The rift between UK-IPO practice under English law, and EPO practice under the case law of the Technical (remember that word) Boards of Appeal has widened further. Decision T 1227/05, until recently only available in the original German, has been published in the EPO Official Journal (OJ EPO 11/2007). This decision marks a significant development in the EPO’s interpretation of Article 52 EPC, further narrowing the scope of ‘excluded’ subject matter, in particular relating to computer programs and mathematical methods.
The decision concerned an application by Infineon Technologies for a computer-implemented method for numerical simulation of a circuit, in which a particular type of electrical noise (termed ‘1/f noise’) was simulated through the use of a mathematically-defined random number generator. The simulation allowed the circuit to be modelled more accurately, and therefore better optimised before a real circuit needed to be made.
Initially, the examining division had refused the application for constituting a ‘mental act’ or a mathematical method as such, because it did not specify any means for implementing the invention. Adding to the claim that the method was ‘computer-implemented’ overcame that particular objection (after T 258/03), but the board was left with the problem of whether the mathematical formulae in the claimed invention could contribute to the required “technical character” (after T 641/00). The invention, the applicant argued, was an improvement over the prior art because the claimed mathematical method of generating random noise made it possible to simulate noisy circuits on small computer systems not previously powerful enough for the purpose, or to simulate larger circuits with the same computer power. This, the applicant argued, constituted “a technical effect that went beyond the normal physical interaction between a computer program and a data processing system”.
The board, headed by the familiar figure of Stefan Steinbrener, was persuaded by this argument and considered that the simulation method constituted an “adequately defined technical purpose for a computer-implemented method, provided that the method is functionally limited to that technical purpose”, and that all steps relevant to circuit simulation, including the mathematically expressed features, contributed to the technical character of the method. Simulation could therefore be a technical feature, and this constituted neither a mathematical method as such nor a computer program as such, even if mathematical formulae and computer instructions were used to perform the simulation.
The board acknowledged that this finding departed from the previous decision of T 453/91, in which a design method for a semiconductor chip was not deemed to be patentable unless a manufacturing step was included as part of the claimed method. In the board’s view, the importance and assessment of industrial simulation methods were changing, and numerical simulations were becoming a cost-effective alternative to real experiments, evolving into a key technology. Specific technical applications of computer-implemented simulation methods were themselves therefore to be regarded as modern technical methods which formed an essential part of the fabrication process, preceding actual production. Such simulation methods could not be denied a technical effect merely on the ground that they did not yet incorporate the physical end product. In summary, the board considered patent protection to be appropriate for numerical development tools designed for a technical purpose, which could not be denied a technical effect merely on the ground that they did not yet incorporate the physical end product.
The board remitted the application back to the examining division to have a search carried out so that a proper assessment of novelty and inventive step could be made, taking into account all those features making a contribution to technical character.
The IPKat finds it interesting to consider how the same claimed invention might be viewed at the UK-IPO, when applying the Aerotel/Macrossan test rather than the EPO ‘technical character’ test. Construing the claim would not be particularly difficult: the method is one of computer simulation of an electronic circuit using a specified mathematical noise model. The contribution in this case would need to be the alleged contribution, in the absence of a search of the prior art having been carried out. Although this might not be the same as the actual contribution, which could only be determined once the prior art was taken into consideration, it could not, unless the applicant failed to identify what in his view the invention was, be any narrower than the actual contribution. The contribution would therefore be viewed as being the application of the specified mathematical noise model in a computerised simulation method, this having certain efficiency advantages. Since this contribution is limited to that of a mathematical model as implemented by a computer program, the contribution could be no more than a mathematical model or computer program as such, and would therefore be excluded from being considered to be ‘an invention’ under section 1(2)(a) and (c) (or Article 52(2), if you prefer). Having failed the third step, any reference to a ‘technical effect’ could not save the method from being excluded, so (as usual) the fourth step would be essentially ignored.
This is the point where patent attorneys, even those who have repeatedly argued and lost hearings at the UK-IPO, will start to complain that the advantages of the invention have not been taken into account. However, the IPKat has not yet seen a UK-IPO decision following Aerotel/Macrossan that has taken into account any actual or alleged advantages to support a contribution being not excluded when it would otherwise be excluded. It is probably always worth a shot, but one should not expect a different result when using an argument that has tried and failed numerous times before.
This is not, however, a mere theoretical musing on how the different approaches might reach different results. Two recent decisions at the UK-IPO (O/342/07 and O/343/07) have resulted in refusals of applications for computer-implemented mathematical simulations. Both of these related to simulations of oil reservoirs, and the IPKat cannot see why these would not now be considered to be sufficiently ‘technical’ to pass the test the EPO board has approved.
The remaining question is: wherefore the different approaches and different results? There appears to be little hope at the moment for a resolution, particularly since (in T 154/04) the board of appeal (Steinbrener again, of course) comprehensively rejected any reference to the enlarged board of appeal to resolve allegedly conflicting EPO decisions, which the Court of Appeal has now effectively directed the UK-IPO and lower courts to ignore. There is also the prospect of a further hardening of the differences when the Astron Clinica judgment in the High Court is issued (see earlier IPKat posts here and here). The IPKat thinks that the real reason lies in the entirely different legal approach the English courts and UK Patent Office have been taking compared with the approach by the EPO boards of appeal. Whereas the EPO, at least in the form of Steinbrener, clearly takes the view that interpretation of the law should evolve as technical progress evolves, the English approach is now trying to go back to basics and interpret the law as it is written, without regard to how technology has developed since the law was first put in place. The former approach depends on the ever-changing definition of what is "technical" (which the IPKat challenges anyone to provide a stable definition of), and departs substantially from the wording of the EPC, whereas the latter tries to assess what the claimed invention actually is, in terms of what it contributes to public knowledge, and whether it is excluded. While the different approaches continue, there seems to be little hope of reconciling English patent law with that of the EPO.