A Review of the “Problem and Solution” Approach to Inventive Step under Article 56 EPC

A. Kennington (GB)

This is the third part of a long article, which is being published in three parts in successive issues of epi Information.


Part 3 – Modifying the Comvik Formulation of the Problem

The first part of this article considered the correct formulation of the problem in the problem and solution approach to analysing inventive step. It proposed that the formulation is only correct if the problem is known or obvious in view of everything made available to the public before the priority date. This rule was derived from the requirement in Article 56 EPC that there is inventive step if the claim is not obvious in view of the state of the art, and the definition of the state of the art in Article 54(2) EPC that the state of the art comprises everything made available to the public. Consequently, if the problem is not obvious in view of everything made available to the public, the problem and solution analysis cannot establish a lack of inventive step under Article 56 EPC. However, this rule appears to conflict with the treatment of non-technical features in the formulation of the problem under the so-called Comvik approach.

The second part of the article reviewed the origins of the Comvik approach and the circumstances under which it was developed. It concluded that treatment of non-technical features in the analysis of inventive step has developed since the approach was first proposed. It also concluded that in early decisions T 1053/98 Canon and T 0931/95 Pension Benefits Systems the same conclusions could have been reached without incorporating novel non-technical features into the problem. However, in decision T 0641/00 Comvik it would not have been possible to reach the same conclusion without the presence of the novel features in the problem. Therefore there is unavoidable conflict between the Comvik approach to formulation the problem and the rule that the problem must itself be obvious.

This part of this article considers whether it is appropriate to modify the Comvik approach to the formulation of the problem, and proposes a modification to the approach to make it consistent with Articles 56 and 54(2) EPC with the proposed rule that the problem should be obvious.

Possibility of Modifying the Comvik Approach

The Comvik approach, following decision T 0641/00, can be summarised as providing two rules for the treatment of non-technical features in a claim, where those non-technical features do not combine with technical features to contribute to a technical effect (although the Comvik decision itself does not explicitly identify these as separate rules). The first rule is that such features cannot contribute to the inventive step of the claim, and in particular they cannot form part of the solution in the problem and solution approach. The second rule is that such features can be included in the formulation of the problem, for example as a constraint that has to be met, regardless of whether the features are novel or not.

The first rule (non-technical features cannot contribute to an inventive step) is well established and is at the core of the current EPO practice in the treatment of non-technical features in claims. I do not wish to challenge this rule. On the contrary, I regard it as correct. This rule has been justified on the basis that inventions are inherently technical, so that anything that is wholly non-technical cannot be an invention, and this view is reflected in the wording of Article 52(1) of EPC 2000 which states that patents shall be granted for inventions "in all fields of technology". On this basis, if a claim includes a feature that is non-technical and does not combine with technical features to contribute towards a technical effect, that feature can be regarded as not being part of the invention. Article 52(1) requires that patents shall be granted for inventions that have an inventive step (as defined in Article 56). It may be that the non-technical feature of the claim meets the requirements of Article 56 (not obvious over everything that was made available to the public). However, such a feature is not part of the invention and so it cannot provide the invention with an inventive step. This line of reasoning leads directly to the rule that non-technical features cannot contribute to the inventive step of the solution.

On the other hand, I propose that the second rule (novel non-technical features can be part of the problem) should be modified to state that non-technical features may be included in the formulation of the problem, regardless of whether those features are recited in the claim or not, provided that the inclusion of those features in the problem is known or would be obvious in view of everything made available to the public before the date of the claim.

I think that it is reasonable to consider modifying this rule because the Boards of Appeal have developed their understanding of the problem and solution approach further since the time of decision T 0641/00 Comvik. In T 0641/00 Comvik, it was considered that the problem has to be a technical problem, whereas more recent decision T 1784/06 Comptel (21 September 2012) makes it clear that the problem and solution approach does not need to involve a technical problem (see part 11 of the Reasons for the Decision). Also, decision T 1689/07 Proctor & Gamble, Colour-changing absorbent article (6 November 2009) states clearly that a claim may have an inventive step even if the novel technical features are provided purely for a non-technical purpose. It seems difficult to reconcile this decision with the practice of including a novel non-technical purpose in the formulation of the problem. Thus there has clearly been a development of ideas about the nature of the problem in the problem and solution approach.

Also, the two rules set out above appear to be sufficiently independent that the proposed modification of the second rule does not have any impact on the first rule. Article 52 EPC requires that an invention must be technical and must have an inventive step. Therefore the solution in the problem and solution approach cannot rely on non-technical features that do not contribute to a technical effect, because they are not part of the invention. On the other hand, Article 56 EPC requires that an invention has an inventive step if it is not obvious over the state of the art, and Article 54 EPC defines the state of the art as everything made available to the public in any way before the priority date of the claim. Therefore the problem can be based on anything known before the priority date of the claim regardless of whether it is technical or not. Thus the need to preserve Comvik's first rule, concerning non-technical features in the solution, does not prevent a reconsideration of Comvik's second rule, concerning non-technical features in the problem.

Having established that it is reasonable to reconsider Comvik's approach to the formulation of the problem, there are several reasons to favour modification of that approach.

First, the Comvik approach to the formulation of the problem, in which novel non-technical features of the claim are included in the problem, appears to have become accepted practice without any consideration of its compatibility with Article 56 EPC.

Decision T 0641/00 Comvik, and the two preceding decisions T 1053/98 Canon and T 0931/95 Pension Benefit Systems, did not consider whether this approach to formulating the problem was compatible with Article 56. After decision T 0641/00 Comvik, it was regarded as an established part of the case law of the Boards of Appeal that non-technical features of the invention, even if new, could be included in the statement of the problem. Subsequent decisions, such as T 0531/03 Catalina Marketing (17 March 2005) and T 0764/02 Online Resources & Communications Corp (20 September 2005) explicitly regard the jurisprudence as having been settled by T 0641/00 Comvik and do not examine whether it is compatible with the EPC.

In decision T 0154/04 Duns Licensing, the appellant attacked the legality of the Comvik approach as a whole, and this decision contains a lengthy defence of it. However, the Board's argument is based entirely on Article 52 EPC, and there is no consideration of whether the approach to formulating the problem is consistent with Article 56. I am not aware of any decision of the Boards of Appeal that has considered whether the practice of including novel non-technical features in the problem can be permitted under Article 56.

In retrospect, it may seem surprising that there was no consideration of whether this practice was compatible with Article 56. However, this may be the result of the way in which the whole approach to non-technical features was regarded at the time. I have characterised the Comvik approach as consisting of the two rules that I summarised above, and I have argued that these two rules are substantially independent of each other. However, that is not how the approach is discussed in decision T 0641/00 Comvik itself. The Board in Comvik does not seem to have considered that it was establishing two separate rules, but instead it seems to have regarded the manner of formulating the problem simply as a convenient way of establishing that non-technical features could not improperly contribute to the inventive step. Since the exclusion of non-technical features from the consideration of inventive step was justified by reference to Article 52, it may not have been apparent at the time that there was any issue of compatibility of the formulation of the problem with Article 56.

If the Comvik rule for formulating the problem is compared with Article 56, it appears that there is inevitably a conflict. Article 56 defines inventive step in terms of obviousness "having regard to the state of the art", and Article 54 defines the state of the art as comprising "everything made available to the public by means of a written or oral description, by use, or in any other way". There can be no justification under these Articles of the EPC for basing an obviousness argument on anything that is novel. This immediately conflicts with the statement in decision T 0641/00 Comvik that it is correct to formulate the technical problem to include non-technical aspects "whether novel or not" (Reasons for the Decision, part 7).

In decision T 0641/00 Comvik, part 7 of the Reasons for the Decision begins

"The technical problem should not be formulated to refer to matters of which the skilled person would only have become aware by knowledge of the solution now claimed. ... Thus a problem should not contain pointers to the solution or partially anticipate it
However, in the Board's view this principle applies to those aspects of the subject matter claimed which contribute to the technical character of the invention and hence are part of the technical solution."

The principle, that the problem should not contain matters that derive only from the solution, and should not contain pointers to the solution, is fundamental to ensuring that inventive step is judged only against the state of the art and does not include any hindsight reasoning. Board does not explain why it considers that this principle applies only to aspects of the claimed subject matter that contribute to the technical character of the invention. It is understandable that aspects of the claimed subject matter that are not part of the technical solution should not be regarded as contributing to the inventive step (since such aspects are not part of the invention under Article 52 EPC). However, this does not mean that such aspects should be regarded as part of the problem. The removal of a feature from one side of the scales does not imply that it should be added to the other side. Instead, the novel non-technical features should simply be ignored. Once a novel but non-technical aspect of the claim is included in the problem, the question of inventive step is no longer being judged against the state of the art alone.

In conclusion, it seems that there has never been a proper evaluation of the compatibility of the Comvik formulation of the problem with Article 56 EPC, and there is at least a question over whether the formulation can in fact be permitted under Article 56.

The conflict with Article 56 EPC can be seen also in the comment included for several years in part G-VII, 5.4.1 of the Guidelines for Examination in the EPO (it has been removed in the 2015 edition)

The requirements specification is not deemed to belong to the prior art; it is merely used in the formulation of the technical problem.

This appears to be an explicit admission that this formulation of the problem involves the inclusion in the problem of features that are not in the prior art. Since Article 56 EPC requires that obviousness is judged "having regard to the state of the art", there is an immediate conflict between this formulation of the problem and Article 56.

It might be argued that there is no problem in including novel features in the formulation of the problem so long as they are non-technical, since non-technical features cannot be part of the invention and therefore do not matter. However, this cannot be correct. Article 56 EPC states that an invention shall be considered as involving an inventive step if it is not obvious having regard to the state of the art. This clearly excludes the possibility that a finding of obviousness (lack of inventive step) can rely on a feature that is not part of the state of the art. There is no leeway in Article 56 to base an argument of obviousness on anything that is not in the state of the art. Thus, if there is a finding of obviousness in view of the requirements specification, and the requirements specification does not belong to the state of the art, the finding of obviousness is not based on the state of the art and so it does not establish a lack of inventive step as defined in Article 56 EPC.

A second reason for favouring modification of the Comvik formulation of the problem is that the reasons given for formulating the problem in this way do not in fact support it, and in some cases actually contradict this formulation of the problem.

Thus in decision T 0154/04 Duns Licensing, part 16 of the Reasons, defending the Comvik approach, begins "For the purpose of the problem-and-solution approach developed as a test for whether an invention meets the requirement of inventive step, the problem must be a technical problem (see the COMVIK decision T 641/00 (supra), Reasons Nos. 5 ff.)". The decision then states that there are difficulties in defining an intelligible technical problem without referring to non-technical features. Next, the decision states "The Board, therefore, allowed in COMVIK an aim to be achieved in a non-technical field to appear in the formulation of the problem as part of the framework of the technical problem that is to be solved, in particular as a constraint that is to be met (Reasons No. 7)". However, the basic premise of this reasoning, that the problem has to be a technical problem, is no longer good law as has already been discussed above. Consequently the Comvik formulation of the problem is justified in T 0154/04 Duns Licensing on the basis of a legal requirement that no longer exists.

Decision T 0154/04 Duns Licensing continues in part 16 of the Reasons by stating "Such a formulation has the additional, desirable effect that the non-technical aspects of the claimed invention, which generally relate to non-patentable desiderata, ideas, and concepts and belong to the phase preceding any invention, are automatically cut out of the assessment of inventive step and cannot be mistaken for technical features positively contributing to inventive step. However, this contradicts the actual situation, since the non-technical features are not "cut out" of the assessment of inventive step according to the Comvik formulation of the problem, but are in fact transferred to the opposing side of the assessment as part of the problem to be solved. If they were truly "cut out", these features would not appear either in the formulation of the problem or in the features contributing to the solution.

Similarly, the decision in T 0641/00 Comvik states that "where a feature cannot be considered as contributing to the solution of any technical problem by providing a technical effect it has no significance for the purpose of assessing inventive step" (Reasons, part 6). However, if a feature is to have "no significance for the purpose of assessing inventive step" it should not be taken into account in that assessment. This implies that in the assessment of inventive step, such non-technical features should not be included in either the problem or the solution. Therefore the Comvik formulation of the problem is directly contrary to the fundamental objective of the treatment of non-technical features in decision T 0641/00 Comvik itself.

A further problem with the logic of the Comvik formulation of the problem can be seen in decision T 0154/04 Duns Licensing at part 16 of the Reasons, where it states "Since only technical features and aspects of the claimed invention should be taken into account in assessing inventive step, ... it is irrelevant whether such a non-technical aim was known before the priority date of the application, or not." The conclusion clearly does not follow from the premise. The requirement that only technical features are to be taken into account does not imply that a non-technical aim can be included in the problem. Specifically, the inclusion of a non-technical aim in the problem makes it inevitable that this non-technical aim will be taken into account, which is directly contrary to the stated requirement that only technical features and aspects should be taken into account. Also, it is not explained how the desire to avoid taking non-technical features into account can justify the incorporation of novel features into the problem.

Thus it can be seen that the reasons given in decision T 0641/00 Comvik and in its defence in decision T 0154/04 Duns Licensing do not in fact provide any reason to maintain the Comvik approach to the formulation of the problem (new and non-obvious non-technical features of the claim may be included in the problem). It is sufficient, in order to satisfy these reasons, simply to use the Comvik approach to non-technical features in the solution (non-technical features that do not contribute to a technical effect cannot be used to provide a solution with an inventive step).

A third reason for moving away from the Comvik formulation of the problem is that it appears to make the patentability of a claim depend on what non-technical features it contains, in direct contradiction of its objectives. This has arbitrary and contradictory consequences for the patentability of different claims to essentially the same invention.

Decision T 0641/00 Comvik considers that the problem has to be technical, and therefore non-technical features will not normally be included on the statement of the problem. However, part 7 of the Reasons in decision T 0641/00 Comvik states that "where the claim refers to an aim to be achieved in a non-technical field, this aim may legitimately appear in the formulation of the problem as part of the framework of the technical problem that is to be solved, in particular as a constraint that has to be met". Thus, the non-technical features that "may legitimately appear in the formulation of the problem" are those that define a non-technical aim that appears in the claim. Claim 1 of the 1st auxiliary request in decision T 0641/00 Comvik included the feature of "the selective activation being used by the home database for distributing the costs for service and private calls or among different users" as the final clause of the claim. Since the justification for including this feature in the problem was that it appeared in the claim, it must be concluded that if this feature had not appeared in the claim at all it could not have been included in the problem, according to the rule for formulating the problem as set out in the decision.

However, the finding of lack of inventive step in decision T 0641/00 Comvik relies entirely on the inclusion of this feature in the problem. If the final clause of the claim had simply been deleted, this non-technical aim would not have been present in the claim. As a result, it would not have been permitted to include it in the problem. In the absence of this feature in the problem, the Board could not have used the reasoning that led them to find that the novel technical features of the claim were obvious. It appears, from the reasoning in the decision, that this situation would have obliged the Board to find the claim non-obvious and hence inventive. In this way it seems that the reasoning of the decision, if followed strictly, implies that the simple deletion of a non-technical aim from the claim would have rendered the claim inventive, without any change to the technical features. Such a result, implied by the reasoning of decision T 0641/00 Comvik, would be the exact opposite of the intention of the Board, who wished to ensure that inventive step would not be dependent on non-technical features.

It also follows from the reasoning of decision T 0641/00 Comvik that if the same technical features are combined with different non-technical aims in different claims, the non-technical aim included in the formulation of the problem would be different for the different claims. These different formulations of the problem could lead to different conclusions concerning the obviousness or non-obviousness of the technical solution, and therefore different conclusions about the patentability of the claim. Thus the same combination of technical features could be patentable or non-patentable depending on which non-technical aim was stated in the claim. Given that an invention must be technical, and non-technical features cannot contribute to inventive step, it seems contradictory that the patentability of a claim could depend on which non-technical features are recited in the claim.

It must be concluded from this result that the precise method of formulating the problem set out in decision T 0641/00 Comvik cannot be sustained.

Finally, it is generally accepted that the patentability of an invention is not affected by the reason why the invention was made. An invention is patentable or not on the basis of the features of the invention, irrespective of why the inventor invented it. Therefore it cannot be right to treat the inventive step of the technical features of an invention differently depending on whether the underlying purpose for making that combination of technical features is a technical objective or a non-technical objective.

Consider a situation where the technical features could be provided either for a technical purpose or for an aesthetic (non-technical) purpose. According to the Comvik formulation of the problem, it is permissible to include the non-technical aim in the formulation of the problem even if this non-technical aim is entirely new. On the other hand, it would not be permissible to include the technical aim in the problem if this aim is entirely new, since the problem may not contain pointers to the solution. If the claimed technical features would not be obvious in the absence of the aim, but become obvious if either aim (technical or non-technical) is considered, then the inventive step of the claim depends on whether the technical aim or the non-technical aim is considered at the time of formulating the problem. This has the effect that the patentability of the same set of technical features depends on why the invention was made.

Since the patentability of an invention should depend only on the technical character of the invention, and not on the reason why it was made, it does not appear possible to support the Comvik manner of formulating the problem.

Proposal – Refined Comvik Approach

For these reasons, I suggest that it is now correct to consider that the law has developed since decision T 0641/00 Comvik, as the Comvik approach has been refined and established, and it is now justified to abandon the Comvik manner of formulating the problem.

In particular, it is now recognised that the Comvik approach involves explicitly ignoring, in the assessment if inventive step, those features that do not contribute to a technical effect, and this principle can be applied independently of the proposal to formulate the problem by reference to non-technical aims stated in the claim. This separation of the approach into two distinct parts is not present in decision T 0641/00 itself, but has only been recognised during the application of the approach in later decisions. In decision T 0641/00, the incorporation of non-technical features of the claim into the statement of the problem was not seen as a separate principle from the removal of non-technical features from the assessment of inventive step, but instead it was seen as the manner in which the removal of non-technical features was to be achieved.

With experience of applying the Comvik approach gained over more than ten years since decision T 0641/00 was issued, we can now see that the incorporation of novel non-technical features into the formulation of the problem is unnecessary. It is also problematic. It conflicts with the requirement of Articles 54(2) and 56 EPC that obviousness is considered having regard to everything made available to the public before the date of filing of the European patent application. Additionally it appears to make patentability depend on the reasons why the invention was made and on the choice of what non-technical features (if any) are recited in the claim, rather than depending on the technical features of the claim.

Accordingly, I propose that when a claim includes non-technical features, inventive step should be assessed using a refined Comvik approach in which those features that do not contribute to a technical effect are not regarded as providing an inventive step to the claimed solution, but such features are not included in the statement of the problem unless they are known or obvious.

This approach can be supported in the EPC in the following way. (i) It follows from the wording of Article 52(1) that an invention is inherently technical. Therefore only those features in a claim that are technical, or which contribute (on their own or in combination with any other feature(s)) to a technical effect, can be regarded as part of an invention. Article 52(1) also requires that an invention must have an inventive step. Therefore a claimed feature that does not contribute to a technical effect cannot contribute to an inventive step because it is not part of the invention. (ii) An inventive step is present if the combination of features having a technical character or effect are not obvious over what is made available to the public anywhere in the world (Article 56 in combination with Article 54(2) EPC). Therefore claimed non-technical features cannot be considered as part of the problem if they have not been made available to the public, but conversely such non-technical features can be included in the problem if they have been made available to the public.

To some extent, this already appears to be the practice of the Boards of Appeal. For example, decision T 1784/06 Comptel (21 September 2012) explicitly upholds the Comvik approach that only features contributing to the technical character of the claimed subject-matter enter into the examination of inventive step (T 1784/06 at part 2 of the reasons for the Decision), but does not mention the inclusion of non-technical features of the claim in the formulation of the problem. In fact, this decision stresses that the problem does not need to be technical, in contradiction to the decision in T 0641/00 Comvik where it is stated that the problem must be technical.

However, the Comvik formulation of the problem is still being used in some cases. For example, in decision T 2063/09 BlackBerry (29 October 2013) the Board considered that the claimed feature, that a user had a fixed monthly data limit in an image repository, was a non-technical feature that could be added to the problem as a constraint to be met. It does not seem to have been necessary for the Board to taken this approach, since such data limits are very well known and it could easily have been established that they were part of the prior art (which appears to have been the view adopted in the examining division decision under appeal). If the present proposal for formulating the problem is adopted, the Board would not be able to add such a feature to the problem simply because it was in the claim and instead it would have to show that the problem (including the claimed non-technical feature) was obvious in view of the state of the art. In T 2063/09, it would probably have been easy for the Board to have done this.

Conclusion

The purpose of the problem and solution approach is to determine whether the claim is obvious over everything made available to the public before the priority date of the claim (Article 56 EPC, together with Article 54(2) and Article 89). Since the problem and solution analysis relies on an assessment of whether the claim provides an obvious solution to the problem, this approach can only be valid if the problem is itself known or obvious over everything made available to the public before the priority date of the claim. If the problem is not obvious over what is already known, no valid conclusion can be drawn from the problem and solution analysis because obviousness of the solution over the problem fails to establish obviousness of the invention over the state of the art.

Since the assessment of inventive step has regard to everything made available to the public, it is acceptable for the problem to start from any known disclosure, and it is acceptable for the problem to include non-technical features provided that they are known or obvious. However, since an invention is inherently technical (Article 52(1) EPC), the solution provided by the claim cannot derive inventive step from non-technical features that do not contribute to a technical effect. On the other hand, the claim can derive an inventive step from a technical feature that is new and not obvious over what is already known even if the technical feature is an obvious solution to or consequence of a non-technical problem or consideration that is itself unknown and not obvious over what is already known. The mere fact that a technical feature has a (new and non-obvious) non-technical aim or motivation cannot be used to deprive that feature of inventive step over what is already known. In order to be relevant, that aim or motivation must itself be already known or obvious, and this applies equally to non-technical aims and to technical aims.

This will require a change in the manner of formulating the problem in cases where a non-technical aim is stated in a claim, but I believe that this change is both necessary and justified for the reasons set out above.

The rule, that the problem must be obvious over the state of the art, should make it easier to determine whether the problem has been stated in a permissible manner or whether it inadvertently includes an impermissible pointer to the solution. This may provide an objective way to resolve disputes over the proper formulation of the problem in some cases.