Saturday, 22 September 2012

T 56/09 – So Close

I sometimes disagree with the way in which the problem-solution approach is implemented, in particular in inter partes proceedings. The present case might illustrate what I mean.

The opponent filed an appeal after the patent had been maintained in amended form by the Opposition Division.

Claim 1 as maintained read as follows:
A syringe by which medicine may be apportioned in preset doses from an ampoule, which syringe has
- a dose setting (2, 23) and injection mechanism comprising a dose setting member which comprises an injection button (11) on a carrier rod (13) and may be moved in one direction to preset a dose and in the opposite direction to inject the preset dose, and
- a piston advancing mechanism comprising a piston rod and a piston rod drive,
- the dose setting (2, 23) and injection mechanism being coupled to the piston advancing mechanism through a unidirectional coupling transmitting only movement of the dose setting member in the dose injecting direction to the piston rod drive (17),
in which syringe
- an air shot button (26, 30) is provided,
- the operation of which acts only on the piston advancing mechanism side of the unidirectional coupling to enable influence on the piston advancing mechanism to advance the piston rod a distance corresponding to expulsion of a fixed volume of medicine.
There was a discussion on whether document D6 qualified as closest prior art:

[6] The [opponent] submitted that the subject-matter of claim 1 was also not inventive over a combination of D6 and the general knowledge of the person skilled in the art.

In the so-called problem-solution approach for assessing inventive step in an objective manner, the first step is to determine the closest prior art, i.e. the most promising starting point or “springboard” towards the invention.

In the present case, as already mentioned in T 1166/05 [3.2], “D2 and D6 disclose less features of claim 1 of the main request than D1”.

More particularly for D6 the Board considered that
“D6 does not refer to a syringe by which medicine may be apportioned in set doses from an ampoule. The syringe according to D6 has a closed piston chamber (19) into which liquid medicine may be drawn and subsequently be injected by a piston (17). Furthermore this syringe does not comprise an air shot button but a system that automatically eliminates any air from the piston chamber when this chamber is filled.”
Hence it is clear that the syringe disclosed in D6 is not of the same type as that claimed, since it is not a syringe by which medicine may be apportioned in set doses from an ampoule, and as already mentioned above it has fewer features in common with the subject-matter of claim 1 than D1.

Thus, according to established case law, the syringe disclosed in D6 is further away from the subject-matter of claim 1 or the invention than the syringe disclosed in D1.

In other words, D6 is a weaker starting point than D1 for a lack of inventive step reasoning, so since the subject-matter of claim 1 is inventive starting from D1 it is also inventive starting from D6.

It may well be that the Board is right in this particular case, which I have not studied in detail, but I believe that this line of reasoning is not without danger.

I like to compare the assessment of inventive step with mountaineering: the skilled person is faced with the problem of reaching the summit of a mountain. He/she is a good walker and capable of reading a map and following a path, but unable of climbing. We want to find out whether the invention is obvious or inventive, i.e. whether the summit under consideration can be reached without climbing or not. In my illustration, the prior art documents correspond to base camps that exist on that mountain.

Consider Mount Narish and its two base camps A and B:

Mount Narish, as seen from the north and the south, respectively
A is very close to the summit S, but in order to get from A to summit S, you have to climb the deadly north face. The skilled person cannot take this route. So, at first sight, the invention is inventive.

However, when starting from more remote base camp B, there is an easy way to summit S, altogether avoiding the deadly north face. Admittedly, it takes longer to get to the summit, but the path is never too steep and no climbing is required. So indeed the invention does not involve an inventive step.

Now, as can be seen from the map, camp B is farther away from the summit than camp A:

A topographic map of Mount Narish
One might say that B is not the closest prior art. But one would come to the false conclusion if B was discarded without further ado. As a matter of fact, B is the better springboard, but in order to find out, you have to take it into account: you need to follow both paths to see whether one of them does not require climbing.

What is the practical conclusion from all this? I would say that a Board should never discard prior art which the opponent claims to be the best point of departure, just because it does seem more remote than another. If the opponent claims this particular prior art to be the most promising springboard, it will, in general, have good reason to do so. The Board should check it out before dismissing the attack, based on mere proximity arguments. To put it differently, time would be better spent if the discussion of whether something is indeed “the” closest prior art was replaced by a discussion whether the prior art invoked by the opponent did indeed lead to the invention without any inventive step.

I therefore fully agree with decision T 967/97 where the Board made the following statement:
“If several paths to the the solution (Lösungswege) along which the invention could be obvious offer themselves to the skilled person, then the reasoning (ratio) underlying the problem-solution approach requires [the Office] to examine the invention with respect to all of these paths to the solution before a decision confirming the presence of an inventive step can be taken.”
Decision T 1514/05 and others have since then followed this approach, which, incidentally, would also help to (slightly) raise the bar.

Should you wish to download the whole decision, just click here.

The file wrapper can be found here.


Anonymous said...

Wholeheartedly agree.

The alternative, that only one of the cited prior art documents is a priori the allowable closest prior art with regard to which inventive step may be assessed, leads to a bizarre situation as follows.

If D1 (for example) had not been cited or uncovered in a search, the invention would have been deemed obvious, but if it had been subsequently cited or uncovered, (even by the Proprietor!), that which was previously found obvious would then become non-obvious, following the reasoning of the Decision. But it is odd to think that the existence of one single alternative starting point merely having more in common with the invention claimed should render the obvious inventive.

The better approach is that of T967/97. However, this implicitly requires reformulating the first step of the classic PSA test as "identify a reasonable starting point for the skilled man", and then adding a final step of "if the conclusion is one of the existence of inventive step, repeat the process with another reasonable starting point". Only unreasonable starting points should be discarded in this analysis.

I enjoyed the mountain analogy, by the way...

Sander van Rijnswou said...

Sorry Oliver, I'm not with you on this one. I agree with the board here.

Many inventions look like this: (Clearly well known object) + (New feature).

A clever strategy to make that look obvious is the following: First find the new feature, anywhere you can, ignoring the known object as much as you need to. Say you find it in D6. Then you argue as follows. The difference between D6 and the invention is: (Clearly well known object). But that is a clearly well known object! So.. it is trivial to add, everybody knows it. Most likely it is even common general knowledge.

In this case, you find the new feature in D6, which is a syringe of a completely different type. The problem is to add the well known syringe.

This approach will invariably lead to hindsight. The skilled person was tasked to improve ampule based syringes. Starting from that it was not obvious to find the improvement. Sure, if we gift the inventive feature, first and then task the skilled person with a redesign he may be able to do so, but that is the wrong way around.

The caselaw book cites T 439/92, which says that, sure you are free to pick a different closest prior art. But...that will define the framework of further development. In this context, sure you could take D6 as closest prior art, but you will never get the syringe of the type claimed.

The mountain analogy is inherently flawed. It assumes that there is peak that everybody can see. But that is hindsight, in reality nobody knows there is a peak or where it lies.

oliver said...

No need to be sorry, we do not always have to agree.

As I said, my remarks were not directed at this particular case. Whether the Board was right or not in its particular choice of the closest prior art is not my point.

I do not think that my analogy is flawed. We examine inventive step once the invention has been made and claimed. So the peak is there and visible to all. The question is : was there an obvious way to get there ?


Anonymous said...

"The skilled person was tasked to improve ampoule based syringes." - but was he?

The skilled person is equally free to start with non-ampoule based syringes, and to seek to modify those in an obvious manner. If he can arrive at the claimed invention in an obvious manner from any realistic starting point, then the invention is itself obvious. The quality of obviousness or inventiveness attaches to the claimed subject-matter in and of itself, not the circumstances of its creation. Otherwise, one skilled in the art would not be free to modify the known D6 syringe in an obvious manner, which defeats the public policy objective of the inventive step test - to make that which is known or obvious free for all to use.

On the other hand, there is a real argument, starting from D6, as to whether one skilled in the art realistically would have contemplated modifying D6 to the claimed invention, because the skilled person saw the two types of syringes as distinct species, and would not have considered that interbreeding them would solve any problem.

The Board, however, never gave us its thoughts on that basis. I think that is a shame, since D6 might well not have been a realistic starting point, but perhaps not for the reasons advanced.

MaxDrei said...

Here's my view:

The EPO prides itself on taking a "real life" view of obviousness, free of hindsight. One has to be fair to the inventor. So follow the piston ring decision (T570/91 from memory). Give full faith and credit to the inventor's statement of technical field and subjective problem, and find a prior art starting point that fits with that, and define the objective problem relative to that. If the claim is directed to a compressor ring, you can't start with an ic engine piston ring, for those wanting to improve compressor piston rings don't start from ic rings but only from the best available compressor ring starting point. Of course, if all you claim is a piston ring, unlimited to ic engines or compressors, either starting point is fair.

This is IMPORTANT for no judge will entertain PSA reasoning if it is unfair to the inventor or relies on hindsight knowledge of what technical feature combination solves the objective technical problem.

I for one prefer the fairness of EPO PSA over the subjective "Obvious because I say it's obvious" approach of the judges in some jurisdictions. PSA can deliver BOTH fair protection AND legal certainty, but ONLY if it is operated in accordance with the piston ring case.

Myshkin said...

I also agree with Oliver. It could be that in this particular case there was simply no way that a skilled person starting from D6 would arrive at the claimed invention in an obvious manner, BUT the reason for this cannot be that there just happened to be some "better" document on file (and/or in the prior art).

The conclusion that a claim is inventive is always conditional on the prior art that is available. It is always possible that some piece of prior art was missed that would have rendered the invention obvious or would even take away novelty. But imho it would make no sense if this could happen in the other direction.

I find the terminology "closest prior art" unfortunate. I would prefer "suitable starting point".

I think T 570/91 confirms that in principle no piece of prior art can be excluded:
"4.2. The assessment of inventive step of the subject-matter of an independent claim has to take into account the state of the art (Article 56 EPC, first sentence), that means everything (including all documents) made available to the public before the date of filing of the European patent application (Article 54(2) EPC).

Instead of making the effort of trying to proceed one by one from each single available prior art disclosure to the claimed invention, it is simpler to start from that available prior art disclosure which is "closest" to the invention, i.e. that which is the most promising starting point for an obvious development leading to the claimed invention. This closest prior art may be one particular embodiment of several disclosed by a document. After having made the above choice, the other prior art disclosures can therefore be neglected as starting points to assess inventive step because they are less promising than the closest prior art disclosure."

Yes, it is SIMPLER to consider only the "best" prior art, but this is only a practical trick useful for avoiding excessive work (imagine having to provide reasons for each document present in the EPO's database that was published before the filing date, and then for all internet disclosures).

Point 4.4 again confirms that the skilled person is completely free in choosing a starting point. The Board adds that the skilled person is then bound by this choice, but of course that only applies to the specific inventive step reasoning starting from that prior art. It does not mean that the Division or opponent is bound by a particular choice of "closest" prior art and is somehow prohibited from considering other documents.

I can also agree with point 4.5: "In most cases it is appropriate to look particularly at those prior art disclosures which have something to do, be it explicitly or implicitly, with the problem set out in the application or opposed patent." This is of course true for most cases (but not for all).

oliver said...


Your remark concerning the expression « closest prior art » vs. « suitable starting point » made me smile. In my original draft I had a very similar remark. I took it out because I think the EQE has rooted these terms so deeply in most of us that we will not be able to get rid of them, even if they are unfortunate.

Myshkin said...


True, and many EQE papers actually reinforce the idea that there can only be one "closest prior art". When doing the papers this is quite helpful, but it is not reflecting reality.

MaxDrei said...

How about a mountain range in which there are twin peaks, A and B, each one with a topography very much like its neighbour.

Suppose the claimed subject matter is the summit of Peak A.

Suppose the choice of prior art starting points is between i) D1, in the foothills of Peak A and ii) D6, high up on the shoulder of Peak B.

Is there any other legitimate and fair starting point than the D1 basecamp?

Anonymous said...

I find the moutaineering analogy very interesting.

However, I don't think that the selection of the closest prior art should be made while considering the path that it is going to be used for arriving or not at the invention.

The "closest prior art", as I understand it, is the most promising one that the skilled, but not inventive, person would have used as a starting point for solving the problem.

This person would indeed select the starting point that is the closest to the summit.

I think that if the skilled person can anticipate the "deadly north face", start from a remote location and use a non direct path (even though easier)for arriving at the invention, he or she is inventing.

I would agree if points A and B were at a same distance from the summit. In this case, the skilled person, even not inventive, would select the easiest path.

Myshkin said...

"The "closest prior art", as I understand it, is the most promising one that the skilled, but not inventive, person would have used as a starting point for solving the problem."

Since the technical problem to be solved can only be formulated after identifying the differences between the claimed invention and the closest prior art, this problem cannot play a role in selecting the closest prior art.

The novelty requirement is intended to prevent a patent being granted on inventions that have already been made available to the public. The way I see it, the requirement of inventive step is intended to ensure that the public is free to take any known invention and modify it in obvious ways.

Reformulating this, inventive step should be denied if there is any disclosure in the prior art and any obvious way of modifying that disclosure that would result in the claimed invention. So in principle anything can be used as "closest prior art", but only those that allow an obvious modification into the claimed invention are relevant.

Note that the obvious ways of modifying a known disclosure are not determined by the claimed invention. Instead, they are determined by the various problems that, starting from the disclosure, the skilled person would pose and be able to solve.

However, when examining a claim for inventive step, it is again not helpful to consider modifications that do not lead to the claimed invention anyway. So the examiner (after selecting a closest prior art) should work backwards: first determine what modification would be necessary to arrive at the claimed invention, then determine whether the modification is an obvious solution to the problem it solves.

So the problem solution approach is not so much a way of determining whether the skilled person would consciously arrive at the invention starting from nothing, as it is a way to test whether the claim covers embodiments that only differ in an obvious way from something already known, without having to go through all possible disclosures and all possible obvious modifications.

MaxDrei said...

Anon points out, 28 Sept, that the objective problem cannot be set till the prior art starting point has been fixed. Right.

Myshkin says the same.

it appears to me that neither has appreciated my comment of Sept 24 that one identifies a "realistic" prior art starting point by reference to the subjective 8not the objective) problem announced by the Applicant in the description of the application as filed.

Start from that, find a prior art starting point that is "realistic" ie fair to the inventor having regard to his/her story of technical field and problem addressed. That uniquely identifies the mountain to be climbed. That gives you a base camp on the right mountain.

From that base camp, on the right mountain, follow EPO-PSA. Is everybody OK with that?

Myshkin said...

"it appears to me that neither has appreciated my comment of Sept 24 that one identifies a "realistic" prior art starting point by reference to the subjective 8not the objective) problem announced by the Applicant in the description of the application as filed."

I have appreciated the comment, but I do not agree with it :-).

In my view (which I have tried to explain), the whole idea of inventive step is that the claim should not cover any embodiment that can be arrived at by an obvious modification of a disclosure in the prior art. In principle *any* disclosure in the prior art can be the starting point. In fact, nothing in the application (such as its subjective technical problem) apart from the effective filing date can in any way influence what is obvious and what is not obvious.

However, since it is evidently impossible to first list everything that is obvious, and then to check whether the claim covers something in that list, one has to be selective in what starting points and what modifications to consider. This selection can (and for practical reasons must) be made with full knowledge of the invention (*). It will usually, for these practical reasons, make sense to only consider "realistic" starting points, but the fact remains that any starting point that can be obviously modified into something falling under the claim is sufficient to show a lack of inventive step.

"Start from that, find a prior art starting point that is "realistic" ie fair to the inventor having regard to his/her story of technical field and problem addressed. That uniquely identifies the mountain to be climbed. That gives you a base camp on the right mountain."

To me it makes absolutely no sense that the inventor by telling a good (or bad) story could alter the prior art that may be taken into account in the examination of one and the same claim.

(*) A selection made with full knowledge of the invention. This does not in any way imply impermissible hindsight, because in principle inventive step requires everything to be taken into account. The selection is only made to ensure that examination ends in a finite amount of time. The more accurate the selection, the higher the quality of the examination.

MaxDrei said...

A pleasure to receive that reply from Myshkin. I agree, to the extent that, if the claimed subject matter is truly obvious, it should not matter what bonus technical effects it delivers. So, what say you, we now debate the "blue squash ball" case?

Traditionally, the ball for playing the racquet game of "squash" in a closed court with 4 white wals, was black. But then the patented blue ball appeared. How can it be, that a claim to a squash ball coloured blue can be valid? Well, the evidence was, that the human eye/brain picks up the flight of a blue ball against a white background faster than it can a black ball. Thus, you get a higher quality game of squash with a blue ball.

Now to Myshkin, who is against deciding obviousness on the cleverness of the story told by Applicant in the application as filed.

Suppose the application as filed includes a "story" about picking up the flight of the ball quickly. Would that help Applicant to get through to issue? Conversely, suppose the application as filed had no such "story". Would the claim in that case be patentable?

As you see, I think the subjective problem has to be taken into account for, generally speaking, it's plainly obvious to colour blue a ball for playing a racquet sport.

Myshkin said...

This is the type of example I conveniently ignored when writing my previous comment ;-)

Your example does not affect my statement that a good or bad story cannot alter the prior art that may be taken into account. If the prior art shows a blue ball suitable for playing racquet ball, there is lack of novelty. Blue balls that are not suitable for playing racquet ball might not allow for an obvious modification to a (still blue) racquet ball, but if one does, there is lack of inventive step. The subjective technical problem does not come into play here.

The other case is where one starts from a known black racquet ball. Painting a black ball or any other object to give it another color is an obvious modification, and to select blue as the new color is normally a technically arbitrary, and therefore obvious, choice. However, in the particular case of a racquet ball the selection of blue has a technical effect which is not expected on the basis of the prior art.

In this case I do agree with you that it is important that the description discloses this effect. This is somewhat conflicting with one of my earlier remarks:
"In fact, nothing in the application (such as its subjective technical problem) apart from the effective filing date can in any way influence what is obvious and what is not obvious."

I would still say a blue racquet ball given a particular effective filing data is either obvious or not, irrespective of the contents of the patent application, but there is somehow an additional requirement that the application shows that the applicant actually realised he made the invention. That means that lack of a good story can hurt. Otherwise an applicant could include claims to all possible colors and only after filing the application figure out which color gives a surprisingly good result.

Btw, the Guidelines are very negative on "bonus effects". Probably too negative, though.

MaxDrei said...

OK. Now me. I'm glad you think that I have an effect that is "technical". Some are not so sure. Compare the case about the electric fence with yellow and black stripes, the better to confine cows to their allotted plot of grass.

The applicable prior art depends on what you claim. "A squash ball" is not the same as "A ball for playing squash". Claim a "Ball for playing football" and you cover a tennis ball. Blue footballs are obvious. But claim a squash ball and you achieve novelty because you thereby exclude all balls except those that meet the requirements laid down by the Squash Racquets Association. (We could insert explicit in claim 1 the already implicit SRA specification if necessary).

The only objective problems you can invoke against an obviousness objection are those you can derive from the application as filed. The app as filed has to make the solution plausible. Thus, 1) getting a European patent on your blue squash ball depends decisively on the subjective problem you set out in your app as filed and 2) EPO-PSA is fair and free from hindsight but does require that you teach your invention in your application as filed.