EP 3 744 649 B1 relates to a conveyor device for conveying flexible containers hanging in a vertical position along a packaging line including one or more filling units and a sealing unit.
The aim is to have a structure in which the linear guide (8) is perpendicular to the vertical axis of the left articulation (6a) and to the vertical axis of the right articulation (6b), so that the two arms (3a,3b) stay parallel even in a curved position of the packaging line. The container is fixed between the two arms.
Brief outline of the case
The OD held that claim 1 as granted lacked IS over D1/D1t=JP 2018172140 (A in ESR)+ D2=US 20160176659 (Not in ESR). and maintained the patent according to AR1.
The board held that claim 1 as granted indeed lacked IS, but the same applied to claim 1 as maintained.
The board remitted to the OD for further prosecution.
The decision is interesting in view of the definition of the problem to be solved and the combination of pieces of prior art relating to different domains.
The proprietor’s point of view
The proprietor emphasized that, the differentiating feature over D1 has the following effects:
– easily adapting to different sizes of flexible containers,
– opening and closing the flexible containers during motion, and
– maintaining parallelism between the gripper arms throughout the entire path, including straight, curved and transitional sections.
These effects should therefore be reflected in the problem to be solved, which should be formulated as
“how to provide an alternative parallelism-keeping mechanism, wherein the conveyor device can quickly adapt to different sizes of flexible containers and being adapted to open and close said flexible containers during motion and either in a straight, curved or transition section of the path”
When discussing the combination with D2 the proprietor contended that the skilled person aiming to solve the problem formulated would not even consider the teachings of D2, as it belongs to a remote technical field.
The OD also failed to take the teaching of paragraph [0016] of D2 into account.
This paragraph clearly states that it is preferable that the gap between the left and right movers remains constant.
This teaching makes clear to any skilled reader that the parallelism keeping mechanism of D2 is inherently incompatible with the conveyor device of D1.
This is not surprising at all, because as the conveying device of D2 is for conveying rigid boxes, this document cannot provide a solution which would be used by a skilled person in the field of conveying flexible containers, where being able to open and close the container during conveying is a mandatory requirement.
The board’s decision
On the objective technical problem
The board held that the various effects mentioned by the proprietor are not achieved. The proprietor has not shown any passage of the patent in suit, and none is apparent to the board, from which it can be derived that the effects are directly and causally related to the particular parallelism-keeping mechanism claimed, which is the only distinguishing feature of claim 1.
Furthermore, contrary to the patent proprietor’s view, the board is not convinced that claim 1 implicitly comprises the feature of opening and closing the flexible containers “during motion”.
The fact that the device of claim 1 has to be suitable for flexible containers does not necessarily imply that it has to be conceived for opening and closing the flexible containers “during motion”.
The board held that, in case of filling flexible containers with liquid product, as in D1, which is not excluded from claim 1 of the MR, filling is best performed when the conveyor is stopped, not during motion. As a consequence, not only can opening and closing the flexible containers “during motion” not be seen as implicit in claim 1, but it also cannot be seen as a technical effect of the claimed subject-matter.
The effects of easily adapting to different sizes of flexible containers and opening and closing the flexible containers are also achieved in D1 such that they cannot be related to the distinguishing feature.
Finally, with respect to the technical effect mentioned by the proprietor of maintaining parallelism between the gripper arms throughout the entire path, including straight, curved and transitional sections, the board considers that it only reflects what a “parallelism-keeping mechanism” aims to do.
As a consequence of the above, the board was of the opinion that the problem to be solved has been correctly formulated in the decision under appeal as:
– how to provide an alternative to the parallelism-keeping mechanism disclosed by D1.
On the combination of pieces of prior art
The board was not convinced that the skilled person would have disregarded D2 because this document belongs to a different technical field, namely the field of conveying rigid boxes.
The fact that D2 deals with a conveyor device for rigid boxes would not prevent a skilled person from taking its teaching into account because, as the technical problem is to find an alternative parallelism-keeping mechanism, the skilled person would concentrate on this detailed aspect and specifically look for parallelism-keeping mechanisms suitable for being used in the device of D1.
For the board, the skilled person, looking for an alternative parallelism-keeping mechanism for the conveyor device of D1 would immediately realise, from figures 1 and 2 of document D2, and from paragraphs [0029] to [0032] of the description thereof, that D2 does not belong to a remote technical field as alleged by the proprietor, but rather to a neighbouring technical field to that of D1 and claim 1 of the MR, since it also discloses a conveyor device for conveying containers along a packaging line using gripper arms.
In addition D2 also deals with the problem of keeping arms parallel at all times while holding containers during conveying, in particular along curved paths.
D2 in fact discloses a mechanism keeping two arms mutually parallel even when the distance between the left and right articulations is varied due to the relative movements of the left and right movers during transport along both the straight and curved sections of the path, de facto including also transitional sections.
Comments
It might be tempting for a proprietor to “charge” the OTP, as it will then be more difficult to attack the claim for lack of IS.
If the proprietor does so, he will have to show that those effects are disclosed, at least implicitely, in the application as filed.
The present decision is a good confirmation of T 176/84, OJ 1986,50. In this decision it was held that, when examining for IS, a skilled person would consider the state of the art in the specific technical field of the application, look for suggestions in neighbouring fields or a broader general technical field if the same or similar problems arose, and if he could be expected to be aware of such general fields.
Whether the containers are flexible or rigid, maintaining the parallelism of the arms holding the containers, also along a curved track, is an absolute necessity.
It follows that the skilled person would combine the teaching of D1, for flexible containers, with that of D2, for rigid containers.
In view of the Sanofi/Amgen and Edwards/Meril decisions of the UPC, commented in the present blog, it could well be expected that the UPC would have considered claim 1 as granted as being inventive. There was no pointer in D1 to look for an alternative solution, as the problem of keeping the bags between the two arms even in a curved section I already solved in D1.
On the procedure
The ESR mentioned three documents of category A.
The patent is classified, inter alia in B65G54/02. D2 bears the same IPC classification.
It is thus surprising that D2 was not found during the European search.
Comments
37 replies on “T 2044/23-On the OTP and the consultation of neighbouring domains”
I can see that “finding an alternative” solution might (depending on the nature of the claimed subject matter) strongly motivate a skilled person to look in neighbouring technical fields. But in general I am wondering how “technical” the problem of “finding an alternative” really is. I mean, finding an alternative only becomes a problem if the teaching in D1 is not free to adopt. If the D1 solution is patented, for example. If the objective problem is to avoid infringing a patent, that is a legal and commercial problem, for sure, but is it a technical one?
I guess the EPO’s jurisprudence is well-settled on this issue but I am both curious and lazy, so I am asking you, Daniel. Otherwise, I am wishing you and your family a good Christmas, and all the best for 2026.
Answering my own question I would say that it is irrelevant whether the prompting is legal or commercial, just so long as what you end up with is a problem in technology. Such as to find a different machine design.
@MaxDrei
I share your doubts. Why make it such a big deal to define the “objective technical problem” when at the end of the day, the definition can be as vague as the search for an alternative ? For the “could vs. would” analysis, there must be some specific content in the definition of the problem.
Mr Hagel,
From various comments you have made here and in other blogs, I allow myself to conclude that you do not seem to be truly convinced by the problem-solution approach.
It is an acquired fact from the case law and the practice of the EPO that, in the absence of a different effect with respect to the CPA, the “objective problem” at the EPO, remains to find an alternative. As I have explained to Max Drei, the problem remains technical, and this is good enough.
If it needs a pointer to come to the solution of a technical problem as required by the UPC, then alternatives would not be patentable, as it can be decided that the original problem, i.e. the “objective problem” for the UPC, is already solved and the skilled person has no incentive to look for another solution. I am not convinced that this is the right approach.
For the surplus, any feature which is not technical cannot justify IS. I am still waiting to see what the UPC will say on this topic.
Mr Thomas,
Thank you for your wishes. I wish you and your beloved ones the best for 2026, and I look forward to other lively exchanges.
To answer your doubts, I am « truly convinced » by the principle of the PSA and by the EPO’s willingness to mitigate hindsight in the assessment of IS. But as to the implementation, I am not always « truly convinced ». It depends. I am not convinced when the PSA is a three-step sequence with the selection of a CPA as the first step. It is my view, and I have advocated it for several years on the blogs, that the first step (or a preliminary to the usual three-step sequence) should be the definition of the « art » of the skilled person.
In this respect, I have noted with great interest the addition to the CLR in the July 2025 version of sections I.D.3.2 « Approaches to selecting the closest prior art » and I.D.8.1.2 « Point at which the skilled person enters the stage3, which signals a major divide between the BOAs precisely over this issue.
The CLR refers to a first approach in which it is for the deciding body to select the closest prior art, and to a second approach in which he selection relies on the skilled person. In respect of that second approach, the CLR emphasises the « firm conviction in T 2759/17 that, in line with the established case law, the skilled person was the relevant point of reference right from the start of any inventive-step assessment » and states that « excluding the skilled person for part of the inventive-step assessment would lead to artificial and thus technically meaningless results. »
The CLR seems to find quite convincing, if not endorse, the reasoning of the board in T 2759/17 when it explains : « The board
found that unlike the first approach, the second approach was based on a technically meaningful and thus realistic scenario. More specifically, the skilled person was normally confronted with a certain purpose or effect to be achieved in a certain technical field. With this in mind, the skilled person would then look for a prior-art disclosure that was in the same technical field and aimed at the same or a similar purpose or effect. »
The CLR mentions that « further decisions in which the boards of appeal at least implicitly indicate that the skilled person may select the closest prior art include T 1841/11, T 2057/12, T 1248/13. » There are more recent decisions along this line, such as T 900/21, T 1760/22, T 749/23.
It seems section I.D.3.2, with its description of a first and second approach and the list of decisions having used the second approach, is worded as a bellwether to an EBA referral. Time will tell. This was a request for a referral in T 749/23 but it was dismissed.
Mr Hagel,
Thanks for your comments, which I read with great interest.
As far as the book CLBA is concerned, I would like to add that, whilst it is edited by a department within the BA, it is not vetted by those or by the presidium of the BA. After all, the CLBA represents the opinion of the drafters but not necessarily that of the BA as a whole.
Nevertheless, it is correct that in Reasons 5.3.1 and 5.3.2 of T 2759/17, the board acknowledged that it was possible to distinguish two approaches for the selection of the CPA.
In the first approach it is the deciding body which decides on the selection of the CPA. The skilled person does not come into play until later on when the CPA has been identified and the problem formulated. This approach implies that each and every disclosure within a document can be selected as the starting point for assessing IS.
In the second approach, the skilled person comes into play as early as when the closest prior-art disclosure is being selected. Under this approach, the skilled person is held to generally look for a disclosure that aims at a purpose or effect that is the same as or at least similar to the one underlying the patent in question.
It is therefore possible to consider, at a glance, that there is an apparent divide between boards.
However, in the end a realistic starting point, be it chosen by the deciding body or by the skilled person, is a disclosure that aims at a purpose or effect that is the same as or at least similar to the one underlying the patent in question. This is a well acquired fact in the application of PSA. Iin my opinion, the distinction between the two approaches is thus more academic than based on technical reality.
The deciding body will generally not chose as starting point which has a purpose or effect that is not the same as or at least not similar to the one underlying the patent in question.
In T 749/23, Reasons 46, the board held that if IS is to be acknowledged, the claimed subject-matter must be inventive starting from any starting point in the prior art. Conversely, if the invention is obvious to the skilled person from at least one starting point, an inventive step is lacking.
In T 749/23, Reasons 47, the board added that there is also no requirement that an embodiment must be exemplified for it to be considered in the assessment of IS. There is, likewise, no requirement that the prior art contain any pointer, suggestion or incentive to select a particular embodiment for further development for this embodiment to qualify as the starting point in the PSA. In this respect the position of the BA is different from that of the UPC.
As the patent subject-matter of T 749/23 was eventually revoked on the basis of the prior art available, it was “not decisive in the case in hand which approach is taken to determine the starting point for the inventive step analysis”, cf. Reasons 79+80.
Section I.D.3.2 of the CLBA 11th edition, 2025,, also refers to T 1450/16, in which the Headnote says the following:
“In the application of the problem-solution approach for the assessment of inventive step, the person skilled in the art within the meaning of Art 56 enters the stage only when the OTP has been formulated in view of the selected “CPA”. Only then can the notional skilled person’s relevant technical field and its extent be appropriately defined. Therefore, it cannot be the “skilled person” who selects the closest prior art in the first step of the PSA. Rather, this selection is to be made by the relevant deciding body, on the basis of the established criteria, in order to avoid any hindsight analysis.” T 1450/16 is clearly in favour of the first approach.
Since in my opinion, the result is the same whatever approach is taken a referral to the EBA to decide which approach should be taken appears superfluous.
I would also like to draw your attention to T 1078/23, commented in the present blog on the choice of the CPA when IS is to be denied. In T 1078/23, the board referred to T 1450/16 and agreed with its conclusion. In my comments, I added that it would be interesting if the EBA would decide which of the two approaches for selecting the CPA should be adopted, or even if both were acceptable.
In view of the present developments, I do not think that a referral to the EBA is necessary.
In its decision UPC_CFI_1/2023, point 8.6 (Sanofi/Amgen), the CD Munich held that the assessment of IS starts from a realistic starting point in the prior art. There can be several realistic starting points. It is not necessary to identify the “most promising” starting point. In decision UPC_CoA_528-529/2024, the CoA UPC decided that the patent was valid, i.e. reversed the decision of the CD Munich.
The position of the UPC appears different than the position of the boards. For the boards, it is only necessary to assess IS from a plurality of starting points if IS has to be acknowledged. If IS is denied, it is not necessary to look at different starting points. One route allowing to deny IS is enough.
Mr Hagel, we differ. My interest is in fixing what is “technical” and what is not. I have no problem with defining the objective problem as that of finding an alternative (provided that both the objective problem and the claimed solution are “technical”). Often, the claimed alternative is obvious, but sometimes it is not. That is all fine with me.
I like it, that your viewpoint, mine and that of DXT are not the same.
Happy Christmas to you too, and all the best for 2026.
Dear Max Drei,
My answer is very simple: any patent is a technical solution to a technical problem, and if it is solved by an alternative it remains, as such, technical and is neither a legal nor a commercial problem.
It should not be forgotten that the purpose of the patent system is to foster technology by disclosing teachings. If an alternative teaching is available and patentable, the patent system has done its job.
I thus agree with your own reply. It is indeed irrelevant whether the prompting is legal or commercial. I simply tried to find more flesh for the argumentation.
Circumventing an existing patent might prima facie look as a legal or commercial problem, but as long as it contributes to the evolution of technique it might be worth a patent.
Let me also wishing you and your family a good Christmas, and all the best for 2026! I am looking forward to continue our discussions.
It occurs to me that there is a category difference between “finding an alternative” and other OTP’s. With most OTP’S, it takes invention to find even one solution to the problem. But finding an alternative (motor, pump, or toothbrush design) requires no invention at all. One can create thousands of alternatives without even thinking at all.
But that is all irrelevant. What matters is whether the alternative that you are claiming is or is not obvious. And that we can decide, efficiently, reliably and accurately, using EPO-PSA.
Perhaps it is this “category difference” which makes people leary of the “alternative” line in EPO case law?
Dear Max Drei,
I fully agree with your point of view: “What matters is whether the alternative that you are claiming is or is not obvious. And that we can decide, efficiently, reliably and accurately, using EPO-PSA”.
I would think that lots of people think that finding an alternative might be obvious. This is by no means the case, and it should not be forgotten that, in the PSA, finding an “alternative” is defined in view of the OTP and the CPA, and not from the “subjective problem” defined by the applicant.
When the UPC finds it necessary to actually call the “subjective problem” defined by the applicant the “objective problem”, it shows clearly that it dislikes the PSA and actually wants to create confusion.
Daniel, I wonder whether you have grasped what I am trying to get across with my “category difference” point.
You write that it is by no means the case that finding an alternative is obvious. Now hang on a minute, please. Not everything that is new is inventive. There is near-infinite scope to design a “new” consumer product without that new design being patentably inventive. So when patent lawyers with no EPO experience hear that, in their specific case, their precious claim, the OTP is to “find an alternative” their knee jerks, their eyebrows jump up, and (in the manner of John McInroe) they scoff “Umpire, you cannot be serious!”
Do we agree? I hope so. How many in the UPC litigator and judge community have a career history which includes serious day-to-day experience of EPO-PSA? My guess: not many.
Dear Max Drei,
I never wanted to give you the impression that finding an alternative is obvious. Some alternatives are obvious, some are not. It depends on the specificity of the case.
I do however follow you when you consider that patent lawyers with no EPO experience can get agitated when they hear that the OTP is to find an alternative.
At the UPC the OTP is the problem the inventor wanted to solve, irrespective of the prior art revealed by a documentary search.
Diverging decisions between the UPC and the BA seem programmed.
Nice, that this thread about the CPA continues on, past the Christmas holidays. Having read the new comments, some thoughts arise in my mind. Do both EPO and UPC hold that the obviousness enquiry shall start from a single prior art source? Do both hold that the only valid candidates for this CPA are sources that are “realistic” starting points? Is there sometimes heated debate between the litigants, whether or not any given prior art candidate qualifies as “realistic”? Is the outcome of the enquiry whether the nominated CPA is “realistic” sometimes decisive to the entire enquiry into obviousness?
If the answer to all these questions is “Yes”, then it seems to me that the question who shall set the CPA (tribunal or skilled person) is otiose, a mere distraction, a debate that brings us no further forward in the enquiry whether the claimed subject matter was obvious to the notional “person skilled in the art”.
In other words, I am sceptical that the case law of the Boards of Appeal reveals two different approaches to the task of fixing of the CPA. I suspect that the Boards might actually agree amongst themselves that there is in fact only one approach, one which they all follow.
Dear Max Drei,
As I have explained to Mr Hagel, it might look, at a glance, as if there were two ways to determine the CPA at the EPO. But it is not sure at all.
At the end of the day, what matters is that the CPA represents a “realistic” starting point, for instance, that it is a disclosure that aims at a purpose or effect that is the same as or at least similar to the one underlying the patent in question. This is a well acquired fact in the application of PSA. In my opinion, the distinction between the two approaches is thus more academic than based on technical reality.
I therefore share your scepticism that “the case law of the Boards of Appeal reveals two different approaches to the task of fixing of the CPA”.
The UPC and the boards have a similar definition of the “realistic” starting point: it is a disclosure that aims at a purpose or effect that is the same as or at least similar to the one underlying the patent in question.
There are however differences between the UPC and the boards of appeal. The UPC does not consider that there is no CPA, all realistic starting points have to be taken into account. When denying IS, it is sufficient that one “realistic” starting point exists at the EPO. The notion of OTP is fundamentally different at the UPC and at the EPO. At the UPC it is the problem the inventor wanted to solve. At the EPO it is the problem resulting from the effect of the difference of the claimed subject matter and the CPA. At the UPC a pointer is necessary in the prior art to arrive at a lack of IS.
The UPC and the EPO claim officially to be in favour of harmonisation: I wonder which type of harmonisation they are thinking of?
For the UPC it is simple: the decision of the instance having decided first should be taken over by the instance deciding subsequently. In view of the procedural time limits at the UPC, the latter is sure to win every time. What the UPC wants is not harmonisation, it is jurisdictional imperialism. And who will be the beneficiary? Follow the money….
OK, Daniel, the OTP is different. But let’s stay on the notion of the prior art starting point. My interest is in fixing what is or is not “realistic”.
I have always assumed that the tribunal gives full faith and credit to what the patent declares to be the technical field and the subjective problem addressed by the inventor. What prior art starting point is “realistic” for attacking that specific problem?
The search results might change the notional problem into a different one, namely the truly “objective” technical problem, but does one then circle back and ask whether the already fixed prior art starting point is no longer realistic, having regard to the newly fixed objective problem?
Might the search for a starting point with more destructive potential, one that was originally not “realistic”, then prompt another shift in the definition of the OTP? How much circling back is permissible?
Dear Mr Thomas, dear MaxDrei,
With my answers to your last comments, maybe this thread is going to make it to 2026 !
To Mr Thomas :
I have been surprised by your reservations regarding the emphasis in the 2025 CLB edition on the existence of two different approaches for the selection of the CPA. You suggest that the CLB is drafted by a specialised group and its content has not been vetted at the highest level. But the foreword to this edition has been signed by Mr Josefsson, who insists that the CLB is « an integral and highly valued resource for the work of patent attorneys in particular, the judiciary in the EPC contracting states and beyond, as well as in-house and outside counsel ». This clearly means that Mr Josefsson assumes full responsibility for its content.
In addition, the fact that the divide between the Boards has been presented by the addition of entirely new sections in the 2025 edition and described in detail makes it a high visibility issue. It is difficult to imagine that it has not been thoroughly vetted and personally approved by Mr Josefsson.
In my view, it makes full sense to have the skilled person enter the stage at the outset of the IS assessment, for a number of reasons.
This is consistent with the mention of the skilled person in Art 56.
The definition of the « art » of the skilled person delineates the segments of the prior art which the skilled person will consider as starting point.
This definition is necessary to determine the CGK of the skilled person. The CGK are a very significant element in the assessment of IS (and of other issues as well).
The « art » can be defined on the basis of the claims to be assessed and of the application as filed. The technical problem to be solved belongs to the « art ».
To MaxDrei :
The skilled person may include a non-technical person and the CPA may belong to the CGK of a non-technical person. This is exemplified in T 2101/12 and T 1049/19.
The distinction between technical and non-technical is now blurred as a result of the global digitisation of human activities.
Foreword to the eleventh edition
The first edition of the “Case Law of the Boards of Appeal of the EPO” (CLB) publication was printed in 1993, compiling six special editions of the Official Journal on important decisions of the Boards of Appeal up to that time. By then, the Boards of Appeal had decided on some 5 000 cases. More than three decades after the first edition of the CLB and almost 50 years since the first decision of the Boards of Appeal was issued, the Boards of Appeal have dealt with over 67 000 cases and issued over 45 000 decisions. The present new edition reflects almost 9000 important decisions from the Boards of Appeal and provides updates on the evolution of the case law through to the end of 2024.
The CLB has since become an integral and highly valued resource for the work of patent attorneys in particular, the judiciary in the EPC contracting states and beyond, as well as in-house and outside counsel. This was confirmed by the results of a survey conducted by the Boards of Appeal in 2024, in which 85% of patent attorneys replying to the survey stated the main reason for using the CLB was for the preparation of cases and/or applications.
Since the 10th edition in 2022, which was the first digital-only edition of the CLB, there have been many changes. Interim revisions of selected chapters were introduced for the first time in 2024, keeping users updated on the most recent and important changes between triennial editions. These included in particular the case law and approaches following Enlarged Board decisions as well as the developing jurisprudence on the application of the revised Rules of Procedure, which came into force in 2020. These interim revisions will continue on an annual basis and will be available in the HTML version of the publication, allowing for a more dynamic and up-to-date insight into the development of our jurisprudence.
Furthermore, our regularly published abstracts of decisions, which summarise key aspects of selected BoA decisions as they are published, are available every month on a dedicated webpage. On a quarterly basis, the abstracts are updated on the “New Decisions” sidebar in the relevant section of the HTML version. This enables users to read the abstracts in the context of the relevant chapter more regularly and stay updated on recent case law on a certain topic.
I wish to sincerely thank the Legal Research Service of the Boards of Appeal for their dedication and commitment to provide this important trilingual resource, not only every three years but to ensure more frequent updates for the user community and our key stakeholders. My thanks also go to the EPO’s Language Service for the translation of the text spanning over 2 000 pages, as well as the Publications and IT teams for their co-ordination and support to publish the 11th edition.
I hope you find it useful.
Carl Josefsson
President of the Boards of Appeal
Mr Hagel,
As the CLBA is actually devised by a department of the boards of appeal, it cannot be politically envisaged that the president of the boards of appeal does not say anything when a new edition of the CLBA is published. I would limit the influence of Mr Josefsson to this. My aim is not to belittle what Mr Josefsson says, but to put it in the right perspective.
For the rest, it is all flowery words, as most of what comes from the EPO. I think here especially about all the wonderful definitions on which aspects the EPO relies to show that the quality of the work delivered by the EPO is necessarily the best in the world. Words, nothing more.
I cannot say much more for fear of revealing my contacts, but I can assure you that as such, from information I did not get by hearsay, the CLBA represents the views of its drafters and not more. This is why I always advise not to limit the citation of a decision to what is said about it in the CLBA. Once selected in the CLBA, it is then of utmost importance to look at the actual decision. The CLBA is certainly very useful as first pre-selection means, but once a decision has been spotted in the CLBA, you better look at the decision itself.
I do not give much to the divide between the boards when it comes to decide what should come first, the art or the skilled person. To me the discussion is the same as to decide what was first: the hen or the egg.
As explained to Max Drei, it is the inventor/applicant which defines the scene. The CPA has to belong to the scene defined in the application/patent. Whether one decides first what the CPA is or what is the notional skilled person is in my view irrelevant or merely academic. By defining the art the inventor/applicant automatically defines the skilled person in the art.
Dear Max Drei,
A “realistic” starting point is a piece of prior art which stems from the same technical domain as that of the claimed subject-matter and which has same purpose or effect as the claimed invention or deals with a similar technical problem, and in the end has the most features in common with the claimed subject-matter. That it has the most features in common with the claimed subject-matter only comes at the end, and can never be used a sole selection criterion.
Just an example: a hair curler is a cylindrical object with a plurality of holes and a stent can also be considered as a cylindrical object with a plurality of holes.
If the claimed subject-matter is a hair curler, no sensible person would take a stent as realistic starting point, even if lots or features are common between the hair curler and the stent. If the claimed subject-matter is a stent, the same goes the other way round.
It is indeed the inventor/applicant which sets the scene. Unless for pioneering inventions, the inventor wants to solve a technical problem which is present and not solved in the prior art.
The prior art taken as “realistic” starting point is the “realistic” prior art known to the inventor/applicant knew before filing. This “realistic” starting point defines the original technical problem known in the EPO jargon as the “subjective” problem.
Now comes a fiction: the inventor/applicant is meant to have been aware of everything which was made available to the public before filing. Through a documentary search prior art can be found which is closer to the claimed subject-matter, taking into account the technical domain of the invention.
This prior art has in principle more features in common with the claimed subject-matter than the inventor/applicant started with.
This closer prior art becomes then the actual “realistic” starting point and replaces the “realistic” starting point from which the inventor started, and defines the “objective” technical problem which is then different from the “subjective problem defined by the applicant/inventor.
It can well happen that there is no better prior art available as “realistic” starting point. Then the “subjective” and the “objective” technical problem are identical. The claimed subject-matter lacks inventive step when there exists a piece of prior art, unknown to the applicant/inventor which allows to solve the technical problem underlying the invention.
I do therefore not see any need to iterate the search for a “realistic” starting point.
OK, Daniel. Thanks for your answers. They leave me thinking that I was seeing difficulties where none exist.
First, in order to fix the technical field and the technical effect recited in the specification, we do not need to invoke the concept of the person skilled in the art.
Second, the way that “realistic” is defined seems to be effective to shut out any of the “circling back” that I was imagining.
Thank you, fh, but I feel that I have not yet had a definitive answer to the question what is or is not a “realistic” starting point.
Me, I think the judge of what is a “realistic” starting point can only be the imaginary “person skilled in the art”. Ergo, you must first settle who is that person. Anybody disagree?
Dear Max Drei,
What was first, the hen or the egg? See my reply above.
I agree with Mr Thomas’ comment that the applicant sets the stage, in that the « art » of the skilled person is based on the claims and on the content of the application as filed but I think it is incomplete. : the claims often need to be interpreted before reaching a conclusion as to the definition of the art, and it is then for the « deciding body » to determine which must apply. To take up the hen and egg metaphor, you need both. The applicant lays the egg and the deciding body is the hen which brings in the interpretation supplement when needed.
To give you an example, it is frequent to find in claims the language « a device (or method) for « . This language may be interpreted in two ways : either as « a device intended for », or as « a device suitable for ». In the former case, the field-of-use indication is a limiting feature, in the latter it is not. The implication for the definition of the « art » is critical. In the former case, the field of use is the prominent element of the « art », in the latter case, the art is that of the generic device, irrespective of any specific field of use. This is a complete change.
On this question of interpretation, G 1/24 comes into play. It may be understood as requiring that the description be consulted in order to resolve the ambiguity. The relevant GL F-IV 4.13.1 may need to be revised for compliance with G 1/24.
Another question for which the definition of the art is helpful is the delineation of the CGK. The CGK are sometimes used by an ED to fill a gap in its line of reasoning leading to a negative IS assessment, without any supporting documentation. The CGK are sometimes trivial, sometimes not so much. A definition of the « art » would help address this issue.
Thanks for that, fh. My initial rreaction is that I find nothing to disagre with. But, meanwhile, DXT has persuaded me that the concept of the skilled person is not needed for the enquiry into what is or is not a “realistic” prior art starting point. I can’t tell from your comment whether or not you agree with DXT on that issue.
It might seem like a trivial question, but in the debate between the EPO and the UPC, how best to run an obviousness enquiry, it might make a significant difference, how costly the enquiry will be.
Dear Max Drei,
I allow myself to refer to my reply to Mr Hagel. I can only confirm that there is no need to refer to skilled person to define what is or is not a “realistic” prior art starting point.
I clearly disagree with Mr Hagel, especially in that the skilled person cannot be made dependent of the form the claim.
It is clear that at the UPC, litigation lawyers will make a song and dance to define the skilled person and the “realistic” starting point. It simply gives more billable hours to the client. How did you say recently: follow the money. I have nothing to add.
They will most probably not try to do so at the EPO as they would get a harsh, but deserved, reply: no need to take us for a ride, we know what we are talking about.
Mr Hagel,
At least you agree that the scene is set by the applicant and is determining when it comes to define the “art” at stake.
I do however have great difficulties in following you, when you consider that the “art” is depending on the form of the claim.
I beg to clearly differ on this point.
When it comes to IS, the state of the art to be taken into account, is that of the specific field of the invention, of neighbouring fields or general fields, see T 176/84, OJ 1986, 50 or T 195/84, OJ 1986,121.
In T 32/81, OJ 1982, 228, it has been held that the state of the art also includes prior art in a non-specific – general – field, dealing with the solution of any general technical problem which the application solves in its specific field. Such solutions of general technical problems in non-specific – general – fields has to be considered as forming part of the general technical knowledge which is to be attributed to those skilled persons versed in any specific technical field.
This means clearly that it is the applicant/inventor that sets the scene. Only from there the notional skilled person can be determined, and certainly not from the form of the claim.
Novelty is missing when the prior art device is “suitable for” the same purpose and has the same features as those claimed, without any purpose even being specified (photographic novelty).
Novelty is given when a prior art device is not suitable for the purpose. If it is not suitable for the purpose, it can also not represent the closest prior art, even if it has the same features as the claimed device. Just think of the example of a stent vs. a hair curler.
G 1/24 certainly comes into play and the relevant GL F-IV 4.13.1 will for sure need to be revised for compliance with G 1/24. Following the case law of the boards following G 1/24, I can well imagine that the Guidelines will say that what matters is the interpretation the skilled person will derive from the wording of the claim. If the applicant/proprietor wants the interpretation found in the description to be taken into account, the claim will need amendment, cf. T 2027/23.
This point of view is valid, whether the description allows a more limited interpretation than that of the claim. If the claim allows a limited interpretation of its subject-matter, but the description allows a broader interpretation, then the claim will be vetted against the broad interpretation found in the description.
I agree that the CGK is sometimes used by an ED to fill a gap in its line of reasoning leading to a negative IS assessment, without any supporting documentation. This is certainly not correct when the CGK is not properly documented. In T 3049/19, the board held that the opposition division committed a substantial procedural violation by using common general knowledge without any supporting evidence.
On the other hand, in T 2526/19 or T 1370/15 (opposition appeal proceedings) and in T 1090/12 (examination appeal proceedings), boards of appeal felt empowered to introduce common general knowledge into the proceedings without any supporting evidence. The catchword in T 1370/15 is very eloquent in this respect.
It remains that the definition of the “art” is set by the applicant/proprietor. If the subject-matter of the claim is a car headlight obtained by injection moulding, the notional skilled person has the CGK needed to optically define a car headlight as well as CGK in the field of injection moulding.
In my 50+ years in the patent business, I never shook hands with a skilled person, but I knew quite well from the subject-matter of the claim, who the skilled person was and what was his CGK. No need to interpret the claim, especially in view of the above definitions of the prior art to be taken into account when assessing inventive step.
Daniel perhaps there is less difference between you and fh than you both suppose. Take for example your stent/curler hypo suppose that my invention is a new node/strut pattern which yields improved mechanical properties in the apertured cylinder. I can choose whether to claim a cylinder, a stent or a curler. My choice fixes the technical field doesn’t it?
Mr Thomas,
Thank you for your comments.
In my view, stating that the applicant sets the scene implies as the primary component the claims in their form. Given the legal function of the claims, the form chosen by the applicant is decisive for the substance. In the case of a field-of-use indication, its form and its link to the claimed features are critical to decide whether it must be interpreted as a limiting feature and as a result, whether the « art » of the skilled person refers to the claimed field-of-use and the CPA must relate to the field-of-use.
From my discussions with EPO examiners following the audit of cases within the SQAP panels, I have noted that they are frequently reluctant to accept that field-of-use indications may be limiting. I must say this is also true for colleagues in private practice.
The mention of a field of use in the claims is a strategic decision for an applicant, whether it will be treated in effect as a limiting feature is therefore of utmost significance. I had pointed out this issue in an article published in Intellectual Property Magazine, July 2014, entitled « Suggestions for strategic drafting of patent applications », available on the blog of Jackie Hutter, a US colleague, at the link : https://thehuttergroup.com/strategic-patenting-guest-post/.
Of note on this topic, a recent decision of the French court de cassation (Lufthansa v. Astronics, 19 March 2025) which confirms the prevailing role of a field-of- use claim limitation (in the case, use in an aeroplane cabin for a voltage supply device). During examination of the patent in suit EP0881145, the ED had not paid any attention to the claimed field-of-use limitation in its first communication finding a lack of novelty.
I have smiled when reading that during your career, you had never shaken the hand of a skilled person. In my decade-long career as a corporate professional, I have shaken the hand of many technical people who were as close as possible to real-life persons skilled in the art. What I found striking in general was their extremely narrow specialisation, consistent with their supposed lack of inventive ability. This is why the emphasis on the field of use is realistic. This is quite the opposite of the highly developed mental agility featured by patent professionals and examiners, who constantly switch from a technical area to another.
DXT might well have a different answer, but what I took from his “never shaken the hand” remark was that the legal fiction called the “person skilled in the art” is deemed to know all in their specific field that has already been made available to the public. No real person can know it all. No real person would claim to know it all. that would be absurd.
That said, an AI can know it all. But can an AI possess “skill” in the art? And can you shake hands with an AI? Not yet, I suppose.
Dear Max Drei,
The skilled person is a notion used to sum up CGK in a given specific or in a more general field. It can even combine the CGK in two fields if the invention at stake can be considered belonging to two fields. Cf. my example of a head light obtained by plastic injection moulding. in this case CGK in optics and CGK in plastic injection can play a role.
As far as AI and CGK is concerned, I refer to T 1193/23, commented in the present blog. The board made clear that a chatbox does not replace the skilled person. Replies from an AI depend from the training data of the chatbot and of the question addressed to it.
The information used by a chatbot is based, at least in principle, on all the information available to it, including documents published well after the effective date of a claim. It is therefore of no use in defining the CGK and the interpretation to be given to a claim.
It will take a long time until we can shake hands with an AI representing the skilled person and its CGK.
Mr Hagel,
I read your comments with great interest, but still cannot agree with you.
Where a field of use is defined in claim, it first depends on the category of the claim.
In a claim to an action, the field of use is a limiting functional feature of the claim. It does however not avoid a lack of N if the prior art discloses a claim to an action with the same features, even without any indication of a field of use, or for a different field of use.
If it is a claim to an entity, the field of use is only limiting if the following features clearly imply specific characteristics linked to the intended use. See the example in the Guidelines of a container for molten steel or for ice cubes.
If the following features necessarily imply a specific use, the examiner is not entitled to ignore the field of use.
If the following features do not necessarily imply a specific use, the examiner is entitled to ignore the field of use.
Last but not least, as explained to you beforehand, G 1/24 has changed the issue. It also depends what is said in the description.
A claim to an entity indicating a given field of use, cannot be broadened in the description to any field of use, if the claimed features necessarily imply a limitation to this specific field of use.
A claim to an entity indicating a given field of use, cannot be limited in the description to this specific field of use, if the claimed features do not imply a limitation to this specific field of use.
In presence of a claim to an entity indicating a given field of use, one has thus to distinguish whether the indication of the field of use is limiting or not.
Irrespective of the above considerations, the problem to be solved by the claimed invention has to be indicated in the description, cf. R 42(1). It is thus the description which sets the art and not the claim.
I had a look at the case Lufthansa v. Astronics and at the decision of the Cour de Cassation. As electronic engineer, I do not think that the claim at stake is necessarily to be limited to a voltage supply apparatus “for providing a supply voltage for electric devices in an aeroplane cabin”.
The plug and socket system of the voltage supply system insures that the voltage is only delivered when the correct plug is properly inserted in the socket. The system is certainly useful in an aeroplane, but could as well be incorporated in the seat of various transport means like a bus or a train. The original search was therefore not limited to airplanes (B64) but to plugs and sockets in general (H01R).
In my opinion, the ED has, in this case, correctly not paid any attention to the claimed field-of-use limitation. The actual description of the patent refers to power supplies in an aeroplane at specific portions, but most of the description is applicable to any plug and socket system in general.
If a claim “for” is limited to a specific field of use, this sets a different skilled person in scene as when the claim “for” is not limited to a specific use, but the description cannot be ignored in either case.
The catchword in T 1370/15 might be eloquent but the decision is one of the worst of all time. Reasons 5.3.9 is egregious. The board decided that they knew what was cgk, could assert what was cgk without evidence, and that it was up to the party to disprove their notion of the cgk. The following statement of the board flies in the face of justice and logic: “The appellant had the opportunity to comment on the board’s view and provide arguments why the facts the board relies on are held to be wrong. A mere denial that such commonly known prior art existed is not sufficient. In the oral proceedings, the appellant said that it could neither confirm nor deny the existence of it. The appellant, however, did not, for example, provide evidence of when the contentious feature, i.e. an EPG in the form of grids or tables, was introduced into the art.” It is lamentable that a board thought this was sound reasoning. It is not for the proprietor to prove a negative. The person making an assertion must bear the burden of proof. The board should easily have provided the evidence if it was as notorious as they claimed.
@ Anonymous,
I can only but agree with you. I find decision T 1370/15 abominable and as you say it defies any justice or logic. A party alleging a fact has to prove it. The party is here the board. If it is such well known CGK, it should also not be at all difficult to prove it.
When I quoted the catchword of this decision as being “eloquent”, you might not have noticed the tongue in my cheek. I should have written “eloquent”. The way the board discharged itself from its onus of proof is a scandal on its own.
The type of misconduct put to light in T 1370/15 and in T 2526/19 began with T 1090/12. In T 1090/12, have a look at Reasons 6. The board’s argumentation is as “eloquent” as in the catchword of T 1370/15.
T 2225/13, Reasons 4.7, T 1066/18, Reasons 1.9, in which T 1090/12 has been quoted are of the same vein.
If an ED or an OD would allow itself to come up with such an argumentation without any supporting evidence, you can take bets that a board would conclude immediately to a SPV, and send the decision back by return of post.
Dear Max Drei,
I agree with you, that your choice does indeed define the art.
However, G 1/24 makes a difference. If you claim a cylinder, you will not be allowed later to rely on a more limiting interpretation in the description stating for instance that what you are actually after is a stent or a curler.
Under G 1/24, broad claims cannot any longer be “shrunk” according to statements in the description.
Under G 1/24, broad claims cannot any longer be “broadened” according to statements in the description. If you claim a stent, it will have to be checked if you allege in the description, that this structure can be used in any cylinder having apertures. But then you might be confronted with more general prior art.
According to G 1/24, all what you write in the description and has an influence on the interpretation of the claim has to be considered.
If you don’t indicate a technical domain, and merely rely on a new node/strut pattern which yields improved mechanical properties in the apertured cylinder, then any technical domain in which a skilled person could envisage to use your invention, will become relevant.
The problem is thus that any broad definition of claimed subject-matter can then be attacked in various ways. Not only the CGK of a general mechanical engineer will have to be considered, but also that of any engineer working in the field of stents or curlers.
Mr Thomas,
Thank you for your comments of 9 Jan and your review of the decision of the French cour de cassation and of the file of the patent in suit. I appreciate your interest.
I agree that for a field of use to be considered a limiting feature, it is safe to link it to claim features. This is what I recommended in the article mentioned in my comment.
As to whether a field of use indication alone should be held limiting, I do not quite agree that it is should be given no weight in the assessment of novelty of a product claim.
A claim must be read as a whole, I see no reason to disregard a field of use indication if the language of the claim makes it clear that the product is intended for that use.
Also, inconsistency between the extent of protection granted by the claims and how the claims should be interpreted for determining patentability should be avoided. This is clear from G 1/24.
In addition, G 1/24 requires that the description be consulted for claim intepretation. This should apply to the question of whether a field of use indication is to be considered limiting. I hope the relevant guidelines will be revised accordingly and simplified.
@MaxDrei – Your comment of Jan 7.
The notion that the skilled person has access to all the prior art is rather theoretical, what matters is the segment of the prior art which the skilled person would consider when looking for a solution to the problem.
Whether the skilled person is a « person » is disputable, but what is clear is that the skilled person and their CGK can be depicted as an Angora cat. An applicant or patent owner will argue for a narrow definition for attacks on IS/novelty, for a broad definition for attacks on Art 83 or 123(2). This suggests it is useful for an opponent to combine both.
Nice, that the thread continues. As to “field of use” what L was taught as a traineee in London in the 1970’s was that “for” in claim 1 is inherently an anti-Angora Cat. When assessing patentable novelty, one reads it as “suitable for”. When the court decides whether the accused embodiment infringes, it reads the “for” as “intended for”. Accordingly, it is generally NOT a good idea for the claim drafter to put their trust in the word “for”. Example: staplers are useful in an office and for surgeons. When claiming a surgical stapler it might not be a good idea to claim “A stapler for surgery”.
Mr Hagel, Dear Max Drei,
Although I agree with Mr Hagel that a claim has to be read as a whole, whether the “for” is limiting or not playing a direct and unambiguous role in the interpretation of the claim, but it cannot be systematically read in the claim as a limitation.
A device having the same features as those claimed is only novelty destroying if the prior art device is suitable for the same use as that claimed. If it is not suitable for the claimed use, it can be forgotten.
If, when assessing the claim as a whole, the indication of use appears not be limiting, any piece of prior art can be novelty destroying or used as prior art for IS in spite of a specified field of use in the claim.
The situation is the same as when the applicant/proprietor files a broad claim, but tries to introduce a more limited interpretation in the description and relies on this interpretation to fend off more general prior art. G 1/24 has clearly put a hold on this.
I thus agree with Max Drei when he explains that it is not a good idea for the claim drafter to rely too much on the limitation allegedly introduced by the field of use.
Max Drei comment on the “surgical stapler” reminds me off I famous case in a previous directorate of mine.
The “surgical stapler” was such, that the next staple would not follow immediately an action on the trigger, what every paper stapler does. The trigger was reused as a kind of hammer in order to insure that the staple was correctly in place and as near as possible to the body it was inserted, here a bone. The stapler could be left in place whilst the staple was pushed in the bone.
This idea was known in the general field of staplers, in presence of a hard underground. Transferring this idea to a “surgical” stapler was not inventive. The applicant nevertheless got a patent, but on something totally different.
As the stapler was a “surgical” stapler, it needed to be sterilised. Next to the use of the trigger as a hammer, the applicant devised a system in which the stapler could be easily dismantled so that its sterilisation after use was much easier. This was the patent he eventually got.
With respect to the hammer, the “surgical” character was not limiting. When it came to the sterilisation, the “surgical” character was limiting.
This example is a further reason for me not to agreeing with Mr Hagel.
It could well be that Max Drei was the representative in the case of the “surgical” stapler, as I remember meeting him regularly in opposition, at the time my directorate was dealing with medical devices.
Hi, Daniel. In the case you describe, it wasn’t me. But my stapler client back then (when surgical stapling was new) often had problems with novelty attacks based on office stapler art