T 1996/20 04-07-2023

Reasons for the Decision

1. The present decision only deals with the issue of inventive step of the subject-matter of claim 1 as granted.

2. In the decision under appeal the opposition decision inter alia found that the 3D printer of claim 1 was obvious to a person skilled in the art having regard to a combination of documents D4 and D5 (cf. point 9.1 of the reasons).

Document D4

3. Document D4 is a disclosure made available on the internet in the form of a blog entry published at a certain date followed by a series of user comments posted at a later point in time. In the present case, the board has no doubt that the content of the blog entry dated 15 June 2013 ("2013 06/15") as well as the first three comments on page 6/7 of 16 June 2013 and 10 October 2013 were made available to the public before the priority date of the patent in suit, i.e. before 29 January 2014. With the exclusion of the last five comments of pages 6/7 and 7/7, posted between 3 December 2014 and 16 March 2016, document D4 therefore constitutes prior art under Article 54(2) EPC.

4. The appellant did not contest the selection of document D4 as a starting point for the inventive step assessment, nor did it challenge the opposition division's finding that features 1.1 to 1.6 and 1.9 were known in combination from document D4. The board has no reason to doubt these findings. Indeed, the author of the internet disclosure recounts their experience with a sheet of flexible "Lexan" material as build surface on a 3D printer. A textured top surface is not disclosed by document D4. Since the purpose clause of feature 1.8 is linked to the textured top surface of feature 1.7, both features 1.7 and 1.8 must be regarded as distinguishing features with respect to document D4.

The objective technical problem

5. In point 9.1.2 of the decision under appeal, the objective technical problem was formulated as "to improve the precision of the printing process by avoiding a relative movement between the printed material and the plate". However, the flexible build plate of document D4 already achieves a strong bond with the print material (cf. page 4/7: "The final print results are great: PLA sticks to the Lexan super good"). Consequently, a more accurate formulation of the objective technical problem is to further improve the bonding of the print material to the build plate. This was not challenged by the appellant at the oral proceedings held before the board.

Document D5

6. Document D5 is concerned with 3D printers. In paragraph [0004] various prior art solutions are disclosed for depositing print material on a substrate removably affixed to a modelling platform. One of these solutions uses a flexible sheet substrate (cf. paragraph [0011]). The importance of a strong adherence of the print material to the substrate is repeatedly acknowledged in paragraph [0006] of the section 'Background of the invention'. In paragraph [0014] a solution is then offered for removably mounting and locking a substrate onto the modelling platform.

7. Despite of being characterised as "substantially rigid" (cf. paragraphs [0015] and [0032], claim 1), the substrate of the preferred embodiment of document D5 comprises a tray 50 which, according to paragraph [0032], "can be flexed slightly so as to assist in releasing a completed model from the substrate 16". This particular property of the substrate is re-iterated towards the end of the detailed description where it is used to "release models from the substrate, either as the sole release technique or in combination with use of the cleaning tank" (cf. paragraph [0046]).

8. The board is therefore satisfied that the teaching of document D5 is sufficiently close to the disclosure of document D4 so that the skilled person would have been prompted to at least consider its teaching when looking for a solution to the objective technical problem.

9. In paragraph [0031] of document D5, the tray 50 illustrated in Figures 3 to 6 is described as follows:

"The tray 50 has an upward-facing, substantially flat modeling surface 52, which has a texture selected for desired adherence and release characteristics according to the particular modeling process parameters and modeling material used to build models on the substrate 20. The texture may range from smooth (rms roughness less than about 0.001 in.) to rough (rms roughness greater than about 0.001 in.). [...] The modeling surface 52 optionally includes a grid 58, defined by grooves in the modeling surface. The grid 58 facilitates accuracy in the positioning of models on the substrate 16."

Further references to a textured modelling surface of the substrate are included in dependent claims 5 to 8, 21 to 23, 38, 41 and 42 of document D5.

10. In the board's view, the skilled person would not have ignored this clear and unambiguous teaching in document D5. This explicit solution to the objective technical problem would have prompted the skilled person to modify the substrate, i.e. the flexible Lexan sheet, of document D4 by texturing its top surface. In so doing, the skilled person would have arrived at the subject-matter of claim 1 of the patent as granted. Hence, the claimed subject-matter is obvious.

Razor blade argument

11. In a first line of argument, the appellant maintained that documents D4 and D5 were incompatible in view of the fact that the razor blade scraper mentioned in document D4 would damage a textured surface. This argument is not found convincing. The corresponding statement on page 4/7 of document D4 only concerns smaller items printed on a Lexan sheet. The same passage suggests that larger 3D prints 'pop off' by bending the flexible sheet. Even if this suggestion leaves some room for speculation (see the author's remark in the passage: "(yet to print anything 'big')"), it at least shows that razor blade scrapers may not be indispensable for removing a 3D print in the context of document D4.

12. Moreover, paragraph [0017] of the patent in suit indicates that a textured surface actually facilitates the removal of items from a flat sheet of flexible material ("does not bond permanently to the surface 104 of the sheet 102 for ease in removing the 3D print 103 from the surface 104 of the sheet 102"). This points away from the idea that a razor blade would be indispensable to remove the 3D print off a flexible sheet with a textured surface.

13. The same applies to the internet disclosures D21 to D24 filed by the appellant in support of its argument that documents D4 and D5 were not compatible. While highlighting the risks of using a sharp object such as a blade or razor to detach 3D prints, document D21 actually provides a list of alternative solutions to razor blade scrapers (see pages 4/12 to 6/12: applying heat or cold, or using dental floss or solvents). Document D21 even contains a suggestion to use flexible plates as a way of easily removing the 3D print (page 7/12: "they are designed to be [...] flexed so that the 3D print simply pops off"). Also document D24 proposes a flexible printing surface as an alternative to scraping (cf. page 3/6: "That's when I discovered NinjaPlates [...] It's basically a flexible printing surface, print the part, let it cool, and give it a good twist"). Documents D22 and D23, in turn, recommend to cool down the heated build platform and to apply dental floss, respectively, in order to remove the 3D print. Hence, none of these documents establishes an absolute requirement for using razor blades in order to scrape off a 3D print from a build plate, let alone an incompatibility between the disclosure of document D4 and the textured surface of document D5.

Documents D15, D16, D21, D23 and D24

14. Documents D21, D23 and D24, as well as further documents D15, D16 relied on by the appellant, have been made available on the internet only after the priority date of the patent. Hence, these documents cannot be regarded as prior art under Article 54(2) EPC.

15. The appellant argued with reference to decision T 1110/03 (OJ EPO 2005, 302) that documents D15, D16, D21, D23 and D24 could nevertheless be used to ascertain facts prior to the priority date.

16. In decision T 1110/03, it was held that a document which is not itself part of the state of the art can provide a basis for an inference about, e.g. the state of the art, common general knowledge in the art, issues of interpretation or technical prejudice etc - an inference which is subject to challenge as to its plausibility (cf. Reasons 2.1). By way of example, decision T 1110/03 mentions a dictionary as an account of the meaning of words and a technical review article as an account of the common general knowledge (cf. Reasons 2.3).

17. Neither of these examples applies in the present case. Although documents D14, D17 and D22 confirm that some aspects, such as the use of Kapton tape, blue painter's tape and hairspray, mentioned in the internet disclosures D15, D16, D21, D23 and D24 were known before the priority date of the patent in suit, the board is not convinced that the entire content of the post-dated blog entries, online forum discussions and question-and-answer sessions, which are not on a par with dictionaries or technical review articles, can be considered indirect evidence for the common general knowledge of the person skilled in the art in the field of 3D printing.

Heated build plate argument

18. A further argument put forward by the appellant against a combination of documents D4 and D5 is that the first document concerns a 3D printer with a heated build plate (HBP) and the second document does not. The board is not persuaded by this argument, since the skilled person would have implemented only the features from document D5 that solve the objective technical problem. In the present case, the question of obviousness is not so much whether the skilled person would have substituted the substrate of document D5 for the Lexan sheet of document D4. Rather, it boils down to the assessment of what document D5 would have taught the skilled person in terms of how the bonding of the print material to the build plate can be improved. And the teaching in document D5 pointed in the direction of texturing the build surface. The fact that the tray 50 shown in Figures 3 to 6 of document D5 has a certain thickness and a matrix of ribs 60 at its lower surface would not have discouraged the skilled person from following the teaching of document D5 and providing texture on the build surface of the 3D printer according to document D4.

Dominant thinking, prejudice

19. The appellant's submissions concerning the "dominant thinking" and the "prejudice" at the time of filing seem to miss the point that document D4, the starting point for the inventive step assessment, already discloses an alternative to the Kapton tape, the blue painter's tape and the hairspray, namely a build plate consisting of a flat sheet of flexible material (cf. page 1/7: "Furthermore, if you print in ABS, there's a good chance you'll somehow mangle the Kapton tape [...] or printing with PLA, rip up your blue painters tape" and page 4/7: "Plus I can cut back on my blue-tape usage"). Even if the solution of document D4 had not been widely known to all users in the 3D printing community, this does not mean that there was an unsolved "tape/glue paradigm" at the priority date of the patent in suit. The fact that some other internet sources D14, D17 or D18 selected by the appellant focused on solutions applying thin disposable layers anew on the build plate for each 3D print does not prevent a finding of lack of inventive step over a combination of document D4 with the teaching of document D5, which was publically available already in 2005.

One of many feasible solutions

20. It may very well be that the textured top surface is just one of several different feasible solutions to the objective technical problem. But it is the only solution to the problem offered by document D5. Neither the chemistry of the build plate nor a change in printing material is in focus in document D5. The appellant has thus not convincingly shown that the teaching of document D5 would have prompted the skilled person to work in a different direction than modifying the surface structure. Regarding the teaching of documents D15 and D16, they do not form part of the state of the art pursuant to Article 54(2) EPC (cf. point 14. above), nor can they be regarded as indirect evidence of the common general knowledge at the priority date of the patent in suit (cf. point 17. above). But even if they were, the content of these documents would not change the inventive step assessment in view of the combination of documents D4 and D5.

Long-felt need

21. The board does not dispute the appellant's argument that producers of 3D printers at the time the application for the patent in suit was filed relied on Kapton foil, blue painter's tape or hairspray. Documents D14 and D17 are proof thereof. However, this does not imply a long-felt need that was overcome by the claimed invention and that is therefore indicative of an inventive step. By providing a flexible sheet as build plate on top of the 3D printer platform, document D4 already proposed before the priority date of the patent in suit a measure to satisfy any need there may have been in the field. In fact, so did document D5 in 2005, at least in the context of the prior art discussed in paragraph [0011] ("The Stratasys**(®) modeling machines have also used a flexible sheet substrate held down on a platform by vacuum forces, as is disclosed in the Comb et al. '008 patent"). Moreover, the board agrees with decision T 1014/92 (Reasons 4.7) that, in absence of a long-felt need, a finding of obviousness based on an objective evaluation of the state of the art cannot be affected by the fact alone that two documents had not been combined by skilled person for a considerable period of time.

Conclusion

22. In sum, the board shares the opposition division's view that the subject-matter of claim 1 as granted does not involve an inventive step so that the ground for opposition under Article 100(a) EPC in combination with Article 56 EPC prejudices the maintenance of the patent as granted. The appeal must therefore be dismissed.