T 1943/15 - Drawings as prior art; open-ended ranges

Key points

• The Board, on drawings as prior art: “Features can therefore be taken from the drawings of a patent publication if their structure and function are clearly, unmistakably and fully derivable from the drawings.”
• The Board agrees with “the conclusion of T 204/83 that dimensions obtained solely by measurements carried out on a diagrammatic representation do not constitute part of the disclosure also applies to patent drawings that accurately portray an invention but do not qualify as construction drawings.”
• The Board, on sufficiency of disclosure: “According to established case law, an open-ended range in a claim does not necessarily result in a lack of sufficiency of disclosure due to the absence of a second boundary to the range. Where it is clear for a skilled person that an open-ended range is limited in practice, depending on the surrounding circumstances, such that the claimed values should be only as high as can be attained above the lower limit, then such open-ended ranges are normally not objected”

EPO T 1943/15 -  link

VII. The appellant [opponent] essentially argued as follows:
x)

Novelty

i) Dimensions from drawings

[Opponent's arguments]

According to decisions T 204/83, T 1488/10 and T 1664/06 there was no general rule for deriving features from drawings, since it was not possible to treat the drawings of each document in the same way. There was no general ban on measuring elements from drawings. In particular, it should be established whether the skilled person would derive any technical teaching from the drawing of a document involving the features in question. In fact, in each of the documents relied upon in opposition, the threads commonly played a relevant role.

When taking account of the relevance of elements shown in a drawing, the quality of the drawing and the background knowledge of the skilled person, it could be appropriate to measure the dimensions of the elements in order to determine the ratio of two dimensions. In that case, the measurements would only be taken to establish whether two dimensions were in proportion to each other. Sometimes, it sufficed for the skilled person to have a perception of a ratio from a drawing without requiring accurate measurements. This was confirmed by decisions T 748/91, T 422/95 and T 1200/05, according to which clearly essential elements may be derived even from schematic drawings.

Reasons for the Decision

2. Sufficiency of disclosure

2.5 According to established case law, an open-ended range in a claim does not necessarily result in a lack of sufficiency of disclosure due to the absence of a second boundary to the range. Where it is clear for a skilled person that an open-ended range is limited in practice, depending on the surrounding circumstances, such that the claimed values should be only as high as can be attained above the lower limit, then such open-ended ranges are normally not objected to (cf. T 487/89, Reasons 3.5; T 129/88 OJ 1993, 598, Reasons 2.2.4; T 1018/05, Reasons 2.3).

[...]

4. Dimensions from drawings

4.1 In the statement setting out the grounds of appeal the appellant [opponent] referred to various decisions concerning the measurement of dimensions in drawings, most of which built on the conclusions of decision T 204/83 (OJ EPO 1985, 310) that "When a feature is shown solely in a drawing without any other clarifying description a careful check should be made to establish whether the mere diagrammatic representation enables a person skilled in the art to derive a practical technical teaching therefrom" (Reasons 4) and that "Dimensions obtained solely by measurements carried out on a diagrammatic representation forming part of a document do not constitute part of the disclosure and cannot, therefore, be regarded as complementing the teaching obtained in the description" (Reasons 7). This led the board in case T 204/83 to deny that a patent drawing showing a coating apparatus in cross-section disclosed a height-to-diameter ratio larger than 0.5 and smaller than 0.66.

Similarly, the board in decision T 1488/10 decided that a length ratio between 0.5 and 0.75 did not form part of the disclosure of a patent drawing showing a cross-section of an ultrasonic transducer (Reasons 3.5).

In decision T 1664/06, with regard to various ratios between diameters and thickness values in a constant velocity joint, the board came to the conclusion that the prior-art document "discloses neither the dimensions of the outer and inner diameters of the cage specified in the present claim nor their size relative to other features and therefore cannot serve either directly or indirectly as a basis for determining the claimed ratios" (Reasons 2.1.2).

In decision T 451/88 (Reasons 2.4) the board referred to decision T 204/83 to conclude that the drawing of a prior-art tyre in a patent had to be distinguished from a scaled construction drawing such that it could not be derived from the prior-art drawing whether a ratio between two distances was between about 0.25 and 0.40.

In contrast, in T 748/91 (Reasons 2.1.1) in the context of an objection against added subject-matter, the board found that the feature whereby the profile depth of an intermediate layer was smaller than the thickness of a superposed surface layer could be read from a schematic cross-sectional view of a bearing, taking into account that the thicknesses of the layers and the ratio of these thicknesses to each other clearly formed essential items of the invention.

Likewise, in T 422/95 (Reasons 2.1.4) an amendment concerning the relative size of a gap and the thickness of a wall was considered to be clearly recognisable from a drawing in the patent application, considering that the patent drawing resembled a construction drawing.

In T 1200/05 (Reasons 2.2), the feature whereby an axle was fixed asymmetrically between two spring elements was deemed to be disclosed in a prior-art drawing that was quite precise.

4.2 According to established case law, it is a prerequisite for the acceptance of lack of novelty that the claimed subject-matter is "directly and unambiguously derivable from the prior art". This is no different when deriving dimensions from prior-art drawings, which are to be regarded as an integral part of a prior-art patent publication (T 169/83, OJ 1985, 193). Features can therefore be taken from the drawings of a patent publication if their structure and function are clearly, unmistakably and fully derivable from the drawings.

4.3 The board observes that patent drawings in the field of mechanical engineering, especially cross-sectional views, often reveal the complete structure of a device to a high degree of detail and go beyond a mere schematic representation of the essential features. Yet this does not mean that such drawings are to be regarded on a par with construction drawings, which may generally be relied on to show dimensions and proportions of elements to scale.

Therefore it stands to reason that the conclusion of T 204/83 that dimensions obtained solely by measurements carried out on a diagrammatic representation do not constitute part of the disclosure also applies to patent drawings that accurately portray an invention but do not qualify as construction drawings.

4.4 However, the question arises as to whether the relative size of dimensions or of ratios between dimensions can be inferred from patent drawings merely by visual perception.

The board concurs with decision T 204/83 (Reasons 4) that it would be impossible to lay down general rules and that each case will depend on the knowledge of the person skilled in the art and the way in which the feature in question is shown in the drawing.

4.5 In the present case, the dimensional constraints at issue set a lower limit for a ratio between two dimensions x and y (x/y >= 0.2 or y <= 5x) and compare two ratios of four dimensions x, y, u and v with each other (x/y <= u/v), respectively.

The board is not persuaded that the "=" variant of either constraint can be derived directly and unambiguously from a patent drawing, irrespective of the degree of detail it shows, without actually measuring the respective dimensions. Inferring from a visual perception only that a dimension y equals five times a dimension x or that four dimensions have the proportional relationship x/y = u/v verges on speculation and does not correspond to the standard of clear and unmistakable disclosure.

This is different when a strict inequality is claimed, which essentially defines an open-ended range. The board acknowledges that a detailed cross-sectional view may in some circumstances directly imply that one dimension is much larger than a second dimension (cf. T 748/91, T 422/95 or also T 1313/04, Reasons 2.2 and 2.3). Nevertheless, these cases pivot on the direct comparison of two dimensions (x > y or x/y > 1). This cannot be placed on an equal footing and judged in the same manner as a constraint that compares a first dimension with a multiple of a second dimension (x/y > 0.2 or y < 5x). Deducing how often a tooth width fits into a second tooth width from merely looking at a patent drawing is not straightforward and therefore tends to be ambiguous. Furthermore, any such comparison includes the mental step of apportioning a length in particular quantities, which can be considered as a way of measuring that length.