Tag: inventive step – no

Sparsely connected neural network/MITSUBISHI – T 702/20 – 7 November 2022

In this decision, the EPO Board of Appeal provides a comprehensive technical and legal overview of a neural network “as such”.

Object of the Invention:

  • the application relates to a neural network (NN) apparatus
  • claim 1 differs from the closest prior art in that the different layers of a NN are connected in accordance with an error code check matrix

Examination Division:

  • distinguishing features do not serve a technical purpose and do not relate to a specific technical implementation
  • distinguishing features pertain to initial, fixed structural definition of an abstract mathematical neural network-like model with unknown input and output data

Appellant (part I):

  • claimed NN apparatus has a new and non-obvious structure
  • claimed NN solved a technical problem by providing effects within the computer related to the implementation of NN (storage requirements)

Board (part I):

  • the network structure only defines a class of mathematical functions which, as such, is excluded matter
  • while the storage and computational requirements are reduced in comparison with the fully-connected NN, this does not translate to a technical effect, for the simple reason that the modified NN is different and will not learn in the same way
  • the NN requires less storage, but it does not do the same thing
  • for instance, a one neuron NN requires the least storage, but it will not be able to learn any complex data relationship

Appellant (part II):

  • NN generally solve technical problems by automating human tasks
  • NN apparatuses are artificial brains and that artificial brains solve an automation problem, because they can carry out various complex tasks, instead of the human, without being programmed specifically for one task or another

Board (part II):

  • no evidence that NN functions like a human brain
  • whilst the functioning of NN may not be foreseeable prior to training and the programmer may not understand the significance of its individual parameters, the NN still operates according to the programming of its structure and learning scheme
  • the claims do not further specify any particular task, i.e. type of relationship to be learned, for the NN à NN does not solve any specific automation problem

Appellant (part III):

  • a technical problem may also be solved if the outputs of the system have an implied further technical use (G 1/19, reasons 137)

Board (part III):

  • claimed learning and use of the NN “to solve a classification problem or a regression problem” can use any data
  • outputs of the NN do not have any implied “further technical use”
  • outputs may be related to forecasting stock market evolution

Conclusion:

This decision provides a technical background regarding NN (reasons 7 to 8.2) and a legal background regarding NN (reasons 9 to 11.3), which provides a good insight into this topic.

Furthermore, the below figure shows according to G 1/19, point 85 and 86 how and when “technical effects” or “technical interactions” based on inter alia non-technical features may occur in the context of a computer-implemented process (the arrows in the figure above represent interactions and not abstract data). In this decision T 702/20 it was discussed whether the NN/ non-technical features contribute to the technical character of the invention via the output side, via the input side and via the technical implementation.

Virtual welding – T 2594/17 and T 2607/17 – 20 May 2021

In this case, a system for virtual testing and inspection of a virtual weldment was claimed. The system has, inter alia, no link to a “real” weldment and no technical implementation. Therefore, the non-technical features are not considered for inventive step and the system is not inventive.

Object of the Invention

  • The claimed system renders a 3D image, which can also be an animated 3D image (of a virtual weldment) and processes this image by simulating testing and inspection of the virtual 3D weldment for a user training. It then displays the processed image and allows the user to move it around.
  • Claim 1 differs from the closest prior art (general purpose computer) in that it is configured to display images of virtual weldments, of virtual testing of those virtual weldments, and of the results of the virtual testing on those virtual weldments. It can also determine a pass/ fail condition based on those results (for a user test).

Appellant

  • The system processed virtual weldments and this indicated that it was part of a virtual reality system for generating the virtual weldments (VRAW system).
  • The claimed system should therefore be considered as a feedback component of the VRAW system, which contributed to the generation of virtual weldments of better quality.

Board

  • The claims do not define or suggest any connection or relation of the claimed system with the VRAW system.
  • The board does not see the claimed system as a feedback component of the VRAW system
  • Only a virtual weldment is generated, mainly for training purposes, and without any link to a particular “real life” weldment.
  • It is not a sufficient condition that the simulation is based, in whole or in part, on technical principles underlying the simulated system or process (see G 1/19 e.g. points 1 and 2 of the Headnote).
  • –> not inventive

Conclusion

The below figure shows according to G 1/19, point 85 and 86 how and when “technical effects” or “technical interactions” based on inter alia non-technical features may occur in the context of a computer-implemented process. In this decision the software/ non-technical features do not contribute to the technical character of the invention via the input data/ output data/ technical implementation. Therefore, the non-technical features are not considered for the assessment of inventive step.

Pedestrians simulation/BENTLEY SYSTEMS – T 489/14 – 26 November 2021

The present decision is the reference decision to G 1/19, which was finally decided after decision G 1/19, as explained below. Among other things, it dealt with the question of whether the output data of a simulation result in the non-technical features of the simulation making a technical contribution and therefore being taken into account for the assessment of inventive step. At the end (conclusion), a bridge is again built to the decision G1/19.

Subject matter main request

  • The invention in the ex-parte case relates to a simulation of the movement of pedestrians through an environment (e.g. a building like a stadium or railway station).
Pedestrians simulation/BENTLEY SYSTEMS – T 489/14 – 26 November 2021

Board

  • The data produced by the method of claim 1, which reflects the behaviour of a crowd moving through an environment, does not contribute to a technical effect.
  • The potential use of such data is not limited to technical purposes, as it can be used e.g. in computer games or presented to a human for obtaining knowledge about the modelled environment.
  • Computer games and presentation of information are examples of non-technical uses of data falling within the scope of claim 1.
  • Therefore, there is no technical effect over the whole claimed scope derivable and the non-technical features are not contribute to the technical character.
  • –> no inventive step

Subject matter auxiliary request

  • Claim 1 of an auxiliary request additionally comprises the use of a profile for a pedestrian based on a set of measured attributes (in response to G 1/19, point 85: “technical input may consist of a measurement”).

Appellant

  • The additional feature leads to a “more accurate simulation”.

Board

  • A “more accurate simulation” per se cannot contribute to a technical character (cf. G 1/19, point 111).
  • Claim 1 does not contain a measurement step, but merely indicates that the attributes are somehow measured.
  • Therefore, no technical effect is derivable.
  • –> no inventive step

Conclusion

  • The below figure shows according to G 1/19, point 85 and 86 how and when “technical effects” or “technical interactions” based on inter alia non-technical features may occur in the context of a computer-implemented process (the arrows in the figure above represent interactions and not abstract data).
  • The subject matter of the main request deals with output data of a simulation, wherein the use is not limited to technical purposes.
  • According to G1/19, points 94 and 95: With respect to potential technical effects, the Enlarged Board added that if claimed data or data resulting from a claimed process was specifically adapted for the purpose of controlling a technical device, the technical effect that would result from this intended use of the data could be considered “implied” by the claim. However, this argument could be made only if the data had no other relevant uses, since otherwise the technical effect was not achieved over substantially the whole scope of the claimed invention.
  • Furthermore, the subject matter of the auxiliary request deals with a profile for a pedestrian based on a set of measured attributes. According to the appellant, this should represent input data in the form of a measurement.
  • In G 1/19 the Enlarged Board had pointed out that technical effects could occur at the input of a computer-implemented process (see figure above) and that technical input could consist of a measurement.
  • The question to be answered is whether the input data/ measuring step and the simulation method are not merely juxtaposed, the output of the measuring step serving as an input for the simulation method, but interact to produce a combined technical effect.
  • An interaction may be present, for example, if the combination amounts to an indirect measurement of a specific physical entity by means of measurements of another physical entity (see G 1/19, point 99).
  • The method of claim 1 provides information about the movement of simulated pedestrians through a modelled building structure. Since the calculated information is neither used in a further step of the method nor specifically adapted for the purposes of an intended technical (and only relevant) use, it has to be investigated whether the information represents a measurement of a physical entity.
  • The modelled building structure does not correspond to a building structure that was measured, whether directly or through measurement of its physical effects on other physical entities.
  • In fact, the modelled building structure need not correspond to any existing building structure. Nor does the calculated information about the movement of simulated pedestrians represent a direct or indirect measurement of any of the real pedestrians (or other physical entities) that were measured in the process of generating the pedestrian profiles.
  • Hence, in the present case no physical entity (or process) can be identified which could potentially be measured by the method of claim 1 in the sense that its physical status or some physical property is described by information calculated on the basis of data obtained by a direct or indirect physical interaction with the entity.

Rain-sensitive parcels/IVECO – T 1806/20 – 17 November 2023

At the EPO, data relating to software-related inventions may, under certain conditions, be regarded as data of a technical character. This means that such data will be taken into account as a distinguishing feature in the examination of inventive step. Data may have a technical character if they are so-called “functional data”. A typical example of “functional data” is data intended to control a technical device. In the following decision T 1806/20, the appellant tried to present “cognitive data” as “functional data”. The appellant relied on the earlier decision T 1194/97, which identifies a type of test method that can be used to determine whether the data are “functional data”.

Object of the Invention

  • A parcel delivery system that seeks to prevent damage to water-sensitive parcels by avoiding delivery to rainy destinations.
  • Claim 1 differs from the closest prior art (CPA) inter alia by rescheduling of the delivery based on the parcels’ sensitivity to water and a rain forecast.

Appellant

  • The prevention of damage to physical objects was a fundamental technical problem that was addressed throughout various areas of technology.
  • The appellant did not dispute that delivering multiple parcels at different destinations and planing such delivery constituted a non-technical logistics scheme.
  • The rescheduling of the delivery based on the parcels’ sensitivity to water and the rain forecast do not form part of the non-technical logistic scheme.

Board

  • Contrary to the appellant’s view, the Board judges that the requirement to ensure that parcels do not get damaged by water forms part of the non-technical logistics scheme.
  • The Board is not convinced by the argument that information about a parcel’s water-sensitivity is functional technical data, because its loss would impair the technical operation of the system (cf. T 1194/97, reasons, point 3.3).
  • It is self-evident that if a piece, either technical or non-technical, of any invention is taken out, it would not work as designed.
  • In the Board’s view, what T 1194/97 is saying is rather that the loss of functional data would make the system inoperable at the technical level.
  • In contrast, if cognitive data is lost, the system would still work but possibly produce results that would be unintended for non-technical reasons.
  • Thus in T 1194/97, the loss of functional data prevented the system from generating any television picture, whereas the loss of cognitive data only resulted in a meaningless television picture resembling snow.
  • In the present case, the loss of water-sensitivity information would not cause the system to stop working; the vehicle would still be guided, and parcels would be delivered. However, it would result in leaving water‑sensitive parcels standing in the rain – an unintended operation comparable to producing a television picture that resembles snow. The reasons why these outcomes are unintended are non-technical. In T 1194/97, it was the cognitive meaninglessness of the television picture to a human viewer; in the present case, it is the prevention of rain damage to a parcel. Hence, judged by the consequence of its loss, the water-sensitivity data is equivalent to cognitive rather than functional data.

Conclusion

This decision deals with “functional data”. Such data is intended for controlling a technical device may be considered to have technical character because it has the potential to cause technical effects. In G1/19, reasons, point 94 the Enlarged Board of Appeal has generalized this as follow:

In the context of the problem-solution approach and the COMVIK approach, such potential technical effects may be considered if the data resulting from a claimed process is specifically adapted for the purposes of its intended technical use. In such cases:

  • either the technical effect that would result from the intended use of the data could be considered “implied” by the claim, or
  • the intended use of the data (i.e. the use in connection with a technical device) could be considered to extend across substantially the whole scope of the claimed data processing method.

Data is considered to be functional data when its loss would impair the technical operation of the system (cf. T 1194/97, reasons, point 3.3). This means, data is considered to be functional data, when the loss of the data would make the system inoperable at the technical level. In contrast, if cognitive data is lost, the system would still work but possibly produce results that would be unintended for non-technical reasons.

E.g. in T 1194/97, the loss of functional data prevented the system from generating any television picture, whereas the loss of cognitive data only resulted in a meaningless television picture resembling snow.

E.g. in T 1806/20, the loss of water-sensitivity information would not cause the system to stop working; the vehicle would still be guided, and parcels would be delivered. However, it would result in leaving water‑sensitive parcels standing in the rain – an unintended operation comparable to producing a television picture that resembles snow. The reasons why these outcomes are unintended are non-technical. In

T 1194/97, it was the cognitive meaninglessness of the television picture to a human viewer; in T 1806/20, it is the prevention of rain damage to a parcel. Hence, judged by the consequence of its loss, the water-sensitivity data is equivalent to cognitive rather than functional data.