T 2006/13 () of 10.1.2017

Summary of Facts and Submissions

I. European patent No 2 231 924 (in the following: "the patent") concerns a method for dealing with faults occurring during the manufacture of a material web.

II. The patent as a whole was opposed for lack of inventive step. The opposition division decided to reject the opposition.

III. This decision was appealed by the opponent (in the following "the appellant").

IV. With the summons to oral proceedings, the Board sent a communication pursuant to Article 15(1) of the Rules of Procedure of the Boards of Appeal (RPBA) indicating its preliminary opinion of the case.

V. Oral proceedings before the Board were held on 10 January 2017, for the course of which reference is made to the minutes. As announced in its letter of 12 September 2016, the duly summoned patent proprietor (in the following "the respondent") was not present at the oral proceedings.

VI. Requests

The appellant requested that the decision under appeal be set aside and the patent be revoked.

The respondent requested that the appeal be dismissed (main request), alternatively that the patent be maintained in amended form on the basis of one of the sets of claims filed as auxiliary requests 2 to 14 with the reply to the grounds of appeal dated 16 April 2014.

VII. Prior art

(a) In the statement setting out the grounds of appeal, and in the reply to it, the parties relied on the following documents which were filed in the opposition proceedings and are cited in the decision under appeal:

D2: WO 2004/044316 A1

D3: "Digital Valves - Intelligent Paper Production Relies on Fully Digital High-response Valves", ipw (International Paper World), 6/2004, page 52

D4: DE 1 054 953 B1

D6: DE 42 43 045 C2

D7: Roeker, G., "Fluidtechnik - Ventilsteuerung im Pulsmodulationsverfahren", Der Konstrukteur, 6/91, pages 32 and 33

D8: DE 80 21 790 U1

Of these D7 and D8 were filed after expiry of the opposition period.

(b) In its submission dated 8 December 2014, the appellant relied on the following document for the first time:

D9: US 4,170,245 A

VIII. Claims of the respondent's main request

Independent method claim 1 as granted reads as follows:

"1. A method for dealing with faults occurring during the manufacture of a material web, said method comprising at least the following characterizing steps of:

- identifying the location of a fault (H) on the surface of a roll (5) or in a material web and estimating or calculating an arrival time of the fault (H) at a roll nip (N) and a dwell time for the fault in the roll nip,

- transmitting from a control system (4) to at least one digital valve unit (100), used for controlling a nip pressure of the roll nip (N), a feed-forward adjustment instruction (F) for dealing with a fault in the roll nip (N) of a calender,

- reducing as per adjustment instruction (F), through the intermediary of at least one digital valve (1), a nip pressure of the roll nip (N) at the latest when the fault (H) arrives at the roll nip (N) and increasing the nip pressure as the fault exits the roll nip."

Independent apparatus claim 11 as granted reads as follows:

"11. An apparatus for implementing a method as set forth in claim 1 for a calender with at least one roll set, which is mounted on a calender body and which comprises at least two rolls (5) between which remains a roll nip (N), enabling a material web to be calendered therebetween, characterized in that the apparatus includes

- a control system (4), which comprises means for identifying the location of a fault (H) in the material web or on the calender roll (5), means for estimating or calculating the arrival time for a fault approaching a roll nip (N) between two rolls, means for estimating or calculating the dwell time for a fault in the same roll nip (N), as well as means for developing a feed-forward adjustment instruction (F), said adjustment instruction enabling the fault (H) arriving in the roll nip (N) of a calender to be dealt with,

- an adjustment system, including at least one digital valve unit (100) which enables the nip pressure of a roll nip (N) to be adjusted according to an adjustment instruction received from the control system by a feed-forward adjustment mode through the intermediary of at least one hydraulic actuator (2) in such a way that the nip pressure of a roll nip can be reduced according to the adjustment instruction (F) through the intermediary of at least one valve (1) in the digital valve unit (100) at the latest when the fault (H) arrives in the roll nip and the nip pressure can be increased when the fault exits the roll nip (N)."

Dependent claims 2 to 10, 12 and 13 define preferred embodiments of the method of claim 1 and the apparatus of claim 11 respectively.

IX. The arguments of the parties, insofar as relevant for the present decision, can be summarised as follows:

(a) Consideration of D7, D8 and D9 in the proceedings

Appellant's case:

Documents D7 to D9 should be taken into consideration because they are highly relevant for the question of inventive step.

Respondent's case:

D7 and D8 should be disregarded because the opposition division already decided not to admit these late-filed documents into the proceedings. D9 should not be admitted into the proceedings because its filing is belated and its content is prima facie no more relevant than that of D2 or D3.

(b) Main request - Inventive step

Appellant's case:

The subject-matter of claim 1 lacks an inventive step in light of D4 alone or in combination with D2 or D3.

As acknowledged by the opposition division, D4 represents the closest prior art. It discloses a method for dealing with faults, such as a thickening, which occur during the calendering of a material web in a multi-roll calender (column 1, lines 1 to 6), which method comprises the steps of:

- identifying the location of a fault in a material web,

- transmitting from a control system (21) to a valve, which is used for controlling the nip pressure of a roll nip, a feed-forward adjustment instruction for dealing with the fault in the roll nip of the calender (column 4, lines 26 to 28; column 5, lines 19 to 24), and

- reducing as per adjustment instruction, through the intermediary of the valve, the nip pressure of the roll nip at the latest when the fault arrives at the roll nip and again increasing the nip pressure as the fault exits the roll nip (column 3, lines 3 to 28; column 4, lines 33 to 69).

The opposition division came to the conclusion that the subject-matter of claim 1 differs from this method only in that:

- the arrival time of the fault at the roll nip and the dwell time for the fault in the roll nip are estimated or calculated, and

- the valve is a digital valve.

However, the first feature is at least implicitly disclosed in D4 for the following reasons.

The teaching of D4 is directed to a skilled person having a university degree and several years of practical experience in the high technological field of paper technology, as evidenced by the university programme in paper technology at Munich University of Applied Sciences.

The calender of D4 comprises an elastic roll (column 3, line 66) and thus is specifically designed for calendering a paper web. The skilled person knows that this paper calender runs at such a high speed that the paper web passes through it in less than a second. For instance, D3 discloses that paper machines run at a speed of more than 2000 meters per minute (first column, second paragraph). D4 aims at minimising the length of uncalendered web in case of a fault in the web (column 1, lines 35 to 41; column 3, lines 29 to 33). It teaches that the steps of momentarily opening and again closing the roll nip are performed either manually by an operator, or automatically by a control device (column 3, lines 3 to 7).

The skilled reader of D4, using his common general knowledge in paper machines, would immediately understand that, in either case, the arrival time of the identified fault at the nip as well as its dwell time in the nip must be estimated or calculated, in order to guarantee that the roll nip be opened and re-closed only as and when needed, whatever the calendering speed and the switching delays (column 1, lines 29 to 31). Otherwise, the aim of D4 would not be achieved, in particular when the web fault passes through the calender in less than a second. When the nip is manually opened and re-closed by the operator, having identified a fault in the paper web approaching the nip, he must mentally estimate or calculate the arrival and dwell times so as to actuate the opening/re-closing of the nip in a precise and adequate manner and thereby avoid damage to the web or the roll while minimising the length of uncalendered web. When instead the nip is automatically actuated by a control device, a sensor is used for detecting that the web fault has exited from the roll nip (column 2, lines 51 to 54). Again, to minimise the length of uncalendered web whatever the calendering speed and the switching delays, the arrival and dwell times must be estimated or calculated to actuate precisely the opening and re-closing of the nip.

In conclusion, the method of claim 1 differs from that disclosed in D4 only in that a digital valve is used in place of the check valve.

This distinguishing feature increases the rapidity and accuracy of the nip pressure control in response to web faults. It solves the problem of further reducing the length of uncalendered web in case of web faults, as acknowledged by the opposition division (see point 2.2.2.2 in the appealed decision).

The claimed solution to this problem is obvious in light of common general knowledge. The check valve described in D4 works in the same manner as a digital valve: it has two discrete adjustment positions (open/shut). It would be an obvious design option to replace the check valve of D4 with a digital valve. Moreover, this modification is rendered obvious by the teachings of either D2 or D3. D2 teaches the use of a digital valve pack to control accurately the nip pressure in a calender roll of a paper machine (page 1, lines 5 to 9 and page 2, lines 25 to 34). The digital valve pack controls hydraulic cylinders which serve to open and close the nip during a path interruption (page 18, lines 9 to 11). D3 also teaches the use of digital valves in paper machines.

Should the Board consider that D4 fails to disclose the feature of claim 1 that the increase in nip pressure as the fault exits the roll nip is the result of a "feed-forward adjustment instruction", this feature would not render the claimed subject-matter inventive. It follows from decision T 850/06 that, in developing an automated process from a known manual process, apart from simply automating the individual steps of the manual process, the skilled person would also incorporate the facilities that automation offers for the monitoring, control and regulation of the individual process steps, provided they fall within the definition of common technical skill. When running the paper calender of D4 at high speed to increase productivity, and seeking to automate and optimise nip pressure control in the event of web faults, the skilled person would replace the check valve of D4 with a digital valve and would inevitably modify the valve control system of D4, so that the valve is triggered by feedforward adjustment instructions for opening and closing the nip. In fact, it is generally known in the art that digital valves must be controlled by feedforward instructions.

The same arguments apply, mutatis mutandis, to the subject-matter of apparatus claim 11.

Respondent's case:

The subject-matter of claim 1 involves an inventive step in view of the teachings of D4, D2 and D3.

Contrary to the appellant's view, D4 fails to disclose the method step of estimating or calculating the arrival time of the web fault at a nip and the dwell time for the fault in this nip, as required in claim 1. Besides, D4 is also silent about the use of a digital valve for controlling the nip pressure, as required in claim 1. Moreover, use of a feedforward adjustment instruction is not disclosed in D4. It only mentions the use of a sensing means being in contact with the web for detecting that the fault has left a nip and then triggering the opening of the check valve to increase again the nip pressure and re-close the nip (column 2, lines 49 to 54, column 4, lines 61 to 69, claim 4). In the calender with three rolls shown in figure 1 of D4, the nip pressure of the bottom and top nips is reduced when a web fault approaches the bottom nip which is arranged upstream of the top nip, such that the two nips are opened to prevent damage to the web or the rolls (column 4, lines 7 to 16 and lines 33 to 50). Once the operator or the sensing means has detected that the fault has left the two nips, they are again closed either by hand or by an automatic control means (column 2, lines 49 to 54, claim 4). A sensing means is used solely to trigger the closing of a nip after the fault has left it, not to trigger its opening. Thus, D4 discloses a simple feedback control of the nip pressure, but not a feedforward control as required in claim 1. In particular, D4 neither discloses the step of transmitting a feedforward adjustment instruction to the check valve controlling the nip pressure, nor the step of reducing and increasing again the nip pressure through the valve in accordance with the feedforward adjustment instruction, as required in claim 1.

To sum up, D4 does not disclose the method steps of:

- estimating or calculating an arrival time of the fault at a nip and a dwell time for the fault in the nip,

- transmitting from a control system to at least one digital valve unit, used for controlling a nip pressure of the nip, a feedforward adjustment instruction for dealing with the fault in the nip, and

- reducing as per feedforward adjustment instruction, through the intermediary of at least one digital valve, the nip pressure at the latest when the fault arrives at the nip and increasing the nip pressure as the fault exits the nip.

These distinguishing features are neither disclosed nor suggested in the other cited documents. D2 discloses digital valves using a feedback adjustment instruction (page 19, lines 12 to 24 and lines 31 to 34, claim 1) and thus provides no motivation to use a digital valve using a feedforward adjustment instruction. D3 is solely dealing with a servo valve in which an internal feedback control and communication are realised in a digital way. Thus, D3 does not show provision of digital valves or use of any digital valves in the sense of claim 1. Therefore, any feature combination of D4 with D2 and D4 with D3 cannot render the subject-matter of claim 1 or 11 obvious in any way.

Reasons for the Decision

1. Consideration of D7 and D8 in the proceedings

1.1 The opposition division already decided not to admit D7 and D8 into the proceedings, using its discretionary power under Article 114(2) EPC.

1.2 It is not the function of the Board to review all the facts and circumstances of the case as if it were in the place of the opposition division, in order to decide whether or not it would have exercised such discretion in the same way. Rather, the Board must confine its review as to whether the opposition division has exercised its discretion according to the wrong principles, without taking into account the right principles, or in an unreasonable way (see Case Law of the Boards of Appeal, 8th Edition, 2016, IV.E.3.6).

1.3 The opposition division decided not to admit D7 and D8 because they were prima facie not relevant with regard to the objection of lack of inventive step. The relevant reasons are set out in point 2.2.2.4 of the contested decision. Hence, the opposition division has correctly exercised its discretionary power under Article 114(2) EPC.

1.4 Consequently, exercising its discretion under Article 12(4) RPBA, the Board decides to disregard D7 and D8.

2. Admission of D9 in the proceedings

2.1 The appellant filed D9 in response to the reply of the respondent to the grounds of appeal.

2.2 The Board shares the respondent's view that the filing of D9 is belated since the factual situation of the case had not changed during the appeal proceedings and that the content of D9 is prima facie no more relevant than that of D2 or D3. In particular, D9 does not address the problem of controlling the nip pressure of a roll nip during the manufacture of a material web, let alone the problem of controlling the nip pressure to reduce down time in the event of a web fault.

2.3 For these reasons, exercising its discretion under Article 13(1) RPBA, the Board decides not to admit D9 into the proceedings.

3. Main request - Inventive step

3.1 The parties agree that D4 forms a relevant starting point for the assessment of inventive step. The Board shares this view.

3.2 D4 discloses a calender with at least three rolls for calendering a material web, in particular a textile web, having transverse or widthwise thickenings, such as seams (column 1, lines 1 to 6; figures 1 and 2, depicting embodiments of the calender, are reproduced below). These thickenings form web faults in the sense of claim 1. The material web to be calendered is passed through nips formed by a top roll, a bottom roll and one or more intermediate roll(s) fitted between the top and bottom rolls, wherein the rolls are arranged to form a substantially vertical stack of rolls. The bottom and intermediate rolls can move in a vertical direction and are pressurised in vertical direction by means of a bottom hydraulic cylinder.

The calender of D4 comprises means for reducing the pressure within the nips before a web fault passes therethrough to open them and thereby prevent damage to the web and the rolls. The calender also comprises means for increasing again the pressure within each roll nip after the web fault has passed it to close the nip and thereby resume calendering.

FORMULA/TABLE/GRAPHIC

D4 aims at reducing the length of uncalendered web to a minimum (column 1, lines 35 to 41; column 3, lines 29 to 33). It teaches that, when a web fault approaches the bottom nip of the calender, all nips are simultaneously opened by cutting off the connection between the bottom hydraulic cylinder and its fluid source, either by hand or automatically by a control device (column 3, lines 3 to 7; column 4, lines 33 to 50 and figure 1; column 5, lines 40 to 46 and figure 2). After the fault has left a nip, it is closed again by increasing the nip pressure, either by hand or automatically by the control device, whereby the latter uses a sensor for detecting the fault's exit from the nip and a check valve (column 2, lines 49 to 54; column 4, lines 61 and 69 and figure 1, check valve 14; column 5, line 46 to column 6, line 2 and figure 2, check valves 34 to 37; claim 4).

3.3 It is in dispute among the parties whether or not D4 discloses the following steps of method claim 1:

(a) "estimating or calculating an arrival time of the fault at a roll nip and a dwell time for the fault in the roll nip";

(b) "transmitting from a control system to at least one digital valve unit, used for controlling a nip pressure of the roll nip, a feed-forward adjustment instruction for dealing with a fault in the roll nip of a calender"; and

(c) "reducing as per adjustment instruction, through the intermediary of at least one digital valve, a nip pressure of the roll nip at the latest when the fault arrives at the roll nip and increasing the nip pressure as the fault exits the roll nip".

3.4 The Board shares the respondent's view that these features cannot be derived from D4.

3.5 Method step (a)

3.5.1 It cannot be derived from D4 that the arrival time of a web fault at a roll nip as well as the dwell time of this fault in this nip are estimated, let alone calculated. The Board is not persuaded by the appellant's argument that, when in the method of D4 the nips are manually opened and re-closed by an operator, he must inevitably estimate the arrival and dwell times of the web fault in one of the nips. In fact, it may well be that the operator opens all nips at a time when he sees that a web fault approaches the bottom nip, and that he closes again each nip only after he sees that the fault has passed it. This does not imply that the operator needs to estimate the arrival time of the fault, or its dwell time, while operating the calender. When in the calender of D4 the nips are automatically opened and re-closed by a control device instead of manually, the web fault approaching the bottom nip and its exit from each nip are detected by sensors. This too does not indicate that the actual arrival and dwell times of the web fault in the nips have been estimated or calculated.

3.5.2 The appellant argues that a skilled person, using its common general knowledge in the high technological field of paper technology, would immediately recognise that feature (a) is at least implicitly disclosed in D4. In particular, the skilled person would recognise that a web fault passes the calender of D4 in less than a second and thus the arrival and dwell times must be estimated to prevent any damage to the web and the rolls.

However, D4 is silent with respect to the running speed of the calender disclosed therein and it cannot be derived clearly from D4 that the web fault passes through the calender in less than a second. D4 is concerned with a calender for webs, in particular textile or fabric webs (column 1, line 5 and column 4, line 29), and it is generally known that the running speed of such a textile calender at that time did not exceed a few dozen meters per minute. This is the reason why D4 teaches that the calender nips can be precisely opened and again closed by hand in the event of web faults. However, it follows from D3 that, in 2004, paper machines produced more than 2000 meters of paper per minute (first column, second paragraph). There is no proof on file or the possibility of inference from any of the documents that the calender of D4 dating from 1959 was meant or capable to be adapted to be run at such a high speed, especially since it provides for manual control.

3.6 Method steps (b) and (c)

3.6.1 A "feed-forward adjustment instruction" as required in steps (b) and (c) means an adjustment instruction transmitted in advance of the actual arrival/exit of the fault in, and from, the nip.

3.6.2 In figures 1 and 2 of D4, after detection of a web fault approaching the bottom nip, a "feed-forward adjustment instruction" is automatically transmitted to the control device 21 to cut off the hydraulic fluid supply to the bottom hydraulic cylinder 8 and thus open all nips simultaneously; this is in advance of the actual arrival of the web fault in the bottom nip. It cannot be derived from D4, however, that the nip pressure is reduced "as per adjustment instruction through the intermediary of at least one digital valve", as required in method step (b).

After the sensor has detected that the web fault has left the bottom nip, the control device 21 re-opens the hydraulic fluid supply to the bottom hydraulic cylinder 8 to close again the bottom nip (column 2, lines 49 to 54). After a sensor has detected that the web fault has left any subsequent nip, the control device 21 automatically triggers a check valve to increase again the pressure at the nip and thus close it (check valve 14 in figure 1; check valves 34 to 37 in figure 2). The increase of the nip pressure as the fault exits any nip following the bottom nip is thus achieved by feedback control of a check valve, and not by way of a "feed-forward adjustment instruction" as required by method step (c).

3.7 Thus, the Board comes to the conclusion that the method defined in claim 1 differs from that disclosed in D4 in that it comprises the afore mentioned steps (a) to (c).

3.8 These distinguishing features further increase the rapidity and accuracy of the nip pressure control in response to web faults. Starting from D4, the technical problem objectively solved by these features can be formulated as how to further reduce the length of uncalendered web in case of web faults.

3.9 The claimed solution to this problem is not part of the skilled person's common general knowledge and is neither disclosed nor suggested in the cited prior art documents D2 and D3. Even though documents D2 and D3 teach the use of digital valves in paper machines, they provide no pointer to all the distinguishing features. In particular, D2 teaches to use a digital valve pack for accurately controlling the opening and closing of the roll nip of a paper calender (page 2, lines 4 to 32). However, D2 teaches that the digital valve pack receives feedback adjustment instructions (page 19, lines 12 to 24 and lines 31 to 34, claim 1), and thus the teaching of D2 leads away from using a digital valve operated by a feedforward adjustment instruction for re-closing the nip, as required in step (c). D3 discloses the use of digital high-response valves for the closed-loop, i.e. feedback, hydraulic control of high-speed paper machines. Thus, D3 also teaches away from providing a digital valve using feedforward control.

3.10 The appellant argues that, when running the calender of D4 at a high speed of, say, 2000 m/min to increase productivity and seeking to automate and optimise nip pressure control in the event of web faults, the skilled person would consider using a digital valve and he would inevitably control the digital valve in the claimed manner.

It is agreed that the mere automation of the manual control disclosed in D4 could not be considered inventive (see e.g. decision T 850/06 as referred to by the appellant) and is, in fact, already disclosed in D4 (see control system 21 in figures 1 and 2). However, starting from the fabric calender disclosed in figure 1 or 2 of D4, there is no motivation to run it at high speed, let alone to use a digital valve with feedforward control for opening and re-closing a roll nip in the event of a web fault. In particular, in the event that the skilled person were to replace the check valve 14 in figure 1 of D4 with a digital valve, there is no evidence indicating that he would also modify the control system in such a manner that the valve is triggered to increase again the nip pressure by feedforward control using estimation or calculation of the time needed for the fault to exit the nip (as required by method steps (a) to (c)), instead of feedback control using detection that the fault has effectively exited from the nip (as disclosed in D4). Contrary to the appellant's assertion, digital valves are not necessarily controlled by feedforward adjustment instructions. For instance, as indicated above, D2 and D3 disclose digital valves using feedback control.

3.11 Thus, when starting from D4, the subject-matter of claim 1 involves an inventive step within the meaning of Article 56 EPC.

3.12 The above reasoning applies mutatis mutandis to the subject-matter of independent apparatus claim 11 as well as that of the dependent claims.

4. In conclusion, the cited ground for opposition of lack of inventive step does not prejudice the maintenance of the patent as granted.

5. Under these circumstances, there is no need to consider the auxiliary requests of the respondent.

6. Board's communication pursuant to Article 15(1) RPBA

6.1 In the oral proceedings, the appellant argued that the Board's communication in preparation of the oral proceedings (Article 15(1) RPBA) was legally defective for two reasons. Firstly, it did not include a definition of the skilled person and his common general knowledge. Secondly, it did not mention the technical problem objectively solved by the invention.

6.2 It is firstly noted that the boards have discretion as to whether or not to send such a communication. In any event the purpose of a board communication pursuant to Article 15(1) RPBA is to prepare the oral proceedings. It is by no means binding nor is it expected to be exhaustive; it merely needs to cite the main points in dispute or, more generally, in need of discussion.

6.3 In the present case, the Board decided to issue a communication in order to streamline the procedure and indeed to identify the main points of dispute in advance of the oral proceedings. Since the parties have neither disputed the common general knowledge of the skilled person nor the formulation of the problem to be solved, the Board saw no need to address these points in its communication.

7. Conduct of proceedings

7.1 In the last written submission before the oral proceedings and at the oral proceedings, the appellant addressed disparaging remarks to the Board, indicating an incorrect perception of the work within a board.

7.2 A board of appeal is a body with a specific composition as foreseen in the relevant legal texts, which acts as a collective body and not as individual members. The EPC and the RPBA provides for an opinion/communication/decision of a board, but nowhere is it foreseen that the personal opinion of (one of) its members be presented as those of the board. In any event, personal criticism is not an appropriate reaction. The good and fair conduct of a case requires that all parties and the tribunal remain objective and argue solely on the basis of the facts of the case.

Order

For these reasons it is decided that:

The appeal is dismissed.