Monday, 29 March 2010

T 392/07 – On Juxtaposition

In the present case the examining division rejected the application, since the claimed method of identifying an integrated circuit (IC) was a juxtaposition or aggregation of features known from the prior art and did therefore not involve an inventive step. The Board comes to the same conclusion.

[3.2] The method of claim 1 differs from the identifying method disclosed in D5 in that the IC is additionally marked with a machine readable optical identification code which corresponds with the electronic identification information.

[3.3] The applicant argued that it was the objective technical problem to create a technique which enabled traceable identification of the individual IC chips through the various manufacturing process steps […]. The board understands this statement of the problem as being directed to the identification of the ICs and not to the identification of the IC chips, as the latter are not a feature of claim 1 and, in particular, since an identification mark on the IC can hardly be seen once the IC is encapsulated in its housing.

[3.4] Document D8 discloses a dense bar code, particularly useful in the identification of semiconductor wafers and/or chips in very large scale IC manufacturing processes where space on chips and wafers is at a premium.

[3.5] Documents D5 and D8 hence disclose tools for identifying ICs either electronically, as in D5, or optically, as in D8. They are concerned with the tools themselves, i.e. how the coding of the information is done in each case. When to use these tools, whether or not to use more than one tool and how to combine these tools when more than one is used, these questions are left open in these documents to the discretion of the skilled person facing a specific situation.

[3.6] The board considers that a skilled person, having the task of providing a further identification on the IC, would use the optical identifier disclosed in D8, as it is an identification method specifically designed for identifying ICs. Both identifiers, the electronic and the optical, correspond forcefully with each other, as they identify the same object. They are two pointers pointing to the same spot. Thus the functional interrelationship between both markers adduced by the appellant applicant is nothing more than the consequence of both markers identifying the same circuit. The board is unable to imagine how two identifying markers of the same object could not correspond to each other.

The problem allegedly underlying the present invention, namely traceable identification of the individual IC’s through the various manufacturing process steps, is independently solved by the electronic and the optical identifiers. They differ in that for reading the former electric contact with the IC is required, while for the latter it is not. A more appropriate formulation of the problem would be therefore allowing contactless traceable identification of the individual IC’s. To this problem an optical identification code is a straightforward solution.

[3.7] The applicant argued that the use of both identification marks provided a greater security than the prior art approaches and had thus a synergistic effect. Apart from the fact that the application does not disclose any security concerns, the board is unable to recognize how the security would be improved by providing two points of attack. In order to interpret a code the coding has to be known. This applies to the electronic and optical identifiers of the prior art as much as to the double system of the application. The board cannot recognize any added security in the claimed identification scheme, since the information encoded in each identifier is revealed when it is decoded. As in the prior art, each identifier encodes the information independently from the other, since one identifier has not to be decoded for accessing the other identifier. The [applicant’s] argument that by decoding only one of the identifiers only a part of the encoded information is revealed is true, but nothing else could have been expected, since by breaking a code only the part of a message encoded with that code is revealed.

[3.8] The appellant applicant further argued that the provision of the optical marking allowed identifying the IC even when the electronic information could no longer be accessed due to failure of the circuit. This, however, only underlines the independence of both identifiers and refutes the presence of a synergistic effect.

[3.9] In the board’s view, the identifying method of claim 1 is an aggregation of features which are in themselves known from documents D5 and D8. Although most inventions are formed by features which are in themselves known from the prior art, for a finding of non-obviousness these features would have to interact with each other to provide something that is more than the sum of the parts. Absent synergy the whole is not more than the sum of its parts, i.e. an aggregation of features, as correctly objected by the examining division. 

[3.10] The board judges, for the above reasons, that the method of claim 1 of the main request does not involve an inventive step.

This decision is interesting because it involves a concept that is not so often present in EPO decisions: the concept of juxtaposition, i.e. the aggregation of features the effect of which does not go beyond the sum of the effects of the individual features (in other words, there is no synergy between the features). This notion was very common in French case law prior to the alignment of French law with the Munich convention. Maybe this is the reason why it has some “vintage” flavour to me.

The notion of juxtaposition is not without danger, because it is of course possible to have a patentable juxtaposition of known elements, in particular if at least one of the known elements was disclosed in an A 54(3) document only (which is not the case here).

Therefore, I would say that the reasoning of the Examining Division and the Board using the short cut “juxtaposition” (thereby following the Guidelines C-IV 11.7) is incomplete. Methinks that it would have been preferable to use the traditional problem-solution approach, starting from either D5 or D8 and showing that the combination of features was obvious. The fact that the claimed device appears to be a juxtaposition of features disclosed in D5 and D8 is indeed a strong indication that the combination was obvious, but it is not as such formal proof for lack of inventive step.

To read the whole decision, click here.


Anonymous said...

I would say the reasoning in point 3.9 is fine. A collection of known features (or sets of features) is obvious if each feature (or set of features) performs its normal function and the combination of these features does not achieve a surprising technical effect.

In terms of the problem solution approach, each feature or set of features is the obvious solution to the partial problem of providing its or their normal function.

"The combination, to be patentable, must produce a different force or effect or result in the combined forces or processes from that given by their separate parts. There must be a new result produced by their union; if not so, it is only an aggregation of separate elements." (Reckendorfer v. Faber, 92 U.S. 347 (1875), on the obviousness of a pencil with attached eraser)

Anonymous said...

There is the logical conflict to show that it would be "obvious" to put arbitrary measures side by side. This is dealt with in the PSA by "defining" arbitrary measures as obvious. This is however not completely satisfactory in my view.

I still find the present approach more convincing. Moreover, at the end of point 3.6 a redefined problem was defined and an obvious solution, namely an optical identifier, found.

Oliver G. Randl said...

Thanks for your thoughtful comments.

What I wanted to point out is that the juxtaposition approach is not really rooted in the EPC and its interpretation by the Boards. It has its roots in much older legal concepts. It would have been more coherent to use the problem-solution approach, which, as far as I can see, provides the adequate tools (partial problems, ...).

But then, if everybody except me is happy with juxtaposition, I shall not object any more. ;-)

Anonymous said...

I am also not happy with juxtaposition in this case. It is clear to me that providing two methods for identification provides a synergistic effect, namely, as asserted by the appellant, to increase the reliability of the identification method.

I agree that a PSA choosing D5 or D8 as CPA would have been better.

I also have the impression the board took here a rather subjective shortcut.

Anonymous said...

@anon of 16:59:
The synergistic effect of increasing the reliability of the identification method is what you expect when you combine two alternative methods of identification. So still a juxtaposition according to the Guidelines.

@anon of 11:33:
I wouldn't call it a logical conflict, but rather the logical consequence of the principle that an inventive step must be based on an unexpected technical effect (so that in particular the term "obvious" does not have the meaning it has in everyday language).

I certainly agree that PSA could have been used as well (since I think juxtaposition is basically a special case of PSA), but juxtaposition seems to be a useful shortcut and to lead to maybe more natural reasoning. (And I think it's nice to see that PSA did not abandon such an old legal concept but rather generalised it.) But please do object! :)

funes said...

For @anon,

what does it have to do juxtaposition with what you expect or not? Where is that in the Guidelines?

Guidelines, C-IV, 11.5:
"A set of technical features is regarded as a combination of features if the functional interaction between the features achieves a combined technical effect which is different from, e.g. greater than, the sum of the technical effects of the individual features. In other words, the interactions of the individual features must produce a synergistic effect. If no such synergistic effect exists, there is no more than a mere aggregation of features (see T 389/86, OJ 3/1988, 87)"

It is still clear to me that the combination of both methods of identification achieves a combined technical effect. Whether the effect is to be expected or not is another question which has to be answered in the light of prior art (that is why PSA is out there...).


ps.- an inventive step must not be imperatively based on an unexpected technical effect (see C-IV, 11.6, e.g. problem invention...)

Anonymous said...

You are indeed correct that GL C-IV, 11.5 does not mention "unexpected". However, the case law does. See e.g. T 473/90, 2.4:
"Therefore, the apparatus of Claim 1 can only be considered as a juxtaposition of means acting in their normal way without resulting unexpected effect."

Or see T 848/94, 4.5.2: "This synergistic improvement of the bond strength was unexpected to one skilled in the art. Its achievement was non-obvious within the meaning of Article 56 EPC."

Regarding C-IV, 11.6, have a look at the Guidelines applicable as of 1 April. But yes, problem inventions might be an exception.

funes said...


Thank you for the case law, I will have a look at it.