What does G1/19 mean for patenting computer simulations in Europe?

The EPO’s Enlarged Board of Appeal has recently issued their decision in G1/19. The full decision can be found here, but what does this mean for the patentability of computer simulations?

The problem is that there exists a general exclusion to the patentability of computer programs under the European Patent Convention (EPC). It is typically required for a computer-implemented invention to bring about a technical effect outside of the mere interaction between the software and computer hardware in order for the invention to be patentable in Europe. However, there has been some uncertainty about whether computer simulations require special attention because they involve some inherently technical considerations, and also because the results of such simulations can bring about some improved technical understanding.

The good news is that the decision doesn’t rule out patenting computer-implemented simulations in Europe, although it does remain difficult. Perhaps most noticeable is that decision doesn’t give any specific criteria for assessing whether a computer-implemented simulation is patentable but does give some examples of previous cases that would now be considered patentable.

Here we delve into some of the key outcomes that arise from the decision.

Physical technical effect not necessarily required

Under some previous decisions of the Enlarged Board of Appeal, in order to be patentable, the claims for any simulation inventions (or design processes) had to include a physical step, such as the resultant manufacture of an article, to be considered to provide a technical effect.

From G1/19, it remains that the provision of a physical step is generally sufficient to establish a technical effect of the claimed invention, but the Enlarged Board also states that it is not necessarily a requirement, and that computer-implemented simulations without such a step can be patentable, at least in theory.

Technical effect

Indeed, the clear message from G1/19 is that this subset of computer-implemented inventions is not excluded from patentability, but also that they don’t enjoy an easier route to patentability than other computer-implemented methods, as may previously have been the case.

As a result, the decision largely focuses on the "technical effect" provided by any invention, and where technical effects could potentially arise. For example, the simulation may provide a technical effect if the simulation is particularly adapted to the internal functioning of the computer. Alternatively, the simulation may provide a technical effect if it receives an input, such as a sensor measurement, or produces an output, such as a control signal, provided the input/output are considered to provide a technical effect. Importantly, this output does not necessarily have to be a physical output, such as the manufacture of an article, but may be, for example, the output of data that has an implied technical use.

Assessing a mix of technical and non-technical features

A key outcome of the G1/19 decision is the confirmation that the COMVIK approach still applies to all computer-implemented inventions, including computer simulations.

The COMVIK approach is a specific way of applying the European problem-solution approach to computer-implemented inventions, to determine whether the claimed invention involves an inventive step. As with any other invention, the differences with respect to the closest prior art are identified, but only those features in the patent claim which are deemed to contribute to the solution of a technical problem, by providing a technical effect, are considered in the assessment of inventive step assessment, whereas the non-technical features are ignored.

This is not to say that all non-technical features should be omitted from the claims. Indeed, these features can be useful to provide context to an invention, and thus help to avoid clarity objections. However, it is worth noting that these features cannot be relied upon to differentiate the invention from the prior art, since they cannot contribute to the assessment of inventive step.

What does this mean for applicants?

It would seem that computer-implemented simulations should be treated like any other computer-implemented invention. In particular, your simulation invention is not excluded from being patentable, provided it can be directly linked to an effect which is deemed technical, physical or otherwise. This remains more favourable than the UK approach to simulation inventions, in which a physical technical effect outside the computer remains a requirement.

It may therefore be worth considering adopting a different filing strategy for simulation-based patent applications, for example to pursue UK protection via a European patent application, rather than a direct UK application if the technical effect is more abstract in nature.