15 July 2020
Today (15 July 2020) saw the hearing of the EPO Enlarged Board of Appeal in G 1/19 – only the second such hearing ever relating to Computer Implemented Inventions (CIIs) at the EPO. While the case that led to this referral relates to computer simulations, it has the potential to have far-reaching consequences for the patentability of not only simulations, but mathematical methods and their application in computer implemented inventions in general (which is reflected in the large number of amicus curiae briefs filed).
While not being physically present, a number of colleagues at Mathys & Squire LLP attended the hearing virtually via videoconference. The EBoA didn’t issue a decision at the hearing, and the hearing itself was relatively brief. We await their findings with keen interest.
Given the current trajectory of technological development, particularly in the field of machine learning and AI, we are hopeful that the first two questions referred to the Enlarged Board will be answered in the affirmative. To do otherwise might risk development of an unduly negative and restrictive standard for the patentability of CIIs.
Only in rare circumstances, normally where case law developed by Technical Boards of Appeal (TBA) over the years appears to diverge, is a matter referred to the Enlarged Board of Appeal (EBoA). For those working in or familiar with the life sciences field, we recently saw such a decision issued in G 3/19 (reported here). In the CII field, the last we had was the referral of the President of the EPO in G 3/08 issued in 2010.
The approach to assessing the patentability of CIIs at the EPO was established in T 641/00 (COMVIK), refined and restated in T 154/04 (Duns Licensing) and expressly approved by the Enlarged Board in G 3/08, the only other referral ever made to the EBoA on this subject.
In summary, the COMVIK approach is first to determine whether the claim includes any technical means (any hardware is enough). Then, if it does (and assuming it is novel) the next step is to determine whether the claim provides a non-obvious technical solution to a technical problem. The EPO does this by assessing which, if any, of the claim features contribute to producing a technical effect. Only those features which do contribute in this way are given any weight when determining obviousness. Those which do not contribute to the solution of a technical problem are deemed to form part of the prior art, given as a “requirements specification” in the objective technical problem to be solved in the “problem and solution” test for inventive step.
A body of established case law in this field has built up over time, and the EPO Guidelines for Examination have also been amended (somewhat recently) to reflect this position.
The revised EPO Guidelines (G-II, 3.3), when discussing the patentability of mathematical methods, note that “A mathematical method may contribute to the technical character of an invention, i.e. contribute to producing a technical effect that serves a technical purpose, by its application to a field of technology and/or by being adapted to a specific technical implementation”. The Guidelines helpfully list a number of examples that may be considered technical, including “simulating the behaviour of an adequately defined class of technical items, or specific technical processes, under technically relevant conditions”.
Importantly, however, the Guidelines and the existing case law suggest that “the mere fact that a mathematical method may serve a technical purpose is not sufficient, either. The claim is to be functionally limited to the technical purpose, either explicitly or implicitly. This can be achieved by establishing a sufficient link between the technical purpose and the mathematical method steps, for example, by specifying how the input and the output of the sequence of mathematical steps relate to the technical purpose so that the mathematical method is causally linked to a technical effect”.
A decision that is often referenced in this field is that of T 1227/05 (Infineon). It relates to a computer-implemented method for the numerical simulation of the performance of an electronic circuit subject to noise. The Technical Board of Appeal held the method here to be patentable because the simulation of a circuit subject to a particular type of noise constitutes an adequately defined technical purpose.
Significantly, in the manner it was claimed, the method was functionally limited to that technical purpose. The Board held that a circuit with input channels, noise input channels and output channels whose performance is described by differential equations does indeed constitute an adequately defined class of technical items, the simulation of which may be a functional technical feature.
By contrast, a claim which attempted “the metaspecification of an (undefined) technical purpose” by making mere reference to simulation of a “technical system” was deemed not an adequately defined class of technical items, and therefore not patentable. This contrast may, we believe, be of particular significance in the present referral.
It may also be worth observing that the EPO Guidelines note that “if the claim were not limited to the numerical simulation of an electronic circuit subject to 1/f noise, the mathematical algorithm defined by steps (d1)-(d3) would not serve any technical purpose and would thus not be considered to contribute to the technical character of the claim”.
The appeal, which gave rise to the present referral to the Enlarged Board, relates to “A computer-implemented method of modelling pedestrian crowd movement in an environment”.
The patent application in question indicates that the main purpose of the simulation is its use in a process for designing a venue such as a railway station or a stadium. This is not recited in the claims of the Main and First to Third Auxiliary Requests but it is referenced in the Fourth Auxiliary Request.
Importantly, at Reasons 10 of the decision referring the questions to the EBA, the TBA held (emphasis added):
“10. As to the technicality of simulating crowd movement, the appellant argued that simulating the movement of pedestrians yielded results which were no different from those obtained by modelling an electron using numerical methods. Like the simulation of an electron, the claimed simulation of the movement of pedestrians was based, at least in part, on the laws of physics.
The Board does not disagree with these observations but is not convinced that numerically calculating the trajectory of an object as determined by the laws of physics is in itself a technical task producing a technical effect.”
The Board went on to state (emphasis added):
“11. In the Board’s view, a technical effect requires, at a minimum, a direct link with physical reality, such as a change in or a measurement of a physical entity. Such a link is not present where, for example, the parabolic trajectory followed by a hypothetical object under the influence of gravity is calculated. Nor can the Board detect such a direct link in the process of calculating the trajectories of hypothetical pedestrians as they move through a modelled environment, which is what is claimed here. In fact, the environment being modelled may not exist and may never exist. And the simulation could be run to support purely theoretical scientific investigations, or it could be used to simulate the movement of pedestrians through the virtual world of a video game.
In this context, the Board notes that the Enlarged Board of Appeal in decision G 2/07, reasons 22.214.171.124, stated that “[h]uman intervention, to bring about a result by utilising the forces of nature, pertains to the core of what an invention is understood to be“. It appears to the Board that using a computer to calculate the trajectories of hypothetical pedestrians as they move through a modelled environment does not utilise the forces of nature to bring about a result in any way different from using a computer to perform any other type of calculation.”
The Appellant had referred during the hearing to T 1227/05 (Infineon) and drawn the analogy that simulating the performance of a modelled electrical circuit (which was held patentable there) is much the same as simulating the behaviour of a modelled environment. In the subsequent written decision, the Board agreed with the Appellant’s arguments that T 1227/05 supported their case but noted that they were “not fully convinced by the decision’s reasoning”. They gave two reasons:
“First, although a computer-implemented simulation of a circuit or environment is a tool that can perform a function “typical of modern engineering work”, it assists the engineer only in the cognitive process of verifying the design of the circuit or environment, i.e. of studying the behaviour of the virtual circuit or environment designed. The circuit or environment, when realised, may be a technical object, but the cognitive process of theoretically verifying its design appears to be fundamentally non-technical.”
“Second, the decision appears to rely on the greater speed of the computer-implemented method as an argument for finding technicality. But any algorithmically specified procedure that can be carried out mentally can be carried out more quickly if implemented on a computer, and it is not the case that the implementation of a non-technical method on a computer necessarily results in a process providing a technical contribution going beyond its computer implementation (see e.g. decision T 1670/07 of 11 July 2013, reasons 9).”
The Board also noted that T1227/05 acknowledged a tension with an earlier decision in T 435/91 which required a step of “materially producing the chip so designed” thus tying the subject matter explicitly to a real-world object. The Board noted the growing importance of “numerical development tools” and in particular CIIs, and that “legal certainty in respect of the patentability of such tools is highly desirable”. The Board therefore referred three questions for consideration by the Enlarged Board.
1. In the assessment of inventive step, can the computer-implemented simulation of a technical system or process solve a technical problem by producing a technical effect which goes beyond the simulation’s implementation on a computer, if the computer-implemented simulation is claimed as such?
2. If the answer to the first question is yes, what are the relevant criteria for assessing whether a computer-implemented simulation claimed as such solves a technical problem? In particular, is it a sufficient condition that the simulation is based, at least in part, on technical principles underlying the simulated system or process?
3. What are the answers to the first and second questions if the computer-implemented simulation is claimed as part of a design process, in particular for verifying a design?
The first question suggests that the applicability of the widely-accepted approach to assessing the patentability of CIIs under inventive step (Article 56 EPC) following the COMVIK decision (T 641/00) is not in question.
However, by using the language “computer implemented simulation of a technical system” (emphasis added) in the question, this presupposes that the system is already technical – which may not be the case here where the system is the behaviour of pedestrians. Instead, the system may be held to be non-technical because, following the reasoning of a very recent decision taken in T 1798/13, it “is not a technical system that the skilled person can improve, or even simulate with the purpose of trying to improve it”.
The use of “as such” language in question 2 was commented on in the amicus briefs because of its apparent relevance to the wording of Article 52(2) EPC, but these issues weren’t raised at the hearing. It remains to be seen whether they will be addressed in the written decision.
The second question seems to address the idea of whether the simulation is (as set out in the Guidelines) “functionally limited to a technical purpose”.
The EPO’s current Guidelines for Examination say that a generic purpose, such as “simulating or controlling a technical system” is not sufficient, but rather that the technical purpose must be a specific one. Moreover, the mere fact that the mathematical method may possibly serve a technical purpose is not sufficient, either – as held in T 1227/05, the claim must be functionally limited to a specific technical purpose, either explicitly or implicitly.
Interestingly, the EPO Guidelines also note (at G-II, 3.3.2) “In the context of computer-aided design of a specific technical object (product, system or process), the determination of a technical parameter which is intrinsically linked to the functioning of the technical object, where the determination is based on technical considerations, is a technical purpose”.
These Guidelines are of course not binding on the Enlarged Board. Indeed, the Enlarged Board’s answer to the second referral question may very well have direct implications for future revisions of these parts of the Guidelines, and if so a direct impact on practice before Examining Divisions at the EPO.
The third questions seems most relevant to the claims of the Fourth Auxiliary Request under appeal – where the claims are directed to a method of designing (a model of) a building structure. The TBA noted that this additional step “arguably strengthens the appellant’s case” that T 1227/05 is relevant yet noted that their view was that it didn’t help as “a direct link with physical reality is still absent”.
During the hearing it was interesting that both the Appellant and the President of the EPO argued in favour of computer simulations as such being technical, with the Appellant arguing that simulations serve a technical purpose and therefore must provide a technical effect.
With respect to the second question, both the Appellant and President made analogies between physical and virtual wind-tunnel tests. They both argued in essence that such simulations rely on technical considerations and therefore have to be regarded as technical – in some sense agreeing with the position set out in the EPO Guidelines at G-II, 3.3.2 but they may perhaps have gone further than this in the arguments they made. The President made further reference to an example of raytracing (commonly used in computer games), noting that the raytracing process is based on technical considerations even if the generated imaged is fictitious.
The EBoA didn’t decide anything today, and the overall hearing was relatively brief. They will likely now take some time to consider the submissions, along with the many amicus curiae briefs before issuing their written decision and minutes of the hearing. However, the consensus view amongst our colleagues who ‘attended’ the virtual hearing is that the first two questions should be held in the affirmative. In our view, deciding otherwise may create a relatively negative and restrictive view on the patentability of CIIs in the future, as well as having an adverse knock-on effect on past cases and in protecting AI inventions more widely.
Sign up to our mailing list to receive Mathys Matters, our monthly newsletter covering the latest IP news, industry insights, events and case law.
If you are interested in receiving quarterly newsletters relevant to our core sector groups - IT & engineering ('Inside Wires') and life sciences & chemistry ('Under the Microscope') - please select your preference(s) below:
Please select your practice area(s) of interest: