Sufficiency of disclosure is a fundamental requirement for patent applications, ensuring that a person skilled in the art can carry out a claimed invention. In this article we consider how the European Patent Office (EPO) may be altering its approach to interpreting sufficiency of engineering patents based on recent decisions in T 0149/21 and T 0500/20.

In particular, a patent is not necessarily insufficient just because there are examples falling within the scope of the claims which are not technically sensible or feasible. In such cases, a patent may still be considered sufficient so long as the skilled person is able, from the patent or from their common general knowledge, to identify the embodiments that would not be technically sensible or feasible.

Introduction: Sufficiency under the European Patent Convention (EPC)

Under Article 83 of the EPC, a patent application must “disclose the invention in a manner sufficiently clear and complete for it to be carried out by a person skilled in the art,” such that the skilled person is able to put the invention defined by the claims into practice without “undue burden”.  In other words, the skilled person must be provided with the information needed to make the alleged invention work.

It is not necessary to provide detailed explanation of every well-known feature of a claim, but it is necessary to at least “describe in detail at least one way of carrying out the invention claimed, using examples where appropriate and referring to the drawings, if any.” (Rule 42(1)(e)).

In the engineering field, different types of insufficiency objections can be raised in various circumstances: For example:

Definition by a result to be achieved 

• • An insufficiency objection can be raised against a claim if its features are defined by a result to be achieved, but the application does not provide the detail required to achieve the result.

Inventions contrary to the laws of physics

• • Where it is not scientifically possible to perform the invention, regardless of the details included in the application (for example a perpetual motion machine), then the EPO will raise an insufficiency objection.

AI-related inventions

• • In the context of AI-implemented objections, an insufficiency objection might be raised if there is lack of disclosure concerning the training data and methods required to achieve the results.
  • For example, if achieving a technical effect is dependent on particular characteristics of the training set, those characteristics must be disclosed by the application, unless the skilled person can determine them using the common general knowledge.

In other technical fields, different types of insufficiency objections can be more prevalent. For example, in the chemistry and life sciences fields, insufficiency objections can arise if the skilled person cannot put the invention into practice ‘over the whole scope of the claim’ (which is sometimes referred to as ‘Biogen insufficiency’). In these circumstances, if the description provides only one, or only a small number, of ways of putting the invention into practice, but the claim defines a range or group (of compounds, for example), an insufficiency objection can be raised if there is a doubt as to whether the invention can actually be worked for all embodiments falling within the scope of that claim (i.e. for all values of that range or embodiments of that group).

These kinds of objections have been less common in engineering fields at the EPO, and generally, non-working examples falling within the scope of the claims are not problematic.

As we will see from the cases below, that situation may be changing as the case law for sufficiency in engineering inventions continues to develop.

Below, we will look at the outcome of two cases, T 0149/21 and T 0500/20, and the common themes in those decisions. We will also see how concepts of these decisions featured in the more recent, citing case T 2401/19.

T 0500/20 – Windmills

In T 0500/20, the EPO considered insufficiency in relation to a claim to a wind turbine which comprised:

• a detection means to detect edgewise oscillations, and
• a yaw controlling system to change the yaw position of the wind turbine nacelle in order to eliminate edgewise oscillations.

Insufficiency was raised as a ground of opposition, and the opponent argued that for some wind turbine designs falling within the scope of the claim, it would not be possible to detect edge-wise oscillations because the resonance frequency of oscillations in the yawing direction would be too close to the frequency of the edge-wise oscillations. The Board disagreed, stating that these isolated cases of non-working or not ideally working embodiments were not relevant to the issue of sufficiency considering the large number of conceivable and realisable alternatives.

The Board reasoned that the opponent’s approach misapplied the Biogen approach. The Board even stated that such an approach can be rejected off-hand when there are no ranges defined. Furthermore, the Board decided that conceiving of an example which falls within the scope of the claim but does not achieve the claimed effect is not enough to demonstrate insufficiency. Such examples were deemed not to prove that the claimed concept does not work and to instead reflect the limitations that are inherent in any technological endeavour.

T 0149/21 – Rolling Mill

In T 0149/21, the Board of Appeal considered insufficiency in relation to a patent for reducing energy consumption of a mechanical rolling mill used to roll metal sheets. Claim 1 of the patent specified:

a. identifying a type of fault that led to an unplanned rolling break and feeding that to a control device, and
b. the control device determining the probable duration of the rolling break based on the identified type of fault.

In the described example, the probable duration is determined by reference to a “fault matrix”.

The Board considered the claimed invention to be insufficiently disclosed for two reasons.

Firstly, the Board noted that step (a) encompassed both manual identification and automatic identification of faults, but the skilled person would not be able to provide automatic identification from the patent’s disclosure or based on their general knowledge.

Secondly, the Board reasoned that the claim encompassed identifying both previously known faults and previously unknown faults. Since the determination of a probable duration of a rolling break based on a fault matrix can only be made for faults that are already known to that matrix, the Board reasoned that the disclosure is insufficient for the claim to be carried out over its entire scope.

The proprietor argued that it was sufficient that at least one way of carrying out the invention was shown, and that the requirement of whole range sufficiency was not applicable to mechanical inventions, citing the above-mentioned T 500/20 among other cases. The Board disagreed, reasoning that the requirement that the claimed invention be capable of being carried out over the entire claimed area is not limited by subject area.

The proprietor also argued that consideration of all the possible types of faults was excessive and used examples such as a nuclear bomb exploding or an airplane crashing into the rolling mill. In response to this, the Board made a distinction between scenarios which are objective and technically reasonable for the expert reader, and scenarios which are only theoretically possible.

The Board’s view was that the embodiments of automatic identification of a fault and of identification of previously unknown fault types were examples of embodiments falling within the scope of the claims which the skilled person would consider technically reasonable. As such technically reasonable embodiments of the claim would not be implementable based on the disclosure, the invention was found to be insufficient.

T 2401/19 – Small diameter video camera head

Within seven months of T 0149/21 (rolling mill) having been issued, it had already been cited in another decision of the Boards of Appeal. Specifically, in T 2401/19, which related to a small diameter video camera. Claim 1 of the patent related to a video camera defined by a number of features, including a video camera head comprising an objective lens assembly and a CMOS sensor, the claim defining that the maximum outer diameter of the video camera head is 1.1mm (i.e. the claim encompassed video camera heads with outer diameters between zero and 1.1mm).

The opponent argued that the patent did not disclose how to carry out the invention over the whole scope, in particular for very small values that still fell within the scope of claim 1 (e.g. below 0.5mm). In addition, the opponent argued that no reasonable lower limit for the maximum outer diameter was derivable from the patent.

The opponent referenced T0149/21 (rolling mill) to support their arguments, concluding that claim 1 specified a result to be achieved, rather than clearly defining a technical teaching of how to achieve said result. On the other hand, the proprietor argued that an implicit lower limit could be calculated for the outer diameter of the video camera head from the description and would disregard as nonsensical diameters too small to work.

The Board was not convinced, however, and found that the skilled person would not immediately disregard a smaller video camera head as being nonsensical based upon the size of components in the worked example.  This was because the patent discloses that “a compromise must be made based on the primary goal of the device, i.e. whether a small diameter is more important than a high-quality image”, the board concluded that the skilled person would have understood that the size could be further reduced at the expense of some image quality. However, the skilled person would not be able to determine the tipping point at which further size reduction would be considered to yield insufficient image quality for the invention to be considered worked.

Ultimately, it was decided that a lower size limit would exist, below which resolution would be too low and other technical constraints would come into play. However, the skilled person could not derive this lower limit, and so would not be able to distinguish between embodiments which are objective and technically reasonable for the expert reader and scenarios which are only theoretically possible.  Because of this, the Board found that the patent did not provide sufficient disclosure for the skilled person to put the claimed invention into effect across the scope of the claim.

T 0500/20 and T 0149/21: Sufficiently consistent?

On first reading, it may seem that the decisions in T 0500/20 (wind turbine) and T 0149/21 (rolling mill) are inconsistent in their interpretations of sufficiency, with T 0149/21 requiring sufficiency across the entire scope of the claim, and T 0500/20 disregarding non-working embodiments.  However, on closer consideration, there appears to be some common logic.

The interpretation used in T 0500/20 arguably applies only where the concept can be broadly defined, and the underlying principles and requirements for achieving the effects provided by an invention can be readily understood by the skilled person.

This is often the case in chemical patents where the effects of the invention rely on a property which is known to be common to a class of compounds. In such cases, it is generally reasonable to expect that a detailed disclosure in relation to one of the compounds would allow the skilled person to understand the underlying principles of the invention, and to apply the principles appropriately across the class of compounds. Analogously, in T 0500/20, the underlying principle of edgewise oscillation detection is apparent to the skilled person, and the skilled person can readily identify and exclude embodiments which would not work, despite theoretically falling within the scope of the claim.

This is consistent with the approach applied to the rolling mill of T 0149/21. The invention relies on “types” of disruptions causing machinery stoppages for a predictable length of time. However, as the claim wording encompasses various types of disruption which would be technically reasonable to the skilled person, but unknown to the control device making the required correlation, the disclosure was found to be insufficient for the skilled person to work the invention for all technically reasonable embodiments.

Moreover, in T 2401/19, the finding of insufficiency was based on the skilled person not being able to readily identify and exclude non-working embodiments which would fall within the scope of the claims.

Conclusion

In summary, it is common in these recent cases that a claim must be sufficient over the entire scope of its “technically reasonable” embodiments. All technically reasonable embodiments which fall within the scope of the claims must be sufficiently disclosed.

This requirement does not necessarily extend to examples that, despite falling within the claim scope, the skilled person could and would exclude as not being technically sensible or feasible. However, such an argument could only apply if the relevant skilled person could derive, from the patent or common general knowledge, whether or not a given embodiment is unreasonable.