EPO approach to examining computer implemented inventions


On Tuesday 25 May 2021, Heli Pihlajamaa, director of patent law at the European Patent Office, gave a presentation to CIPA’s computer technology committee on the Patentability of computer-implemented inventions after the Enlarged Board of Appeal’s decision in G 1/19.

G 1/19 relates to the assessment of inventive step in relation to computer-implemented simulations and considered whether to have a technical effect, a simulation must have a direct link to physical reality going beyond its implementation on a computer.

In the presentation, Pihlajamaa discussed how the decision of G 1/19 emphasised that the approach in T 641/00 applies to all computer implemented inventions including simulations. In T 641/00 it was decided that while a claim can have a mix of technical and non-technical features, novelty and inventive step can only be based on the technical features. In T 641/00 it was also stated that when assessing inventive step, the claim should solve a technical problem which the skilled person in the particular technical field might be asked to solve.

Pihlajamaa provided some key takeaways of the decision in G 1/19. In particular, Pihlajamaa emphasised that:

  • The technical effect of any simulation must be achieved over the whole scope of the claim i.e. substantially all embodiments falling under the claims must show the technical effect.
  • The simulation of a non-technical system (e.g. the weather or human behaviour) may contribute to technical character e.g. through automatically closing windows on a building based on a weather forecast.
  • There are no special rules for application of the approach in T 641/00 to a design process, but a design process is as a rule a cognitive activity.

Pihlajamaa also discussed how G/19 could be considered to depart from the decision in T 1227/05. T 1227/05 related to a resource-saving numerical simulation of a circuit subject to 1/f noise. In that case, the board considered that simulation can constitute an adequately defined technical purpose for a computer-implemented method provided that the method was functionally limited to that purpose.

In discussing how G 1/19 departs from T 1227/05, Pihlajamaa, emphasised that given the decision in G 1/19 it would not be possible to argue a claim is technical on the grounds of:

  • the accuracy of a simulation in representing physical reality;
  • the simulated system or process being technical;
  • the simulation reflecting technical principals underlying the simulated system;
  • the skilled person being from a technical field of the simulated system; and
  • the simulating being of an adequately defined class of technical items (this was the criteria in T 1227/05).

Pihlajamaa considered that G 1/19 was a positive decision since it brings a uniform approach to assessment of inventive step with regard to all CII categories and that the decision itself constituted a study book of possible technical (and non-technical examples). To this end, Pihlajamaa discussed examples from G 1/19 in relation to what can be considered technical. In particular:

  • A technical effect can be achieved by having a technical output e.g. a control signal for controlling a technical device. However, while output data can be technical, it was noted that cognitive content of data is as a rule not technical and calculated numerical data reflecting the physical state or behaviour of a system or process modelled in a computer usually does not contribute to the technical character of the invention.
  • A technical effect can be achieved by a technical input e.g. measurements.
  • A technical effect can be achieved by adaptation of the computer-implemented process to the internal functioning of the computer (e.g. better use of storage capacity or bandwidth). However, in the discussion afterwards it was noted this was unlikely to apply to a more efficient algorithm, since this is subject matter the EPO generally consider non-technical.

In addition, Pihlajamaa noted that a potential technical effect (i.e. a technical effect that would be produced when the output data were used according to an intended technical use) could be relied upon if the claim at least implied the intended technical use provided the claims did not also encompassed relevant non-technical uses. Pihlajamaa also noted that a virtual or calculated technical effect of the output data may exceptionally be relied on if intended use of the data is limited by the claim to only technical uses. However, a virtual or calculated effect cannot be relied on if non-technical uses of the simulated output are also encompassed by the claim e.g. gaining scientific knowledge about a technical or natural system.

In summary, in the presentation, Pihlajamaa noted that while G 1/19 marks a departure from T 1227/05 and could be considered to change how computer-implemented simulations are assessed, there are still several examples of when computer-implemented simulations can be considered technical. Thus, with careful drafting, it is still possible to argue that several forms of computer implemented simulation are technical and so it should still often be possible to obtain patents for this subject matter before the EPO.