The objective technical problem should not be that much of a problem

Reading time: 9 minutes

We have seen some interesting decisions lately that deal with basic principles in the assessment of inventive step / obviousness. For instance, the Supreme Court provided guidance how to deal with a non-enabling disclosure as a closest prior art document; see this Blog here. But there is yet another decision of potentially wider relevance, i.e. in that it apparently re-defines how to formulate the ‘objective technical problem’.

It’s silly season now, and I could not help but reflect a bit about the typical learning curve of a patent attorney trainee when getting acquainted with such basics. If you are in a rush, feel free to jump over here.

Patent law is somewhat mysterious. The basics of patentability are very simple: The claimed subject-matter shall

  • be new (‘just don’t claim exactly what has been out there already’); and
  • involve an inventive step (‘a bit creative, please – not just obvious’);

On the merits, that’s pretty much it. Admittedly, there are some side aspects to observe, e.g.

  • it should have industrial applicability (‘it should be useful’);
  • it should not be excluded from patentability by law (‘don’t do what is explicitly forbidden’);
  • it should be sufficiently disclosed (‘tell the world how to make use of it’).

But that doesn’t seem to be a big deal. A quite common first thought of a patent attorney trainee goes like:

Well, that’s it? Why do I have to wait three years to sit (and pass!) the exam?

Learning by doing

Later, one realizes that things are much more complicated than it seemed at first glance — and that it’s probably not a bad idea to test these basic concepts in real cases for a while, and one is grateful for some supervision before actually sitting the exam (while still at least hoping to pass).

Thereafter, it’s like driving a car. Passing the exam does not yet mean that you are a good driver. Still, at some point sooner or later and with some experience in the field, almost everybody will fairly get along on the streets.

This is no invention. No way!

It’s the same with patent attorneys: We are said to have quite firm opinions (when we finally have made up our mind), and it’s our bloody job to transfer this opinion to whoever needs to be convinced, every single day.

But patent attorneys constantly fail. Just look at all the negative office actions in the mailbox every day. Still, we know how to play the guitar. No doubt that a case only failed because the specific piece of music was not well perceived by the audience.

But once in a while, even the basic concepts are challenged. Things are heavily on the move in recent times, e.g. in terms of the assessment of infringement under the DoE or the abolition of the infringement test for SPCs. Still, all this is not too much out of the ordinary.

But when the common general understanding of how to assess inventive step / obviousness is whobbling, I am getting concerned. That’s my day-to-day business! Do I still play the game correctly?

Assessment of ‘inventive step’

In knowledge of what the invention is, it’s hard (if not impossible) to objectively evaluate whether or not it had been obvious before the invention has seen the light of day. As soon as you are aware of the invention, the unavoidable hindsight bias undermines any objective assessment.

Case law should be predictable, and individual judgments must be ‘right’. That’s a tricky balance — and whenever things are tricky, it’s comfy to have standardized tools at hand. This is where the EPO’s problem-and-solution approach comes into play (Guidelines, G-VII, 5), and which the FPC has repeatedly indicated to apply. It aims to avoid hindsight bias, and it does so with a rigid framework to be worked through. In brief, the problem-and-solution approach involves three main steps:

  1. determining the ‘closest prior art‘;
  2. establishing the ‘objective technical problem‘ to be solved; and
  3. considering whether the claimed invention would have been obvious to the skilled person.
Close, but hard to reach

Step i), i.e. the question of whether or not something should be considered as the ‘closest prior art’, is mostly a major issue in any discussion of inventive step. I won’t dive into this now, but I strongly recommend Oliver Randl’s Post on an his Blog ‘K’s Law‘. (The Blog is dormant now because he became a member of the Boards of Appeal.) His comparison of assessment of inventive step with mountaineering is just great.

Further, one may well discuss for quite a while in step iii) what a skilled person would actually have done or not. The EPO’s could-would approach requires that

[…] the prior art as a whole […] would (not simply could, but would) have prompted the skilled person, faced with the objective technical problem, to modify or adapt the closest prior art while taking into account of that teaching, thereby arriving at something falling within the terms of the claims, and thus achieving what the invention achieves.

The gist is that hindsight should not be an issue if there is a pointer or an objective incentive to combine two pieces of prior art. Fine. But …

The definition of the ‘objective technical problem’

Everything hinges on the definition of the objective technical problem in step ii), and this is what this Post is all about. According to the EPO Guidelines (G-VII, 5.1),

[…] one establishes in an objective way the technical problem to be solved. To do this one studies the application (or the patent), the ‘closest prior art’ and the difference (also called ‘the distinguishing feature(s)‘ of the claimed invention) in terms of features (either structural or functional) between the claimed invention and the ‘closest prior art’, identifies the technical effect resulting from the distinguishing features, and then formulates the technical problem. […]

It is noted that the objective technical problem must be so formulated as not to contain pointers to the technical solution, since including part of a technical solution offered by an invention in the statement of the problem must, when the state of the art is assessed in terms of that problem, necessarily result in an ex post facto view being taken of inventive activity (see T 229/85).

That’s how I’ve learned to play the game. A plain and simple mechanism:

  • identify the distinguishing feature (e.g. bolt and nut instead of a nail);
  • identify the technical effect of the distinguishing feature in the context of the invention (e.g. ease of detachability)
  • define the objective technical problem based on the technical effect of the distinguishing feature (e.g. facilitate detachability)

Does O2015_008 lower the bar?

When re-reading the judgment O2015_008 of 14 March 2018 (commented on this Blog here), I stumbled upon a general remark in ¶66 that I had overlooked before. It reads as follows (translated from French):

Nevertheless, the question must be asked whether the objective problem that is used for the problem-and-solution approach […] finds a basis or motivation in the closest prior art document or is part of the standard problems for the skilled person, and does not incorporate elements of the proposed solution into the patent (cf. T 59/90, r. 8). This is important to avoid hindsight bias.

(A reasoning along these lines, but in less explicit terms, can also be found in ¶33 of the decision O2015_018 of 15 June 2018, with the very same reference to T 59/90; see this Blog here.)

First, I just don’t know for sure what is meant with ‘standard problems’; but maybe this refers to e.g. ‘less expensive’, ‘faster’, ‘more efficient’ or similar things.

But be that as it may, I feel that the whole approach is flawed.

Why should it be required to have a basis or motivation for the ‘objective technical problem’ in the ‘closest prior art’?

Let’s take a simple example:

The invention is a car with an allegedly clever engine — but there is a lot of other stuff recited in the claim, e.g. wheels, mirrors, bumper, a radio, etc.

Obviously great

The ‘closest prior art’ shows almost everything, in particular the allegedly clever engine. So this really is a reasonably chosen ‘closest prior art’. In fact, it is almost novelty destroying. It is only silent about the radio.

Car radios are perfectly known, but in no way mandatory. Thus, there clearly is no implicit disclosure of a radio in the ‘closest prior art’, and one has to deal with it as a matter of obviousness.

In my perception, a proper definition of the objective technical problem will somehow need to address entertainment and/or information in a car, while carefully avoiding to mention a radio.

The alleged invention correctly just cannot survive the problem-and-solution approach because there are a lot of car radios in the prior art as a whole, and the skilled person would surely have integrated them as a straight-forward solution to the objective technical problem to be solved.

But I am not so sure about this if the reasoning in ¶66 of O2015_008 was correct. To be frank, I don’t even have a clue how the objective technical problem should then be defined. There just is no ‘basis or motivation’ towards entertainment or information in that document.

Alternatively, let’s start from a car with wheels, mirrors, bumpers and a radio as the ‘closest prior art’, and then try to combine it with the (not anymore) ‘closest prior art’ showing the engine. Assume the clever engine is just shown somewhere in the smallprint, without any indication of advantages or further explanations. It’s just there, and it’s not immediately apparent why it actually is clever. You will then have a hard time to hunt this April Fool’s joke of an ‘invention’ down with the problem-and-solution approach, in the absence of any pointer towards the invention, i.e. to use  the engine from the prior art in a car with wheels, mirrors, bumper and — believe it or not — a radio.

Frankly, this cannot be it. I feel there must be a bug in a test that does not reliably sort out such non-inventions.

Struggling with this, I sought for guidance in T 59/90 (¶8) referred to in ¶66 of O2015_008. But there is none. It might be the last paragraph of ¶8 that appears to support the view taken in ¶66 of O2015_008, at least at first glance. It reads as follows:

Daraus muss geschlossen werden, dass aus [dem nächstliegenden SdT] das der patentgemässen Aufgabe zugrundeliegende Problem nicht bekannt war, und demzufolge auch die Lösung der Aufgabe aus [dem nächstliegenden SdT] in Verbindung mit dem allgemeinen Fachwissen nicht nahegelegt gewesen sein kann.

Loosely translated:

It must be concluded that the actual technical issue underlying the objective technical problem was not known from [the closest prior art], and therefore the solution of the problem could not have been obvious from [the closest prior art] in connection with the general technical knowledge.

However, the Board did not deal with the formulation of the objective technical problem here. It took the objective technical problem as it had been properly defined beforehand, and then summarized its assessement according to step iii), i.e. the could-would approach.

At closer scrutiny, it turns out that all this was only intended to sum up the reasons in respect of a single attack based on the ‘closest prior art’ per se, only in further view of the common general knowledge of the skilled person. Thus, it had essentially been argued by the opponent that the ‘closest prior art’ in itself would render the claimed subject-matter obvious for the skilled person. In that case, I might agree that the ‘objective technical problem’ needs to have a basis or motivation in the ‘closest prior art’ because it is not combined with yet another document but rather is considered self-contained, only seen with the eyes of the skilled person (common general knowledge). Where else should the pointer come from, if not from the ‘closest prior art’ in such a case?

But as soon as the ‘closest prior art’ is considered in further view of yet another document, the basis or motivation may well be derived from that further document.

And this is exactly what has been done in T 59/90. The Board moved on to assess in ¶9 whether there is any incentive towards the solution of the objective technical problem in any other prior art document:

It remains to be checked whether a suggestion for the solution of the objective technical problem can be derived from any of the other documents on file.

Note that the Board did not touch the wording of the ‘objective technical problem’ when moving on to check the combinations with the other documents on file. This is the problem-and-solution approach; nothing more, nothing less.

Further, I strongly doubt that this decision is of much relevance at all. It had been only distributed to Chairmen of other Boards (classified as ‘C’), and it has been cited by other Boards only once in 25 years (T 1009/96), for an unrelated aspect. And it is not that this Board had cited it by its own volition. Rather, it had to bother with T 59/90 because a party brought it up. It had been argued that a piece of prior art cannot be considered as ‘closest prior art’ at all because it did not deal with the specific problem of the invention. But this is not my point here, and the argument failed anyway. What is more, T 59/90 is not even mentioned in the reference book ‘Case Law of the Boards of Appeal‘. Thus, the decision has effectively been ignored by other Boards for a quarter of a century. Rightly so, because there is nothing special in it.

Concluding, I don’t see that T 59/90 supports the view taken in ¶66 of O2015_008 (except for the special situation where lack of inventive step is argued based on a single reference, i.e. the ‘closest prior art’ per se), and the above example shows how it would almost inevitably lead to obviously wrong results. In my perception, there is no need to spice up the well established framework of the problem-and-solution approach with yet further requirements. Rather, it can and should be applied as set forth in the Guidelines, for that case law remains predictable. And I have no reason to doubt that it leads to ‘right’ decisions — if applied correctly.

UPDATE 26 Sep 2018:

Decision T 910/90 states the following in ¶5.1 (and I fully concur with this):

Dabei kommt es nicht darauf an ob diese Aufgabe bereits im nächstkommenden Stand der Technik angesprochen ist, sondern darauf was der Fachmann beim Vergleich des nächstkommenden Standes der Technik mit der Erfindung als Aufgabe objektiv erkennt.

Loosely translated:

It does not matter whether this problem is already addressed in the closest prior art; rather, it is decisive what the skilled person objectively recognizes as the problem when comparing the closest prior art with the invention.

Reported by Martin WILMING


Fullscreen view (new tab)

T 59/90

Fullscreen view (new tab)


¶33, translated from French:

However, [the definition of the objective technical problem according to the problem-and-solution approach] is only valid insofar as the technical problem thus determined can reasonably be presumed to be a problem that arises without inventive effort within the general framework of the closest prior art document. Otherwise, this approach would focus the disclosure of the closest prior art and its interpretation in the direction of the invention, and introduce hindsight. (cf. T 59/90, ¶8)


Enter your name and email address below to get notified of new posts by email.

BREAKING: Non-enabling disclosure as closest prior art

Supreme Court
Case No. 4A_541/2017 | Decision of 08 May 2018 | ‘Appeal against decision O2015_011 of 29 August 2017

As to the background of this decision, I suggest to first have a look at the review of the first instance judgment on this Blog here.

In (very) brief, AstraZeneca’s EP(CH) 2 266 573 B1 had been revoked for lack of inventive step over Howell in view of McLeskey. The parties heavily dissented on whether or not Howell contained an enabling disclosure or not. The FPC held that it did, and that the objective technical problem to be solved in accordance with the problem-and-solution approach was only to provide an alternative formulation for sustained release.

AZ appealed and essentially argued that the FPC erroneously assumed a (concrete) pharmaceutical formulation in Howell, thus incorrectly defined the distinguishing features and the (objective) technical problem, and then wrongly concluded for lack of inventive step.

The Supreme Court indeed agreed and remitted the case for re-assessment of inventive step on the basis of an ‘objectively correct’ definition of the problem to be solved; see ¶2.3.5. Further, the Supreme Court notes in ¶2.3.3 that a correct definition is:

Provision of a castor oil-based formulation (i.e. a formulation consisting of castor oil and known suitable solvents and adjuvants in a certain composition and amount) for the administration of up to 250 mg fulvestrant for the treatment of breast cancer, which is well tolerated, has a uniform release profile and achieves the therapeutically decisive concentration of fulvestrant in the blood plasma over a longer period of time.

It is rare (to say the least) that the Supreme Court re-defines the objective technical problem to be solved in the assessment of inventive step. It did so here because the FPC got the meaning of an enabling disclosure wrong:

[…] weil [die Vorinstanz] von einem unzutreffenden Begriff der Ausführbarkeit der Lehre ausgegangen ist […]

What the heck …?!

Really? The FPC got such a fundamental point of law just wrong!?

This demands for a closer look!

The Supreme Court holds that the FPC failed to recognize the concept of feasibility or sufficient disclosure with the conclusion that ‘the technical teaching of [Howell] could basically be reworked.’ For a technical teaching consists not only of the problem but also of the solution.

Die Vorinstanz hat den Begriff der Ausführbarkeit oder der hinreichenden Offenbarung verkannt mit dem Schluss, dass ‘die Lehre [in Howell] grundsätzlich nacharbeitbar war’. Denn eine technische Lehre besteht nicht nur aus dem Problem, sondern auch aus der Lösung.

However, Howell only suggests that there was a compatible and pharmaceutically effective depot formulation available — but the actual composition of this formulation is not disclosed.

Further, the Supreme Court holds that the FPC did not expand on which concrete formulation the skilled person(s) would have found on the basis of the information in Howell without undue burden and without involvement of an inventive step. The general knowledge that steroids such as fulvestrant can be dissolved in castor oil with certain excipients and solvents in such a way that compatible injections can be produced is not sufficient for the feasibility of a technical teaching, in the Supreme Courts’ view.

In sum, the Supreme Court thus holds that the FPC applied an incorrect legal concept of feasibility with the assumption that no concrete technical formulation was required in order to affirm that there well is a technical teaching in Howell; ¶2.2.4.

Die Beschwerdeführerin rügt im Ergebnis zu Recht, dass die Vorinstanz von einem unzutreffenden Rechtsbegriff der Ausführbarkeit ausgegangen ist mit der Annahme, es bedürfe keiner konkreten technischen Formulierung, um die Offenbarung der technischen Lehre — d.h. hier der in [Howell] beschriebenen Depot-Formulierung — bejahen zu können. […] Die Vorinstanz hat die Offenbarung einer technischen Lehre durch [Howell] zu Unrecht bejaht.

Further down the road, the Supreme Court concluded that the FPC failed in the identification of the differentiating features. It just could not do it correctly because there is no specific disclosure in Howell.

But still, Howell undoubtedly is pre-published. Now, what to do with it?

The FPC had left it undecided whether or not a non-enabling disclosure is ‘prior art’ under Art. 54(2) EPC. But the Supreme Court took over — and answered even more than that. The Supreme Court holds that Howell belongs to the state of the art (contrary to what the EPO typically does, i.e. to just ignore it in toto; see Guidelines, G-IV, 2). But, in the Supreme Court’s view, such a piece of prior art cannot be assessed with the problem-and-solution approach because the objective technical problem is then just to find a working solution to what is insufficiently disclosed therein, and there is an inherent motivation to search for that solution; ¶2.3.2.

Der Beschwerdegegnerin ist zwar zuzustimmen, dass die Information von [Howell], in der das Problem formuliert und Ansätze für die Lösung mitgeteilt werden, zum Stand der Technik gehört. Wird jedoch ein Dokument als ‘nächstliegender Stand der Technik’ beigezogen, das keine technische Lösung offenbart, […]  wird der ‘Aufgabe-Lösungs-Ansatz’ verlassen. Denn die objektive technische Aufgabe ergibt sich unmittelbar aus [Howell], wenn darin dem fachkundigen Adressaten mitgeteilt wird, dass es eine Lösung für das Problem gibt […]. […] Eines Vergleichs von Merkmalen zur Ermittlung der objektiven technischen Aufgabe bedarf es in diesem Fall nicht. […] Aufgrund der Information in Howell, dass es eine Formulierung [gibt], wird das Fachteam jedenfalls dazu motiviert, geeignete Depotformulierungen zu suchen.

Frankly, it is now me who is not (yet) enabled to fully grasp the implications of this decision. Time will tell, as always.

Reported by Martin WILMING


Supreme Court
Case No. 4A_541/2017 | Decision of 08 May 2018 | ‘Appeal against decision O2015_011 of 29 August 2017

AstraZeneca AB
Actavis Switzerland AG

Panel of Judges:

  • Dr. Christina KISS
    • Dr. Kathrin KLETT
    • Dr. Fabienne HOHL
    • Dr. Martha NIQUILLE
    • Marie-Chantal MAY CANELLAS

Court Clerk:

  •  Dr. Matthias LEEMANN

Representative(s) of Appellant / Defendant:

  • Dr. Michael RITSCHER (MLL)
  • Dr. Kilian SCHÄRLI (MLL)

Representative(s) of Respondent / Plaintiff:


Fullscreen view (new tab)


Fullscreen view (new tab)


Fullscreen view (new tab)


Enter your name and email address below to get notified of new posts by email.