solution remain very relevant to the assessment of obviousness. The real-life problem is not only a solution to achieve a certain outcome, it also includes a host of constraints, such as constraints linked to the field of use, practical and sometimes regulatory factors, which must be also met. The person skilled in the art is aware of all these factors, this allows him to eliminate otherwise possible options. Even though the definition of the problem under the PSA needs some adjustment, this should be viewed as complementing the definition of the problem in the application, because the technical insights provided by the application most likely retain relevance for the assessment of obviousness. Instead of the Board’s dogma, my view is that a nuanced approach should be applied.
In response to Francis Hagel, I jump to the defence of the notion of the OTP. The EPO has to strike a balance between the legitimate interests of the inventor(patent owner) and the public. Obviousness has to be judged objectively. The patent application is written with full hindsight knowledge of the invention and near to full knowledge of the state of the art, enabling the drafter to frame the invention and its purported technical effects to optimise its progress in the obviousness enquiry. The public, ie the EPO Examiner, has to be given the means to block the progress to grant of subject matter which is in reality obvious even if, when based on the cleverly drafted specification, the claim can resist the Art 56 EPC objections formulated by the ED.
ReplyDeleteThe proof of the pudding is in the eating. Where is your evidence that EPO-PSA works in opposition proceedings to the systematic and unfair detriment of patent owners?
Of course, it goes more or less without saying, that advocates for the Applicant/Patentee have to be on guard for illegitimate obviousness argumentation under EPO-PSA. I would hazard a guess, that many cases fail as a result of poor advocacy rather than illegitimate aspects of the EPO-PSA jurisprudence.
@MaxDrei
ReplyDeleteT 2497/22 reported by Peter on June 25 illustrates my comment.