If you follow this blog and attended the conferences hosted by AIPPI in February 2020, you will remember Luis Vilalta’s outstanding lecture on computer-implemented inventions (available in AIPPI’s publication about the XXXV Jornadas de Estudio). The lecture examined the European Patent Office’s (EPO) decisions on the fulfillment of patentability requirements by the invention under patent application number 03793825.5 (a computer-implemented simulation of the movement of a crowd of pedestrians through an environment).
This matter is particularly interesting because (i) the EPO had to determine whether a specific type of computer-implemented inventions (in which the method steps - a simulation in this case - exist only on a computer, with no link with external physical reality) can be considered to produce a technical effect and thus be patentable; and (ii) the decision could be generally applied to other computer-implemented inventions.
On March 10, 2021, the Enlarged Board of Appeal issued decision T 489/14 (the “Decision”) on case G1/19.
The Decision first recalls the requirements applicable to computer-implemented inventions. Patents can be granted for any inventions in all fields of technology, including computer programs. Article 52 of the European Patent Convention (EPC) covers abstract or intellectual activities, including computer programs, which would only be excluded from patentability if the patent exclusively refers to them “as such,” but not if they produce a technical effect. Therefore, a computer-implemented invention will not be excluded under article 52 EPC and thus be patentable if the claim explicitly refers to the use of a computer, a computer program or other technical means.
For an invention to be patentable, it must involve an inventive step (article 56 EPC). For the inventive step assessment, the EPO applies the problem-solution approach: (i) determining the “closest prior art;” (ii) identifying the objective technical problem to be solved by the invention based on the technical differences between the invention and the closest prior art, and on the technical effect of these differences; and (iii) considering if the claimed invention would have been obvious to a “skilled person” or an “expert” based on the closest prior art and the objective technical problem.
Regarding computer-implemented inventions, the Enlarged Board of Appeal considers that applying the problem-solution approach requires an intermediate step, i.e., determining the features of the invention that contribute to the invention’s technical effect. Individually considered, these features may be non-technical (e.g., a computer program as such), but considering the invention as a whole they may contribute to solve the technical problem. To determine whether an invention involves an inventive step, only the features that contribute to the invention’s technical effect must be considered.
The invention’s technical effect must go beyond the normal physical interactions between the program and the computer on which it is run. This effect will often be a technical effect on a physical entity in the real world. However, it may also be a technical effect within the computer, the system or the network.
The simulation examined by the Decision is a method comprising only numerical input and output, without interaction with external physical reality. However, according to the Enlarged Board of Appeal, a feature of a computer-implemented invention can contribute to the invention’s technical character not only if it relates to a technical effect at the input (e.g., measurement of physical values) or at the output (e.g., a control signal used for controlling a machine), there being no need for a direct link with physical reality in every case. Technical effects can also occur (i) within the computer-implemented method (e.g., adaptations to the computer resulting in a better use of storage capacity); or (ii) when the model used for the simulation constitutes the basis for a further technical use (in physical reality) of the simulation outcomes. However, in case (ii) above, these adaptations to the computer or computer system, or the further technical use, must be specified in the claims.
Some authors claim that this is a relevant decision because (i) it confirms that computer-implemented simulations are patentable; and (ii) the matters examined by the Decision are similar to artificial intelligence (AI) and machine learning issues, and thus can anticipate the EPO’s position on inventions in these fields.