T 0287/23 (Retroactive analysis in a safety system/OMRON) 10-06-2025

1. Technical background

1.1 The invention underlying the opposed patent addresses the difficulty of properly maintaining and diagnosing industrial safety systems after their initial deployment. When a safety operation causes a machine to stop, it can be difficult to identify the root cause, especially if the triggering event was transient or intermittent (e.g. a momentary interference with a sensor). According to the patent, this makes troubleshooting and verifying the correct performance of the safety components challenging.

1.2 The opposed patent therefore proposes a safety system that is designed for ex-post facto or retroactive analysis. This safety system comprises three main functional parts:

- a "safety controller" which executes a safety program;

- a "collection unit" configured to collect an "input value" over a predetermined period, where this input value is defined as the value of an input signal that was "input to the safety controller";

- a "visualisation unit" that uses this collected "input value" to reproduce a behaviour of the safety program over that predetermined period and then visually expresses the operating state of the program at an appointed point in time.

This is supposed to allow an operator to "replay" the safety program's past execution based on its actual input history, thereby facilitating the investigation of the cause of a safety operation such as a malfunction.

1.3 Figure 1 (reproduced below) of the opposed patent schematically illustrates the overall architecture of the claimed "safety system (1)". The safety system includes a safety controller (100), which is the central processing unit running the safety logic. The "safety controller" receives input signals from safety components, such as a safety sensor (16) (e.g. a light curtain), and controls output devices, such as a safety relay (14), which in turn can stop a motor (10).

FORMULA/TABLE/GRAPHIC

A "collection unit", represented conceptually by the arrow labelled "Collection of trace data", gathers trace data associated with the safety controller (100) over a period of time. A support device (200), which comprises the "visualisation unit", is configured to receive this collected trace data. This support device then performs the "Reproduction of behaviour", allowing an operator to analyse the past operational state of the safety program that was running on the safety controller.

2. Main request: claim 1 - inventive step

2.1 In Reasons 3.1.1 and 3.1.2 of the appealed decision, the opposition division found document D11 to disclose all of the features of claim 1 as granted.

2.2 Reasons 3.1.2 of the appealed decision only provides excerpts from D11's disclosure for each feature. It does however not explicitly identify the specific components in D11 that correspond to each feature of granted claim 1. In order to determine whether the appellant's challenge to the disclosure in D11 of the "collection unit" according to feature (c) and, consequently, the basis for the "reproduction" mentioned in feature (d) is well-founded, the board would therefore need to conduct its own profound analysis of D11.

2.3 Instead, the board accepts, arguendo, that D11 indeed does not disclose feature (c) and the sub-feature "based on the input value collected over the predetermined period" of feature (d).

2.4 The appellant ascribes to "the characterizing features of the invention" the technical effect that

- "it becomes possible to evaluate the safety system in the case of a malfunction or the like"

or, in a refinement, that

- "it becomes possible to evaluate the safety system in the case of a malfunction triggered by the input signal" (emphasis as in the original).

The board is not convinced that this technical effect can be credibly associated with features (c) and (d). In fact, feature (c) relates to the collection of an unspecified input value and feature (d) to the presentation of information derived from that value. Moreover, the board notes that claim 1 is not limited to any state of "malfunction". The input signals mentioned in feature (c) could therefore relate to the system's normal operation. Furthermore, the evaluation of a "malfunction" in a complex safety system typically depends on a multitude of factors. While a specific "input value" can be a trigger, it is rarely the sole cause. Other factors, such as the internal logic of the safety program, the internal state of the safety controller, malfunctions in output devices or environmental conditions, can also influence the system's overall behaviour. Correspondingly, the board does not consider the appellant's objective technical problem framed as "allowing a more accurate evaluation of a safety system" to be derivable from technical effects directly and causally related to the technical features of the claimed invention (see e.g. G 2/21, Reasons 25).

2.5 In a further refinement, the appellant argued that the technical effect of the distinguishing features (c) and (d) was to enable a retroactive analysis of the safety program's behaviour. It described a scenario where, by collecting an input value from a safety component (such as a light curtain) and using it to reproduce the program's behaviour, an operator could identify the factor that caused a safety stop (referring to paragraph [0056] of the patent description).

The board was not convinced by this line of argument for two reasons:

2.5.1 First, the technical effect of enabling a meaningful retroactive analysis is not credibly achieved over the entire scope of claim 1. The claim language is overly broad and covers embodiments where this specific technical effect would not materialise. To give an example, the board notes that claim 1 does not specify when the selection mentioned in feature (c) must occur. This means that the "input signals" are not limited to live operational data but could well be test signals from a storage medium (such as a DVD), having been "selected previously" during an early development phase. These test signals could in turn stem from a rudimentary, earlier version of the program, e.g. a version developed while in a debugging phase or as a minimum viable product. Likewise, the "behaviour" being reproduced in accordance with feature (d) is not specified and could in fact relate to the basic function of an early program version rather than the currently deployed operational logic. In such cases, which are actually encompassed by claim 1, the claimed process would offer no practical retroactive-diagnostic value for a deployed system. The appellant's scenario is thus only one advantageous embodiment whose effects however cannot be generalised to the claim as a whole.

2.5.2 Secondly, the board recalls that even if an advantageous technical effect is described in the patent specification, this is not sufficient for acknowledging an inventive step if it is not achieved by the invention as defined in the claims over their entire scope (cf. T 2010/22, Reasons 4.6.1).

2.6 In its preliminary opinion, the board had tentatively considered that features (c) and (d) could provide the technical effect of giving "some indication of which input values to take when the visualisation unit according to feature (d) reproduces the safety program's behaviour". However, upon further consideration of the breadth of claim 1 and on the basis of the parties' arguments exchanged at the hearing before the board, the board no longer adheres to this view, for the following reasons:

2.6.1 The phrase "selected previously" used in feature (c) provides no guarantee that the collected "input values" are indeed relevant to the current version of the safety program or its associated behaviour. As set out in point 2.5.1 above, it is well possible that the selection was made for a rudimentary, earlier version of the program. Using such outdated or contextually irrelevant "input values" to reproduce the behaviour of a different or more mature program version would, from an objective perspective, not provide a useful technical indication and could even be misleading.

2.6.2 The phrase "input signals input to the safety controller" mentioned in feature (c) does not, over the whole scope of claim 1, provide a reliable technical limitation that could justify an "indication" as referred to in the board's tentative technical effect. Claim 1 does not preclude that these signals may originate from a generic storage medium (cf. point 2.5.1 above). In such a scenario, there would be no verifiable link between the signals and the functional context of "the safety controller (100)". A skilled reader faced with signals on a storage medium would have, at least in general, no way of verifying their origin based on their intrinsic physical properties, such as amplitude or frequency. If the origin and relevance of the signals cannot be verified as to whether they stem from the specific "safety controller", they cannot provide a technically meaningful "indication" of which input values to take for a relevant reproduction.

2.6.3 The phrase "based on" used in feature (d) is also broad and does not imply a direct or an unmodified use of the collected "input value". It could encompass (significant) intermediate processing, where the collected input value is transformed before being used in the reproduction. If such an arbitrary processing is considered, any "input signal" could in principle be manipulated to produce a desired reproduction outcome. This would however render the choice of the initial input signal technically meaningless, implying that it could not serve as a credible "indication" that constrains the reproduction in any technical sense.

2.7 In conclusion, it is not apparent to the board how the mere "collection" of unknown "input values" as per feature (c) on the basis of which an undefined "operating state of the safety program" is supposed to be "visualised" in accordance with feature (d) can lead to a tangible technical effect. Therefore, the respondent is right that no technical effect can be credibly associated with the distinguishing features considered by the appellant, at least not over the whole scope claimed. Hence, it is not possible for the board to formulate an objective technical problem that is derivable from effects directly and causally related to the technical features of the claimed invention.

2.8 In the absence of a credibly solved objective technical problem, no inventive step can be acknowledged on account of distinguishing features (c) and (d). The board reaches this conclusion based on two separate, alternative lines of reasoning:

2.8.1 Following the principles established in G 1/19 (cf. Reasons 49, 82 and 124), the problem-solution approach may be concluded without formulating an objective technical problem if the distinguishing features do not credibly achieve any technical effect over the whole scope claimed (see e.g. T 746/22, Reasons 1.5 and T 1699/22, Reasons 2.7). Hence, features (c) and (d) cannot support an inventive step.

2.8.2 In the alternative, even when formally completing the problem-solution approach, the same conclusion is reached. The board finds the reasoning in case

T 1580/23 (Reasons 2.2.8) pertinent in this regard, where it was found that distinguishing features which solve no objective technical problem merely relate to an "arbitrary or non-functional modification" of the prior art and, as such, cannot contribute to an inventive step. The present board concurs with this approach and finds it useful to distinguish between these two terms. Distinguishing features can be regarded as being "arbitrary" when they represent a choice made without any technical reason or purpose, and "non-functional" when they have no objectively derivable technical function within the context of the claimed invention (see e.g. G 1/19, Reasons 33, last two sentences). While the former term focuses on the choice process and the latter on the outcome of the distinguishing features' implementation, both aspects are relevant to the case in hand.

2.8.3 Applying this framework to features (c) and (d), the board finds that the specific modification of the prior art associated with these features is not prompted by any technical consideration but represents a mere design choice. Therefore, using the terminology adopted in T 1580/23, the modification is "arbitrary". Likewise, since the board has already established that features (c) and (d) do not produce any credible technical effect, the resulting modification is

"non-functional". A modification that is demonstrably both "arbitrary" and "non-functional" cannot contribute to an inventive step. In agreement with the conclusion in T 1580/23, the board therefore finds that the subject-matter of claim 1 does not involve an inventive step.

2.9 Hence, the ground for opposition under Articles 100(a) and 56 EPC prejudices the maintenance of the patent as granted.

3. First auxiliary request: claim 1 - inventive step

3.1 Even when considering feature (g) not to be disclosed in D11, the board can identify no technical effect in relation to this feature that would be credibly solved over the whole scope claimed. The appellant did not provide any substantive arguments in this regard and merely referred to the technical effect mentioned for the main request (see points 2.4 and 2.5 above).

3.2 For reasons similar to those set out in point 2.6 above, the board no longer adheres to the tentative technical effect considered in its preliminary opinion for the first auxiliary request, namely that feature (g) allowed for a practical way to provide the "input values" for the "collection unit" mentioned in claim 1.

3.3 While feature (g) specifies that the "input values" are collected "from the safety controller", this does not cure the deficiencies inherent in the broad wording of the other features of claim 1. The reasoning set out in points 2.6.1 and 2.6.2 above therefore applies also to claim 1 of the first auxiliary request.

3.4 In particular, indicating the source of the "input values", as done in feature (g), cannot overcome the fact that claim 1 still fails to specify which "input values" are collected or to ensure that these values are relevant to the current version of the safety program. As explained in point 2.6.1 above, the "input

signal" that is "selected previously" could well be outdated or contextually irrelevant. Furthermore, even if this input signal was collected "from the safety controller", the board's reasoning as set out in point 2.6.2 above remains valid, given that claim 1 still does not preclude a scenario where the controller is merely processing rudimentary signals obtained from a generic storage medium without a verifiable link to its functional safety context. Moreover, specifying only that these potentially outdated or functionally irrelevant signals are sourced "from the safety controller" as per feature (g) fails to establish that a "practical way" of providing meaningful data for analysis is credibly achieved over the entire scope claimed.

3.5 The board therefore finds that feature (g) does not contribute to a credible technical effect, either. This again means that it is not possible to frame an objective technical problem for claim 1 of the first auxiliary request that is derivable from effects directly and causally related to the technical features of the claimed invention. As a result, neither can this claim 1 involve an inventive step (Article 56 EPC).

4. Third auxiliary request: claim 1 - inventive step

4.1 In support of inventive step regarding claim 1 of the third auxiliary request, the appellant argued that the additional feature (h) provided the technical effect of enabling a retroactive analysis of the safety program's behaviour. It explained that by including an "input value" from before the "collection condition" is established, the system captured the "run-up" to a state transition. This, in the appellant's view, allowed for a more accurate and detailed analysis by helping to identify which input changes were causative for the event. The appellant therefore considered the objective technical problem to be "to provide a safety system by which the behavior of the safety program can be analyzed more accurately and in greater detail".

4.2 This line of argument could likewise not convince the board. The alleged technical effect of a "more accurate and detailed analysis" is not credibly achieved over the entire scope of claim 1 for two reasons:

4.2.1 First, the claim does not provide any details regarding the term "collection condition" mentioned in feature (h). As a result, this "collection condition" is not limited to a specific type of operational situation, such as a "state transition" considered by the appellant. If the "collection condition" is arbitrary, the significance of the preceding "input value" is equally speculative and does not necessarily lead to a more accurate or more detailed analysis.

4.2.2 Secondly, and more fundamentally, feature (h) does not overcome the deficiencies already identified in the context of the main request. The analysis enabled by this feature can only be as accurate or detailed as the underlying data allows. As established in point 2.6 above, the claim covers embodiments where the collected "input values" are themselves outdated, contextually irrelevant or from an unverified source. Adding a (single) preceding data point of the same potentially poor quality cannot credibly make the overall analysis "more accurate" or "more detailed" in a technically meaningful way.

4.3 As a result, the board is not persuaded that the objective technical problem recited by the appellant is

credibly solved by the distinguishing features of claim 1 of the third auxiliary request.

4.4 As a consequence, no objective technical problem that is derivable from effects directly and causally related to the technical features of the claimed invention can be identified for claim 1 of the third auxiliary request.

4.5 This in turn means that present claim 1 cannot involve an inventive step, either (Article 56 EPC).