On March 10, the European Patent Office’s (EPO) Enlarged Board of Appeal published its keenly-anticipated decision G 1/19 on computer-implemented simulations.
G 1/19 holds that the EPO’s existing approach for assessing the patentability of computer-implemented inventions (CIIs) should indeed be used to assess computer simulation inventions. According to this approach, whether a claim has inventive step is assessed using only any features of the claim which contribute to the claim’s “technical character”.
In contrast to a leading earlier case, G 1/19 holds that it is not decisive whether a technical or a non-technical system or process is simulated. Simulation of a technical system is not necessarily technical.
Instead, the technical character of a simulation method may lie in a “direct link with physical reality” in the form of an input (eg, the measurement of a physical value used in the simulation) or an output (eg, the use of the result of the simulation as a control signal for a machine).
When the technical character arises from the direct link with physical reality of the output, ie, due to the use of the simulation result, that use has to be at least implicitly specified in the claim.
For example, if technical character arises because the simulation is part of a design process for a product, the claim must at least implicitly specify that the simulation result is used to produce the product.
Alternatively, even if a simulation method has no direct link with physical reality (eg, it is simulation of a non-technical process for a non-technical purpose), it may still be patentable, eg, if the simulation method is performed using a specially modified computer.
The decision is long (68 pages) and complex but the main points are as follows.
Background
The EPO’s general approach to patenting CIIs is called the Comvik approach (based on the Comvik decision, T 641/00). This holds that features of a claim are relevant to deciding whether the claim has an inventive step only if they contribute to the “technical character” of the claim.
For example, mathematical steps are not regarded as technical themselves, so whether they can be used to establish that a claim involves an inventive step depends on whether they contribute to a technical purpose defined by the claims. There are two ways this can happen:
- If the mathematical steps interact closely with the architecture of the computer, eg, if they are a reason for adapting how the computer is constructed. This possibility is rarely relevant to methods carried out on general-purpose computers.
- If the mathematical steps contribute to a “technical purpose” of the claimed method, such as controlling a mechanical system in the real world.
T 1227/05 applied the Comvik approach to computer simulation of physical systems. It held that simulation of a circuit subject to 1/f noise, or any other “adequately defined” physical system, constitutes a “technical purpose”.
This means that mathematical features of a claimed simulation method that result in improved simulation contribute to the technical purpose. Therefore, a claim directed to a simulation process can be patentable, even if the only novel steps are mathematical, and even if the claim does not specify how the simulation results are used.
“G 1/19 emphasises that a ‘direct link with physical reality’ is not the only way that claim features can contribute to technical character.”
Tim Watkin
Lara Sibley
Decision G 1/19
The questions referred to the Enlarged Board asked it to explain the relevant criteria for assessing whether a computer-implemented simulation claimed as such solves a technical problem. However, the board considered that it is impossible to give an exhaustive list of criteria for assessing whether a computer simulation, or indeed any other computer-implemented process, solves a technical problem by producing a technical effect. In fact, it refused to admit the question.
Instead, it answered a narrower question: whether a computer-implemented simulation necessarily solves a technical problem if is based on technical principles underlying a simulated system.
As a starting point, G 1/19 approved the use of the Comvik approach for assessing inventions for computer simulations, based on identifying the features which contribute to the “technical character” of the invention, and assessing whether they are inventive.
G 1/19 explained that “technical effects” or “technical interactions” can arise from technical input to the computer (such as a measurement of the real world) or technical output (such as a control signal for controlling a machine). In either case there is a “direct link with physical reality, based on features that per se are technical or non-technical, and this is in most cases sufficient to establish technicality”.
However, data obtained from simulating a technical process could be used for many purposes, such as to gain scientific knowledge or for teaching. Therefore, even if the data generated by a simulation has uses that are technical, the claim would not have a technical effect over the whole scope of the claim unless it specifies what technical use is made of the data.
On this basis, G 1/19 rejected the broad conclusion of T 1227/05 that simulation of a physical system defined by a claim is itself a technical purpose. It states: “Calculated numerical data reflecting the physical behaviour of a system modelled in a computer usually cannot establish the technical character of the invention. Only in exceptional cases may such calculated effects be considered implied technical effects (eg, if the potential use of such data is limited to technical purposes).”
Fact File
- Comvik decision: the EPO’s general approach to patenting CIIs, which rules that a patent is inventive if it contributes to a “technical character”.
- Production step: G 1/19 holds that the technical character of the invention arises because the simulation results are used to design a product.
More generally, a claim which achieves technical character via a “direct link with physical reality” based on the real-world use which it makes of the simulation results, must specify that use at least implicitly in the claim.
One of the questions referred to the board was whether the law is different if the computer-implemented simulation is claimed as part of a design process. The board held that it is not. Thus, if the technical character of the invention arises because the simulation results are used to design a product, then the claim must at least imply that the simulation result is used to produce the product, eg, the claim could recite a production step.
The decision declined to review decisions of national courts (eg, in the UK and Germany), which have reached different conclusions.
G 1/19 emphasises that a “direct link with physical reality” is not the only way that claim features can contribute to technical character. Alternatively, technical effects can arise within the computer process itself, eg, by specific adaptations of the computer or of a data transfer or storage mechanism, to make them better at performing the simulation. This comment is in line with the Comvik approach.
Discussion
G 1/19 is clearly explained, self-consistent and goes a long way to clarifying this important area of law. In future, it will be more important for the claims of patent applications for computer simulations carried out on general-purpose computers to define the invention as tied directly or indirectly to the real world, either with its input (eg, a measurement) or by a technical effect arising from its output (eg, control of a real-world apparatus).
Patent applications for design methods should mention details of any manufacturing step that can use the result of the design method.
Furthermore, G 1/19 has implications for CIIs more generally. For example, G 1/19 provides useful clarification that for any computer-implemented method “a direct link with physical reality is in most cases sufficient to establish technicality”.
This suggests that any computer-implemented method that takes sensor data as input (eg, a captured image, or recorded sound), or which has effects in the real world, should be held to be patentable.
Tim Watkin is a partner at Marks & Clerk. He can be contacted at: twatkin@marks-clerk.com
Lara Sibley is a senior associate at Marks & Clerk. She can be contacted at: lsibley@marks-clerk.com
Image: Envato Elements / KinoMaster
