Fujikura Ltd & Anor v Sterlite Technologies Limited

Optical fibre cables (“OFCs”) are an essential part of modern communications. Prior to use they are commonly installed into ducts using a high-pressure stream of air to blow the cable along the duct. This is the trial of a patent action concerning the design of the outer sheath of an OFC which is said to improve the air-blown installation process.

The Claimants, “Fujikura”, are the proprietor and exclusive licensee respectively of EP 3796060 B1 (“the Patent”) with a priority date of 11 October 2018.

The Defendant, “Sterlite”, is accused of infringing the Patent by the sale of its Celesta range of OFCs. It denies infringement and counterclaims for invalidity on multiple grounds.

The Claimants were represented by Geoffrey Pritchard KC and Richard Darby, instructed by Dehns. The Defendant was represented by Brian Nicholson KC and Adam Gamsa, instructed by Allen Overy Shearman Sterling LLP. I am grateful to the parties and all their legal representatives for the efficient conduct of the trial.

I am also pleased to record that in accordance with the Patents Court Guide, both parties’ junior counsel conducted cross-examination and delivered oral closing submissions during the trial. This involved me granting permission to the parties to utilise both Counsel to cross-examine the main experts. I considered that it was appropriate to allow this given the clear difference in subject matter between the topics addressed by each and the fact that this did not lengthen the cross-examination of the witnesses overall.

Each side called two expert witnesses in the fields of (a) fibre optic cable design, manufacture and installation (the “Cable Expert”) and (b) operating 3D Optical Surface Profilers (the “Measurement Expert”).

Fujikura’s Cable Expert was H. Paul Debban Junior. Mr Debban has some 40 years of experience in the design and manufacture of OFCs, and particularly fibre optic ribbon cables, working at OFS and its predecessor companies.

Sterlite’s Cable Expert was Ralph Sutehall. Mr Sutehall is a retired former Principal Installation & Applications Engineer within the Communications Division of Prysmian Group, another manufacturer of OFCs. His experience, which spanned some 50 years, was more in the installation of cables than in their design, and he accepted that Mr Debban had greater experience in the latter. However, as noted below, the notional skilled team required expertise in both disciplines.

Fujikura’s Measurement Expert was Casey Michelle Jarvis. Ms Jarvis is an executive director at MVA Scientific Consultants, a microscopy laboratory.

Sterlite’s Measurement Expert was Mark Malburg. Dr Malburg is President of Digital Metrology Solutions, Inc, which provides metrology-related training, consulting and software.

I am grateful to all four experts for the assistance they gave the court. Both parties accepted that the Measurement Experts were fair and impartial and I agree with this assessment.

Criticism was directed by Sterlite towards Mr Debban for sometimes giving long answers, and by Fujikura towards Mr Sutehall for sometimes not answering the question. They were both cross examined for a day and a half and it is not surprising that on occasion their concentration wandered. Further, as both parties ultimately relied on the opposing expert as supporting their obviousness case in oral evidence, the criticisms were ultimately somewhat muted.

It was clear that whilst both Cable Experts had ample industry experience, neither was accustomed to the artificial scenario of giving expert evidence in court as part of patent proceedings. For example, there were times when Mr Sutehall struggled to deal with the notion of what was properly common general knowledge. Nevertheless, I find that both Cable experts were doing their best to fulfil their duties and overall I am unable to place greater weight on one over the other. I comment on individual answers in context below.

The parties also tendered fact witnesses, but none were in the end called.

Fujikura’s first fact witness was Donald Shane Heinze. Mr Heinze is a Quality Engineering Test Technician at American Fujikura Ltd, a subsidiary of the First Claimant. Mr Heinze carried out the analysis for the Repeat Experiment relating to the allegation of infringement. Although Sterlite challenged the robustness of the experimental protocol it did not challenge Mr Heinze’s evidence.

Fujikura’s second fact witness was Eric Christopher Erb, a Test Engineer at AFL. His evidence explained the production of images depicting the analysis carried out by Mr Heinze, as well as the step of choosing the sample to be tested as part of the Repeat Experiment. Again, this evidence was ultimately not challenged by Sterlite.

Sterlite’s fact witness was Sarin Kumar. Dr Kumar is an employee of Sterlite and is Global Head of the Engineering & Technology Communication Cabling Solutions Business. His evidence introduced a Gas Leakage Study into the proceedings that was conducted by Sterlite after the Priority Date of the Patent.

Sterlite also submitted a Product and Process Description (“PPD”). It explains details about Sterlite’s Celesta Products and is signed by Vaibhav Khanna, who is the Global Head of Intellectual Property at Sterlite. It was not challenged by Fujikura. I return to this below.

There was no dispute as to this. The Patent is directed to a team skilled in the design of fibre optic cables, with knowledge of cable manufacture and installation by air blowing. There was some debate about precisely how the skill sets would be split between notional individuals on the team, but nothing turns on this. The parties agreed that the notional team would have around 3 years’ experience. It was not suggested that the notional skilled team had expertise in measurement. The Measurement Experts were called solely in relation to the Claimants’ experiments to demonstrate infringement.

In hearing from Mr Debban and Mr Sutehall, it struck me that both individuals had immense practical experience in this field, albeit coming from slightly different backgrounds (Mr Debban was a designer of cables and Mr Sutehall was an installer). Whilst I recognise that it would be a mistake to characterise the skilled person by reference to the individual experts appearing at trial, I cannot ignore the fact that this field is primarily a practical and empirical one. The skilled team would therefore be less concerned than in some other fields about theoretical hypothesis and more interested in practical, real-world results. This has a bearing on some of the issues that I will come onto, namely construction and the approach to obviousness.

Most of what follows (including the figures) is taken directly from the technical primer/agreed statement of common general knowledge (“CGK”).

Fibre optics are at the forefront of modern networks. Their applications include fibre to the home networks, data centres and enterprise and campus networks. As demand has increased, there has been a consistent trend by optical fibre cable manufacturers to reduce the size of optical fibre cables and increase the fibre density.

Aspects of the construction and performance of optical fibres, fibre units, optical fibre cables and other elements of optical fibre networks are standardised. In the UK and Europe the relevant standards are published by the International Electrotechnical Commission (the “IEC”) and the International Telecommunication Union (the “ITU”), both of which are located in Switzerland.

There is a choice of where and how to install the optical fibre cables in indoor and outdoor networks. One of these is duct installation. Plastic or metal tubes, called ducts, are installed underground or indoors along the route. The cable is then either pulled or pushed into the duct, or blown into the duct using a cable blowing machine.

There are three sizes of ducts. The difference between sub-ducts and micro-ducts is size. Both are used in indoor and outdoor networks. The precise boundaries for the size of sub-ducts and micro-ducts as a matter of CGK was not agreed by the parties. I return to this in the context of the Patent below:

Sub-ducts and micro-ducts (generally referred to as “ducting”) were available at the priority date with different inner surface designs. These designs were intended to aid in the installation process by reducing friction between the cable being installed and the inner surface of the ducting. For example, prior to the priority date, sub-ducts with and without ribs on the inner surface were available (see Figure 1).

Telecommunications grade optical fibre is made from synthetic quartz (SiO2) with the core or centre of the fibre doped with chemicals to give a higher index of refraction than the outer portion of the fibre (called the cladding). For single mode fibres, the core of the fibre has a diameter of approximately 9 µm. The diameter of the cladding over the glass fibre itself is 125 µm. For multimode fibres, the core has a diameter of approximately 50 µm, and the diameter of the cladding over the glass fibre itself is 125 µm.

The difference between the high index of refraction in the centre of the optical fibre and the index of refraction of the cladding results in total internal reflection of the light transmitted into the fibre, allowing the optical signal to be transmitted for long distances.

Optical fibre designs have been standardized globally to ensure that fibres from different vendors can inter-operate. Optical fibres can be classified as either ‘single mode’ or ‘multimode’. A single mode optical fibre allows for the light to take a single path within the core, whereas multimode optical fibre allows the light to make multiple paths within the core. Single-mode and multi-mode optical fibres are used in the same optical fibre cable designs.

Optical fibres may be grouped within cables in different arrangements. The below groupings were used in both indoor and outdoor cables.

The most common grouping method is to house the fibres ‘loosely’ in a tube. If there are multiple tubes within a cable, the tubes are coloured or otherwise marked for identification. Because these optical fibres are held loosely in a tube, it means that they have a small degree of movement within the tube, which helps in dealing with mechanical stresses.

Fibres may be additionally grouped within a cable by fabricating them into optical ribbons. Traditional ribbons are manufactured using 4 to 36, commonly 12, optical fibres arranged side by side and connected using a resin.

One of the most significant advantages of optical fibre ribbons is the ease and speed of splicing during the installation process. Optical fibre splicing is a method for fusing two different optical fibres – used, for instance, when extending or branching optical fibre cable runs or repairing damaged/severed optical fibres. This process involves cleaving the ends of two optical fibres, aligning them, and then fusing them using a fusion splicing machine. The fusion is achieved through an electric arc that effectively welds the fibres together, creating a seamless connection.

Optical fibres within ribbons can be spliced simultaneously using mass fusion splicing techniques, which can splice multiple optical fibres at once. Conversely, loose optical fibres need to be spliced individually – this is because they lack the fixed and ordered planar structure required by a mass fusion splicer. Therefore, ribbons greatly reduce the time and labour costs associated with optical fibre splicing.

Intermittently bonded ribbons (“IBRs”) are a development of the traditional optical ribbon originating in Japan. This optical ribbon design is similar to the traditional optical ribbon except the linear array of optical fibres is intermittently connected by matrix instead of the continuous connection used in traditional ribbons. The intermittent connection breaks the preferential bending of the ribbon structure and allows the use of design rules for a loose fibre cable, but the ribbons reassume their flat shape when released from a cable to support mass fusion splicing.

IBRs have contributed to the development of cables with smaller outer diameters (as compared to a cable utilising traditional optical ribbons with the same total number of optical fibres).

The concept of air blown fibre was developed by British Telecommunications (BT) in the 1980s. This technique comprised blowing an optical fibre (with a foamed polyethylene covering) into a pre-installed empty tube. In contrast to conventional cable installation, the use of this blown fibre reduced the time and effort required for installation.

Following this innovation, BT developed enhanced performance fibre units or “EPFUs”, which are low fibre count units that are encapsulated by acrylate and protected by a thin outer sheath (thinner than one used for a conventional optical fibre cable) or a resin coating that are optimised for air blown installations. Since that time, other companies developed their own versions of these products – collectively, they are generally referred to as ‘fibre units’. Fibre units typically have diameters of less than 2 mm.

Two different cable structures are commonly used in optical fibre cables in the UK and Europe:

In the manufacturing process of multi-loose tube (MLT) cables, groups of optical fibres or ribbons are loosely placed in coloured plastic tubes.

The tubes are helically or SZ stranded around a central strength member. SZ stranding (also known as reverse lay stranding) means that the tubes are twisted around the central strength member a predetermined number of turns in one direction (the “S” direction), then the tubes are rotated the same number of turns in the opposite direction (the “Z” direction).

The stranded tubes may be covered with additional strength members, protective tapes or armour, as required for the application. The cable is then finished with a polymeric outer jacket (also referred to as an outer sheath).

In central core cables, the fibres are located in the centre of a cable. The ribbons or fibres may be enclosed in a plastic tube or wrapped with a tape or surrounded by aramid. The strength members are located to the outside of the cable core (i.e., surrounding the cable core), for example as strength members embedded within the cable jacket. A central core ribbon cable with a layer of steel armour (green) and two steel strength members embedded in the jacket is illustrated in Figure 8.

Both cables have relative advantages and disadvantages in the context of air blown installations:

One of the sources of CGK was commercially available products from the major optical fibre cable manufacturers such as Corning, Prysmian, OFS, Furukawa, Fujikura/AFL and Sumitomo.

At the priority date, a typical optical fibre cable (both indoor and outdoor variations) would have commonly included the following components, in addition to optical fibres.

Strength members used in optical fibre cables are typically either:

Strength members are commonly found in MLT cables and Central core cables. In MLT cables, there is a single rigid strength member at the centre of the cable axis, and the loose tubes would be stranded around it. Additional flexible strength members may be wound over the core if required.

In Central core cables, there are typically multiple strength members, such as aramid fibres, surrounding the core or embedded within the outer sheath. The strength members increase the tensile strength (i.e., the maximum mechanical load that a material can withstand when being pulled before failure) of the cable and decrease the likelihood of buckling or kinks in the cable.

The optical performance and strength of optical fibres is reduced if they are exposed to humidity or water. Water freezing and expanding in cold conditions can damage the optical fibre core. To protect the optical fibres from water ingress, the cables typically contain water blocking materials.

Ripcords are located underneath or within the outer sheath to assist in opening the cable to safely access the optical fibres. These are yarns or threads of sufficient strength to tear through the jacket. Their location can be indicated by a protrusion, a groove, or a line on the cable sheath.

The cable core is covered by an outer sheath (alternatively called a cable jacket) to provide mechanical and environmental protection.

Cables that are directly buried may have a layer of metal armour to protect from rodents, shovels, or other mechanical hazards. Duct cables are protected from these by the duct system.

The sheath may also contain strength members as described earlier.

The outermost layer of the sheath is an extruded polymer coating. At the priority date, 2018, the most common optical fibre cable sheath materials for outdoor cables were high density (HDPE) or medium density (MDPE) grades of polyethylene.

In addition to the conventional smooth outer surface design, there were other, different optical fibre cable (and fibre unit) outer surface designs available in 2018. Two examples are set out below.

Here, the surface of the blown fibre unit has been modified to incorporate small glass spheres, which are sprayed onto the surface of the uncured outer acrylate layer. These glass spheres produce protrusions in the profile of the fibre unit, which enhance the blowing performance of the fibre unit.

The reflective sheath design is the result of manufacturing a thin cable jacket that closely followed the contours of the underlying stranded cable core (i.e., the stranded loose tubes in a loose tube cable). This had the advantages of reducing the cable diameter, and increasing the effectiveness of the viscous drag of the air on the cable together with reducing friction.

Part of the manufacturing process of optical fibres is the application of a cable sheath to the optical fibre cable being manufactured, which is done through a process known as extrusion coating.

This is a process where a protective polymer layer (e.g., high density polyethylene) is applied over the core of an optical fibre cable.

In the sheathing process, a machine called an extruder is used. This extruder contains a hopper containing a solid polymer material, which is fed from the hopper onto a screw in a heated barrel (thereby melting the polymer material). At the end of the extruder is a crosshead where the cable core is guided by an extrusion ‘tip’ into and through a die that shapes the molten sheath surrounding the cable core. The extruded cable then passes through a series of water baths to cool and solidify the polymer, forming a protective sheath.

In the case of smooth optical fibre cable sheaths, the exit profiles of the tip and die are circular since the desired outcome is a smooth surface. However, if a profiled or non- circular outer surface is required, the die will be cut as a negative of the idealised desired profile. In practice, the finished optical fibre cable will deviate from the die profile as a result of factors such as shrinkage due to the cooling of the sheath after extrusion.

There are a variety of methods that have been used in order to install cables into ducting, which generally includes pushing, pulling, and air blown installation.

Air blown installation of optical fibres was patented by BT in 1982. Initially, this technology involved installing air blown fibre units into ducting using high-speed airflow directed by an air blowing machine. Later developments made it possible to install optical fibre cables by blown installation.

An example of a cable blowing machine is Plumettaz’s MiniJet (PO1), which was used to blow optical fibre cables into both ducts and micro-ducts.

Such a cable blowing machine commonly comprises four elements:

In relation to the air inlet unit, there are airtight seals around the ducting on one side of the unit (the ‘duct seal’) and around the circumference of the cable on the other side of the unit (the ‘cable seal’) (see Figure 12).

Air will be fed into the air blowing machine via the air inlet (located between the duct seal and the cable seal), thus pressurising the air inlet unit. Once inside the air inlet unit, the only direction that the air will be able to flow is down the open ducting, taking the optical fibre cable with it.

Various optical fibre cable test procedures are standardised by the IEC. For example, Method E24 (installation test for micro-duct cabling) for air blown installations is found in IEC 60794-1-21:2015.

Furthermore, the air blown installation test track specified in Method E24 of IEC 60794- 1-21:2015 is reproduced below in Figure 13:

One example of non-cable design factor that affects the air blowing installation performance of optical fibre cables is the design of ducting.

At the priority date, it was common general knowledge that there was ducting designed to reduce friction within the bore which, in turn, helps improve installation performance. These designs included having:

The following two cable design factors affected the air blowing installation performance of optical fibre cables.

It was not uncommon for an optical fibre cable to become ‘oval’ (i.e., not circular) as a result of its construction. Depending on the degree of ovality, this can be detrimental to air blown installation if the major axis of the oval cable comes into contact with the walls of the ducting, causing friction.

It was generally accepted that in the context of air blowing installations, the fill factor (the ratio between the cross-sectional area of the optical fibre cable and the cross-sectional area of the inner bore of the ducting) should not be greater than 70% (0.7), whilst the ideal fill factor is between about 40% (0.4) to 60% (0.6).

Cable stiffness can impact air blown installations in two ways:

By the close of the evidence there was very little disputed CGK remaining. As noted above, the cross-examination of Mr Sutehall indicated that he had an overly broad understanding of what might be CGK, at one stage suggesting that all 100 papers delivered at the IWCS conference would have been CGK. I have borne this in mind in the assessment which follows.

Having heard the evidence I am satisfied that the relevant market for OFC design and installation was an international one. Very few of the companies active in the market place were designing and selling product for one market only (i.e. the UK). People in the field knew of international standards and products were marketed as complying with multiples of these.

The main dispute which I am left to determine was about the understanding of micro-duct as a matter of CGK This is relevant to the construction of the Patent. By way of background, the relationship between the external and internal diameter of the ducts in question should be recorded. This was set out as part of the Agreed CGK as follows:

The Defendant argued that micro-duct would have been understood as limited to ducts of less than 16mm outer/13mm inner diameter because this was the UK understanding based on what was written in a 2014 version of IEC Standard 60794-5: “A microduct suitable for installation of microduct cables is a small, flexible, lightweight tube with an outer diameter typically less than 16mm”. Mr Sutehall relied in particular on the use of the word “typically” to support his view.

The Claimants’ position is that by the priority date the CGK definition of micro-duct had expanded to include internal diameters of up to 16mm. They relied on a wider range of later materials including US Standards, international brochures and sections of the 2016 and 2018 FTTH (fibre to the home) Handbook published by the FTTH Council Europe, an industry association. Mr Debban also pointed out various 16mm inner diameter commercial products sold as micro-ducts at the priority date, including a 2016 brochure of Radius products sold in the UK. Mr Sutehall explained that Radius were one of the best duct manufacturers in the world, based in Northern Ireland.

Mr Sutehall acknowledged that the 2016 FTTH handbook indicated that micro-ducts should be defined not just by size, but also by looking at the way in which the duct was being used. Indeed, he had relied on the 2016 handbook as supporting his evidence about “typical” diameters in the IEC standard. The 2018 FTTH handbook was cited by Mr Debban in his reply report and referred to a 20mm outer/16mm inner diameter micro-duct. Mr Sutehall did not comment on this or the Radius document in his subsequent third report. In cross-examination he accepted that if he was being fair he should have commented on the Radius brochure and acknowledged that the 2018 FTTH handbook was the current edition at the priority date, but he emphasised that it was only a guideline and the earlier IEC document was a standard. However, he also accepted that the Radius brochure was consistent with Mr Debban’s evidence that 16mm internal diameter duct was an acknowledged micro-duct size in the UK at the priority date.

Having listened to the cross-examination on this point and looked at the source materials, I am satisfied on the evidence that micro-ducts were not limited in size as a matter of CGK in the way that Mr Sutehall had suggested. They included ducts with an inner diameter of 16mm. Further, the skilled person would have no real problem identifying what was and what was not a micro-duct at the priority date.

The remaining issues of CGK which it had been suggested were live prior to the start of the trial but which turned out on the evidence not in fact to be disputed are picked up in context below.

The Patent-in-suit in this case is a curious beast. That is because it discloses a number of different alleged improvements in the design of OFCs. Yet only one of these is the subject of the claims. I was not told if there were other granted patents in the same family which sought to claim the other design features, no doubt because this is irrelevant to the considerations before me. I will therefore focus on the features in the specification which are relevant to the claims. I refer to some of the other features in the context of the breadth of claim allegations.

The title of the Patent and [0001] make clear that the invention relates to OFCs. After summarising some of the background art the technical problem is set out in [0009] (repeated under “Advantageous Effects of the Invention” in [0012]) as being the provision of “an optical fiber cable which is advantageous in terms of air-blowing characteristics, diameter reduction, and transmission loss while increasing the strength of the sheath”. However, thus far the air-blowing problem has been expressed in terms of a lack of rigidity of the cable, and not the problem which is the subject of the solution in the claims.

The description continues by introducing the figures, including Figure 1A which is described in [0014]-[0015] as follows:

[0014] As illustrated in Fig. 1A, the optical fiber cable 1 includes a sheath 10, a core 20 housed in the sheath 10, and a plurality of tensile strength members 30 embedded in the sheath 10.

[0015]The core 20 has a plurality of optical fiber units 21, and a wrapping tube 22 that wraps these optical fiber units 21. Each of the optical fiber units 21 has a plurality of optical fibers 21a and a binding material 21b that binds the optical fibers 21a.

A detailed description of the structure of an intermittently-adhered OFC follows, much of which is reflected in the CGK section I have set out above.

At [0026] the description turns to the profile of the recesses (12) and protrusions (11) shown in Figure 1A. As can be seen from the figure, these recesses and protrusions are not even and form what the specification refers to as an “uneven shape”. In [0028] the “shoulders” of the recesses at 12a are said to be “formed in a curved surface shape that is radially inward convex”. In [0029] the bottom surface 12b is described as having an arc shape centered on the central axis O, but could also have other shapes including being formed by a straight line between the two connecting portions 12a.

[0034] states as follows:

[0034] The radius of curvature of the outer circumferential surface of the protrusion 11 may be smaller than the radius of the sheath 10 (the radius of the optical fiber cable 1). According to this configuration, the contact area between the protrusion 11 and the micro-duct (details will be described later) becomes smaller. Therefore, the workability when the optical fiber cable 1 is inserted into the micro-duct can be improved. The "radius of the sheath 10" is the maximum value of the distance between the outer circumferential surface of the protrusion 11 and the central axis O. When the maximum value is different for each protrusion 11, the average value of each maximum value is defined as the "radius of the sheath 10".

In other words, some of the protrusions may be smaller than others which may reduce friction and improve “workability”. If this is the case the radius of the sheath overall is calculated by averaging the radii of the protrusions.

The specification then descends into detail about one of the promised advantages that is not the subject of the claims – Maximum Compressive Stress, or stiffness. [0045] explains by reference to data set out in Table 1 that the examples having a compressive stress of less than 11.6N/mm2 do not perform well in the air-blowing tests because they buckle. Table 1 also shows that compressive stress is not necessarily related to the diameter of the cable.

Although it is not necessary to discuss this section further, it does cross refer to Figure 3 which is as follows:

This illustrates the installation process. Note the pump P, the micro-duct D and the seal S attached to the end of the micro-duct D which forms around the OFC and allows an air layer to be generated between the OFC and the duct walls.

At [0050]-[0065] there is another portion of the disclosure which is not relevant to the claims, this time concerning “Wrap Rate”. Similarly there are sections at [0073]-[0082] “Twisted Shape of Sheath”, at [0083]-[0090] “Material of Tensile Strength Member”, at [0091] “Number of Tensile Strength members for protrusions”, at [0092]-[0101] “Set Twist Angle”, at [0102]-[0108] “Low Friction Material” and at [0109]-[0121] “Ripcord”.

There are just seven paragraphs dealing with the subject matter of the claims under the heading “Cross-sectional Area of Recesses” at [0066]-[0072].

The teaching of the specification is that the formation of the friction-reducing air layer between the OFC and its micro-duct (as a result of the OFC’s protrusions and recesses) can be optimised by reference to the cross-sectional area of the recesses ([0066]).

To illustrate this, cables of different recess cross-sectional areas as depicted in Figure 6 are prepared and tested in the test-track of Figure 4. The cross-sectional area of the recesses A is defined as the difference between the cross-sectional area of a virtual optical fiber cable having a closed curve L as the outer circumferential surface and the example being tested ([0067]). It is depicted by the thin shaded area A within each recess in Fig 6 as follows:

[0068] presents the results with the following explanation:

[0068] The closed curve L is usually circular with the central axis O as the center. However, due to the deformation of the optical fiber cable, the closed curve L may have an elliptical shape.

[0061] had previously defined results for a different test as good (OK) or not good (NG). The reader would not understand the NG results to mean that the cable could not be blown at all, just that they did not meet the undefined standard used by the patentee.

The specification goes on to explain the results in the following terms:

[0069] As shown in Table 3, the results of the air-blowing test are not good, in Test Example (3-1) having a cross-sectional area A of the recesses of 5.2 mm². The reason for this is that when the cross-sectional area A of the recesses is significantly large, the sealability between the seal S and the optical fiber cable is deteriorated, and the backflow of air from the inside of the micro-duct D is likely to occur. When the amount of air flowing back from the inside of the microduct D is large, the amount of air intervening between the inner surface of the micro-duct D and the optical fiber cable is reduced, and friction increases. It is considered that this friction made it difficult for the force to be transmitted from the upstream side to the downstream side of the optical fiber cable, and the progress of the optical fiber cable stopped.

So when the cross-sectional area of the recesses is too large, air leaks back around the seal S shown in Figure 3, and the air layer is insufficient to overcome friction between the OFC and the micro-duct.

Further, as [0070] goes on to explain, there is also a stage where the recesses are too small and the seal is too tight around the OFC, so the air layer is again insufficient, not as a result of leaking but because this time not enough air can enter the duct in the first place. The Patent teaches that a recess area somewhere in the middle is desirable, namely between 1.3mm² and 4.8mm²:

[0070] In contrast, in Test Examples (3-2 to 3-5) in which the cross-sectional area A of the recesses is 1.3 mm² or more and 4.8 mm² or less, the results of the air-blowing test is good. This is because the cross-sectional area A of the recesses is sufficiently small, the sealability between the seal S and the optical fiber cable is good, and the backflow of air from the inside of the micro-duct D is suppressed. That is, it is considered that the friction is reduced by the sufficient air intervening between the inner surface of the micro-duct D and the optical fiber cable, and the force can be transmitted from the upstream side to the downstream side of the optical fiber cable.

The further example of a totally smooth OFC is given in [0071] with zero cross sectional area. This is said not to blow far enough because the friction is too large.

Accordingly, [0072] teaches as follows:

[0072] From the above results, it is preferable that the cross-sectional area A of the recesses is in the range of 1.3 mm² or more and 4.8 mm² or less. With this configuration, the sealability between the seal S and the optical fiber cable can be ensured, and the workability of air-blowing can be improved.

As noted above, the claims are restricted to the alleged invention in the Cross-sectional Area of Recesses portion of the specification.

The Claimants have applied to amend the claims unconditionally to bring them into line with the form upheld by the Opposition Division of the EPO. Claim 1 as proposed to be amended (with the addition underlined) reads as follows:

1.

An optical fiber cable (1)for insertion into a micro-duct by air-blowing, comprising:

a sheath (10); and

a core (20) which is housed in the sheath (10) and which has an intermittently-adhered optical fiber ribbon (21) including a plurality of optical fibers (21a) and a plurality of adhesive portions (21c) for intermittently adhering the plurality of optical fibers (21a) in a longitudinal direction,

wherein recesses (12) and protrusions (11) are formed so as to be disposed alternately in a circumferential direction on an outer circumferential surface of the sheath (10), and

the recesses (12) each include two connecting portions (12a) respectively connected to radial inner ends of two adjacent protrusions (11), and a bottom surface (12b) positioned between the two connecting portions (12a), the optical fiber cable characterized in that:

in a transverse cross-sectional view, a cross-sectional area (A) of the recesses (12) is within a range of 1.3 mm² or more and 4.8 mm² or less,

where the cross-sectional area (A) of the recesses (12), which is a cross-sectional area of a space defined by a closed curve (L) tangent to radial outer ends of the plurality of protrusions (11) and all the recesses (12), is a difference in a cross-sectional area of the optical fiber cable (1) with respect to a cross-sectional area of a virtual optical fiber cable having the closed curve (L) that contacts with each radial outer end of the protrusions (11) as an outer circumferential surface,

and the closed curve (L) is a circular shape or an elliptical shape of which center is a central axis (O) of the optical fiber cable (1).

There are a number of immediate points to note about the claim.

First, as in the specification, the recess cross-sectional areas are given in absolute terms, regardless of cable size - within a range of 1.3 mm² or more and 4.8 mm² or less.

Without any limitation to cable size, the claim might well suffer from insufficiency problems. For instance, it would be difficult to imagine that a cable of diameter 100cm would have improved blowability with a recess area of just 1.3 mm².

The proposed amendment addresses this by imposing a limit to the cable size – it has to be something which is “suitable for” insertion into a micro-duct. However, as flagged above, the parties were not agreed as to what was the CGK definition of micro-duct. This was relied on for a non-infringement point and also led to lack of clarity objections to the proposed amendment on the part of the Defendant. I return to these below.

Finally, there is the issue of how to calculate the cross-sectional area of the recesses. The claim explains that this is to be done by calculating the area of a closed curve L tangential to the outer ends of the protrusions and subtracting the area of the OFC from that. Further, the L could be either circular or elliptical in shape with its centre O as the central axis of the OFC.

There was no real dispute as to this as a matter of the construction of the words. The dispute was instead about how to define the comparator in practical terms and whether the claim was so ambiguous that it could not be infringed. Much of this turned on whether the claim is limited to perfectly circular and elliptical cables and, if not, whether the claim is therefore bad for uncertainty insufficiency. There was also a dispute as to whether the Claimants had actually carried the required test out sufficiently in the course of their experiments to prove that the 432F sample Celesta cable tested infringed, and whether this could be extended to other cables in the Celesta range. It is easier to deal with all of this under the heading of infringement, to which I return below.

There was no dispute between the parties as to the approach I should take to construing the claims. The objection to the scope and/or clarity of micro-duct turns on my assessment of the facts/CGK. The significance of the point is that if the Defendant is right and 16mm internal diameter is not a micro-duct, then that would exclude from infringement the Defendant’s 432F cable.

The Defendant went further to contend that the limitation on the size of micro-duct needs to be read alongside the CGK that the “fill factor” of the cable within the duct should not be greater than 70%. Taken together, the biggest cable covered by the claim (i.e. which is “[suitable] for insertion into a micro-duct by air blowing”) would therefore have a diameter of 10.88mm. The Claimants contended that fill factors could be higher, up to 80%, because these still allowed cables to be blown, albeit over shorter distances.

I have dealt with the CGK above. As a matter of CGK micro-ducts could be 16mm internal diameter. To be added to this is the fact that the Patent at Table 8 refers to 12mm outer diameter cables being inserted into micro-ducts. These would only fit down a 16mm internal diameter duct.

As a back-up argument the Defendant contended that the claim was uncertain and the proposed amendment should be refused for lack of clarity. It referred to an AFL/Duraline brochure produced for the first time in Mr Debban’s XX bundle which suggested that micro-ducts could be as large as 20mm internal/27mm external diameter. I reject the contention that this late-adduced brochure was CGK and in any event I have found that the skilled person would know a micro-duct when they saw one at the priority date. There is no lack of clarity.

For reasons which will become apparent when I turn to the prior art, this is one of those cases where it is appropriate to seek to define the technical contribution of the claim, and not just construe it. The technical contribution is also relevant to the breadth of claim insufficiency attacks. In the present case mere construction of the claim does no more than define the cables falling within it, because all the claim does is to set out the limits for the cross-sectional area of the recesses. To assess inventive step and sufficiency, it is necessary to add any teaching from the Patent as to why these limits matter, if at all.

As to this, the key paragraph in the specification is [0069], in combination with the schematic in Figure 3. I have already set both out above. The patentee has identified a problem with air-blowing, namely the tightness of the seal S where the OFC enters the micro-duct. If the seal is too loose, air escapes and there is insufficient air flow to carry the cable along the duct. If the seal is too tight, not enough air can enter the duct and again blowability of the OFC will be poor. The Patent puts forward a range of cross-sectional recess areas which are said to allow sufficient air to enter the micro-duct (and remain there) and thereby to improve the air blown performance of the cable.

Although schematic, Figure 3 of the Patent resembles the photograph at Figure 12 of the Agreed Statement of CGK, which is also set out above. The cable seal on the right hand side of the picture is the same as the seal S shown in Figure 3.

There was nothing in the Agreed Statement of CGK which suggested that those in the field were conscious that the external contours of the OFC could affect the seal in such a way so as to alter the air-blown performance of the cable.

This was supported by the witness statement of Mr Kumar filed on behalf of the Defendant which exhibited a 2020 (i.e. post-priority) Gas Leakage Study. Mr Debban considered that this supported his view that sealability issues were not part of the CGK as they did not appear to have been appreciated by Sterlite prior to 2020. Mr Debban was not cross-examined on this topic.

So there was no awareness in the CGK that this was a problem. Mr Sutehall accepted as such in cross-examination and the only document which he could point to in the case which he thought addressed the issue was the AFL prior art, which it was not suggested was CGK.

The Defendant sought to argue that the technical contribution ultimately relied on by the Claimants was only raised for the first time at trial, had no basis in the evidence and should therefore be rejected. I agree that the sealability advantage gained more prominence at trial than it had done in the written evidence. However, its finds clear basis in the specification of the Patent and is supported by the agreed CGK concerning the air inlet unit of a typical cable blowing machine – which in turn resembles Figure 3 of the Patent. It was also supported by the answers given by the Defendant’s expert in cross-examination.

Accordingly, this appears to be one of those patents in which the contribution includes the identification of the problem as well as the alleged solution. In my judgment the technical contribution is the identification of the tightness of the seal in the pump as contributing to air-blown performance, and the disclosure of a range of recess cross-sectional areas which are said to optimise the seal in such a way as to improve the blowability of cables in micro-ducts.

These attacks were pleaded and opened in multiple ways, but by the time of Closing the Defendant was content to express them under three headings, as follows.

The essential point is as follows. The scope of the claims is said to exceed the technical contribution because merely choosing cables with a recess cross-sectional area within the claims does not guarantee a successful product. To support this the Defendant relies on the Patent itself and the teaching about other necessary features, such as rib twist, which needs to be 10 degrees or larger to pass the patentee’s test. The claim covers cables without a rib twist and according to the Patent these would not work.

The Defendant also relies on the absence of any limitation as to diameter in the claims, and argues that the claimed areas would not provide any technical benefit for bigger cables. Finally, there is a squeeze on the rib designs in the prior art; if it is said that some of those will not work, then the same must be true of the claim because there is no limit as to design there either.

Taking the rib twist argument first, I reject this. I agree with the Claimants that the reader of the Patent would understand the separate disclosures to be both cumulative and independent of each other. Thus, whilst they would understand that the teaching of Table 4 and [0075] showed that cables with a rib twist of over 10 degrees were better than cables with no rib twist, they would not think that cables with no rib twist would not work at all. After all, the cables tested for recess area in the Patent (Table 3) were not said to possess a rib-twist, yet still performed well.

Further, the Defendant did not put forward evidence from its expert to suggest that cables without any rib twist would not perform adequately enough to be air blown at all. This would fly in the face of its own products (the alleged infringements), which do not contain a rib twist but which the Defendant no doubt contends as having effective air blown performance.

It is necessary to apply common sense to the breadth of the claim. The different features of the Patent specification are presented separately and the reader would understand that it was not practical to combine all of them into one claim. The fact that other features are shown also to improve performance through testing and by the use of OK/NG terminology in the specification does not mean that the absence of those features means that the cable does not perform at all.

No doubt the claimed cables could be improved further if certain additional features such as rib twist were added, but without those features I do not consider that the claims are not enabled across their breadth. Just as in Kymab v Regeneron [2020] UKSC 27 where the Supreme Court rejected the notion under the head of classical insufficiency that the breadth of the claim could be divided by reference to irrelevant features (there, the length of a mouse’s tail), here the presence or absence of rib twist is an irrelevant feature as far as the claim to recess area is concerned.

Turning to the allegation based on the absence of any restriction on diameter in the claim, for the reasons given above in relation to micro-duct, I do think that there is a relevant restriction in the claim which prevents it covering cables of size whereby the claimed areas would have no real effect. Moreover, the Defendant made no attempt to demonstrate that there is a cable suitable for inserting into a micro-duct which would not demonstrate improved blowability using the parameters for recess area specified in the claims.

It may be initially surprising to see a claim with apparently universal teaching for a range of cable sizes, and no attempt to scale the effect, but this can be explained by the underlying theory which is based on the seal of the air compressor. Given the overall limit in the size of micro-ducts, there are only so many pumps which could be used to insert the OFCs. Once this teaching is understood, it is credible that cables with a relative narrow range of recess areas could all have improved blowability.

The Defendant attempted to explore the edges of the claim in reply evidence with examples which it said illustrated that at the upper and lower ranges of diameter, cables falling outside the claim would be expected to perform better than those within it. It was suggested that the cables on the right hand side falling outside the claim would perform better than those on the left which did:

But it was not established in evidence that the skilled person would want to make any of these cables. On the contrary, Mr Debban described them as “absurd”. I agree that they amount to extreme examples rather than real world cables which might show that the boundaries of the claim are arbitrary.

To succeed in a plea of AgrEvo obviousness the Defendant would have to demonstrate that there were realistic examples falling within the claim which did not provide the benefit promised by the Patent. The evidence did not establish this. The fact that there may be examples outside the claim which also provide a benefit over the prior art does not render the claim invalid – a patentee is entitled to frame its claim more narrowly than it needs to in order to exclude an accidental anticipation, for example. Based on the evidence before me the claims in the present case are not AgrEvo obvious.

Finally, I turn to the squeeze on protrusion design. The Defendant contended that if the design of the protrusions mattered for the purposes of the prior art and avoiding a finding of obviousness, there must also be an insufficiency given the lack of teaching in the Patent about this.

The problem with this point is that there is an imperfect squeeze between the prior art and the Patent. When faced with the prior art, the skilled person has a huge range of parameters to choose, starting with the cross-sectional area of the recesses. The point about the variety of protrusion designs is that each will produce a different cross-sectional area, and many of these are likely to fall outside the scope of the claims. So it is not just the design of the protrusions alone which might affect whether or not the prior art could be made in way which falls within the claims.

In contrast, starting with the claims the skilled reader is given the target for cross-sectional area. There was no evidence that if they hit this target, then there are realistic protrusion designs which would nevertheless not work. That is why the squeeze is imperfect. Further, the burden was on the Defendant to establish that there were possible protrusion designs falling within the claims which would not work, but it did not attempt to do so.

I reject this argument too. All the breadth of claim attacks fall short of establishing invalidity. I therefore turn to conventional obviousness.

The prior art known as Pedder is US Patent 7,087,841 B2 granted to inventors Fett et al, including Jason Pedder, who was well known in the field, and assigned to Fitel USA Corp. Its original priority date was March 13 2001. The granted US patent was not published until August 2006.

At one stage it was suggested by Mr Sutehall that the contents of the Pedder patent were CGK at the priority date. However, he rowed back from this in his oral evidence. It may not matter to the outcome of the case, but in case it does I find that Pedder was not CGK. Indeed, whilst it again might not matter because they are found in the pleaded prior art, I do not even think there was evidential support for Mr Sutehall’s assertion in his written evidence that ribbed cables were CGK at the priority date. The nearest to this in the CGK was the beaded surface design which it was acknowledged enhanced blowing performance.

The abstract of Pedder explains:

A communication cable for being used in ducts or tubes comprising an outer jacket (1) which is formed with ribs (4) spaced around the periphery of the jacket and extending along the length of the cable. The new profiled cable has a lower index of friction when being installed in the ducts or tubes.

Figure 1 is as follows:

So it can be seen that Pedder discloses communication cables with external protrusions and recesses. The Background of the Invention explains that the problem sought to be addressed by the patent is too much friction between cables and ducts when trying to push or blow them into place. The object of the invention is the provision of cables showing lower friction during installation to improve their performance during air blowing.

The penultimate paragraph in column 1 of the specification reads:

The invention also proposes a method of installing cables in ducts or tubes wherein a cable is used having ribs which extend along the cable. One end of the ribbed cable is introduced in the duct and fed forward. Simultaneously, compressed air is admitted into the duct and flows along the length of the ribbed cable. This has a dragging effect on the cable. The effect is increased by the fact that the surface of the cable is increased by the ribs. Furthermore, the flowing air finds flow channels between the ribs and adjacent inner walls of the ducts, and since the lower channels are narrower than the upper channels, pressure in the lower channels is higher than in the upper channels where the velocity of the flowing air is higher than in the lower channels. This makes a lifting effect onto the cable thus decreasing the pressure of the contact surfaces between the cable and the lower duct walls. Decreasing contact pressure means lowering the friction between cable and duct wall.

So what is proposed is that the ribs on the outside of the cable have an aerodynamic effect which both increases the surface area of the cable (so the air has more to “drag” on) and which creates lift and lowers the friction between cable and wall.

Various rib designs are then described in the detailed description. It is said that the distances between adjacent ribs are preferably larger than the thickness of the ribs at the root end. Further, each cm of circumference is said to carry five to twenty ribs. The height of the ribs has a range from 0.1 to 2mm. The specification continues:

The height of the ribs 4 may vary in a range from 0.1 to 2 mm. Cables of the kind described may have an outer diameter in the range of 1 to 30 mm and preferably between 5 and 20 mm. Dependent on the diameter of the cable, the jacket may include 15 to 90 ribs.

So there is a wide range of possible designs taught.

The specification goes on to acknowledge that a balance has to be achieved because the presence of ribs will increase the contact pressure at a reduced contact surface area, and that this may increase friction overall. However, it is said that the cables of the invention exhibit lower friction, as demonstrated by tests carried out on “a standard cable having an outer diameter of 12.2 mm and the new cable having the same outer diameter plus ribs having a height of 0.5 mm. The inner diameter of the test ducts or tubes for guiding the cables was 20 mm.” The figures do indeed suggest improved blowability performance for the ribbed cables of the invention.

The claims reflect the specification. Thus claims 1 to 5 read as follows:

1.

A cable for being used in ducts or tubes, comprising:

an inner core of signal transmitting means, and an outer jacket (1) of protective material covering the signal transmitting means, characterized in that the outer jacket (1) is formed with ribs (4) which are spaced around the periphery of the jacket (1) and extend along the length of the cable, wherein each rib (4), seen in cross-section, has a first end connected to the jacket (1) and a second, free end somewhat rounded, each rib tapering in that width of the rib reduces from the first end to the second end.

2.

The cable according to claim 1, wherein the ribs have a height in the range between 0.1 to 2 mm.

3.

The cable according to claim 1, wherein the number of the ribs along the periphery of the jacket is between 5 to 20 per cm circumference length.

4.

The cable according to claim 1, wherein the outer diameter of the cable is in a range between 1 and 30 mm.

5.

The cable according to claim 4, wherein the range is between 5 and 20 mm.

There was no dispute that the skilled person would be interested in developing a range of OFCs for air-blowing at the priority date, and that this would include IBRs. This is the starting point for obviousness. However, the evidence was that IBRs were relatively new at the priority date and had not been widely commercialised in the UK. Further, it was known that they would not blow as well as conventional cables. Pedder’s design is applied to a conventional core, and there is no mention of IBRs. Although the evidence was that this would not prevent the skilled person wanting to make an IBR range after looking at Pedder with interest, it needs to be factored into the assessment.

In his written evidence Mr Sutehall noted that Pedder provided guidance on the appropriate rib height, spacing, shape and number which he suggested the skilled team would use when developing their own ribbed sheath optical fibre cables. He also said the team would take into consideration the robustness of the cable, whether it could be manufactured and whether the seal in the air box of the blowing machine could form a tight seal around the cable to minimise air leakage from the air box. He attributed these features to the CGK. As a result he suggested that the skilled team would adjust the design of the ribs shown in figure 1 of Pedder and apply it to an IBR, and would come up with something that Dr Malburg calculated would fall within claim 1 of the Patent.

Dr Malburg’s Fig 47 compared cables of different diameter and protrusion width and showed whether they would fall within the recess area claimed in the Patent. Depending on the parameters chosen it is clear that there were designs of cables which would satisfy the requirements of the claims (in between the two black lines), but also plenty of cables (probably the majority) which would not:

As will be apparent from the discussion of the Patent above, I find that a consideration of the tightness of the seal in the air box was not part of the CGK. There was nothing in any of the papers in the case to support this notion and Mr Sutehall accepted as much in cross-examination. He also accepted that there was nothing in Pedder about sealability. Therefore the approach taken in his written evidence to the question of obviousness was flawed. It injected a hindsight consideration which would not have been in the mind of the notional skilled person at the priority date. The Defendant sought to submit in Closing that an understanding of seal leakage was irrelevant to its obviousness case, but it is clear that it was an integral part of Mr Sutehall’s written evidence.

In his oral evidence, Mr Sutehall took a rather different view. He accepted that the skilled team seeking to implement Pedder would start with Pedder’s own specific embodiment, that is to say a 12.2mm cable with a design which resembled Figure 1. Such a cable would fall well outside the scope of the claims based on Dr Malburg’s graph (with a recess area of around 15mm²). Further, Mr Sutehall acknowledged that there was no teaching in Pedder to tell the skilled team how changing any of the variables (shape of ribs, spacing etc) would affect the blowability of the cable. This would be particularly the case for IBRs, about which the skilled person had little experience and for which predicting blowability would be even more difficult because they bend differently and have different stiffness. Mr Sutehall agreed that the skilled team could therefore not predict from Pedder what the effect of varying Figure 1 would be and that there was an almost endless range of permutations and combinations.

As a result, there is nothing in Pedder to encourage the skilled team to redesign Figure 1 for an IBR in a way which results in something falling within claim 1 of the Patent. On this basis the skilled person never gets as far as considering the sorts of ranges of parameters modelled by Dr Malburg in his Figure 47.

The Defendant sought to suggest that even if the skilled person started off with Pedder’s preferred embodiment, if they also made cables of less than about 5mm diameter with Pedder’s protrusions, Dr Malburg’s graph showed that such cables would fall within the claim. But it was never established that the skilled team would carry out this exercise in parallel. If they made Pedder’s preferred embodiment of diameter 12.2mm with an IBR and found (according to the results in the Patent) that it did not blow well, why would they go on and make further cables to the same design with much smaller diameters? In any event, the Defendant suggested that the normal range of cable size at the priority date was 3.5-15.5mm. It is only at the very bottom of this range (somewhere less than 5mm) that the Pedder design falls within the claim on Dr Malburg’s graph, and it was not suggested that the skilled team would make cables this small, even if they investigated multiple cables in parallel.

Alternatively, the Defendant suggested that the skilled person would redesign the protrusions. Mr Sutehall thought that the shape of the ribs in Figure 1 would not be robust enough to extrude or be installed, but he had to concede that Pedder itself disclosed that the Figure 1 design had been successfully installed at column 2 line 54. Further, even if the skilled team was concerned about the shape of the ribs in Pedder, they would have no idea whether changing them, for example to the triangular design suggested by Mr Sutehall, would still give the benefit of low friction promised by Pedder. In any event, taking Pedder’s 12.2mm cable and varying the 0.5mm protrusions to 0.3mm or 0.2mm would still not fall within the scope of the claims based on Dr Malburg’s Figure 47.

So whilst a skilled team seeking to implement Pedder with an IBR might choose to test a different cable size and different rib design in the course of their work, there is no telling what different parameters they would choose. The assumptions made by Mr Sutehall in his written evidence cannot apply because of the injection of impermissible hindsight. Whilst this exercise could result in a cable which satisfied claim 1 of the Patent, it is far from clear that it would, even if the skilled team were to embark on such a project. Further and significantly, without knowledge of sealability, the skilled person would not choose different parameters with any expectation of success, particularly for an IBR – they would not even be sure that they would maintain the reduction in friction taught by Pedder.

There was a hint in the submissions of Sterlite that hindsight could be excluded from Mr Sutehall’s written evidence because he had provided Dr Malburg with his proposed design and it was only Dr Malburg who had calculated that it fell within the scope of the claims. That may be, but what this submission did betray is the fact that Mr Sutehall did not say in his expert evidence that he had given the views in his report on the prior art before he was aware of claim 1 of the Patent. I do not consider that evidence tendered by experts who have not been subject to sequential unmasking should necessarily be given less weight, as it is sometimes impossible to achieve this, and even when the practice is followed it is often difficult to disentangle their unvarnished opinions from what appears in the report. But I discount any notion that Mr Sutehall’s evidence can be treated as some sort of survey of what a skilled person would have done without knowledge of the invention.

For all the above reasons I find that Sterlite’s obviousness attack over Pedder expressed through its expert Mr Sutehall fell well short of what is required to establish lack of inventive step. As is to be expected, Sterlite also sought to rely on some of the evidence of Fujikura’s expert Mr Debban. Of course, his written evidence did not support a finding of obviousness. So it was necessary for Sterlite to try to establish an alternative case based on the cross examination.

As to this, Mr Debban did give an answer at the end of Day 1 where he agreed that it would be almost impossible to avoid falling within the range in claim 1 if Mr Sutehall’s suggested variant of Pedder was applied to the normal range of cable diameters. This answer in essence reflected Dr Malburg’s graph reproduced above. But it could only assist Sterlite if all the preceding assumptions and steps embodied in Mr Sutehall’s reasoning were valid. As explained above, they were not supported by Mr Sutehall’s oral evidence. Nor did Sterlite establish them during the cross examination of Mr Debban. I have read and re-read the entire passage leading up to the answer given by Mr Debban at the end of the first day, as I was urged to do by both parties. Although Mr Sutehall’s steps were put to Mr Debban, it is clear that Mr Debban did not accept that they necessarily represented the steps of the notional skilled team, and he was not really pressed on this. So whilst he was right to agree that if the skilled team carried out those steps, the result would be some cables falling within the claim, the validity of the steps was not established with him.

So Sterlite cannot rely on the oral evidence of Mr Debban to make up for Mr Sutehall’s concessions given in cross-examination. It has failed to establish its case of obviousness over Pedder.

The AFL document is a sales brochure produced on behalf of AFL/Duraline. It has a copyright notice of 2013 on it. It advertises several different types of OFC.

The first page promoted an Enterprise Blown Fiber Cable in the following terms:

The diagram clearly shows a cable with a ribbed exterior. The text refers to a “patent pending cable design” which “combines a light-weight, high-drag jacketing system that allows the cable to be blown long distances”.

These cables are not intermittently bonded cables, as can be seen from the diagram. They are therefore different from the cables described in the Patent.

Below the description are tables of specifications, including estimated installation distances of up to 3000ft. The following two pages include further specifications for different diameter cables falling within the range.

The fourth page discloses a different type of cable, called the eABF® SWR® Enterprise Air-Jetted Fiber Cable. This is an intermittently bonded or “spiderweb” ribbon. However, in contrast to the cables shown on the first page, this OFC has a smooth outside. This can be seen from both diagrams on the page:

The text focuses on the advantages of the intermittently bonded design:

There follows a page of specifications for this cable.

Finally, on page 6 there is a third cable range described as follows:

As can be seen, this is once again a ribbed design and the same reference to “patent pending” advantages are referred to. There is no mention of intermittently bonded cables.

Overall, it can be seen that whilst there is disclosure of cables with ribbed exteriors, none of these are the IBR designs claimed in the Patent. In contrast to the “patent pending” ribbed designs, the IBR cables are only promoted by reference to a conventional smooth exterior.

Fujikura submitted that AFL would be dismissed as a technical document because it was a “sales” brochure. I think this argument was overplayed. Granted, the document is not meant to be a full technical specification. But it does contain interesting technical information and promises that the new jacketing system can allow cables to be blown long distances. It would be read by the skilled team with interest.

A much better point in Fujikura’s favour in my view is the fact that whilst the new jacket is applied to conventional cables, it is not applied to the spiderweb ribbon IBR design. This is shown with a smooth jacket and is said to be capable of being jetted thousands of feet. As mentioned above, IBR cables were thought to have certain advantages at the priority date but the skilled person would have very little experience designing them.

I consider that the notional uninventive reader of the document would consider that AFL had a reason for not suggesting that the ribbed jacket should be applied to the IBR spiderweb cable. This might come from their CGK, and the knowledge that IBR cables blow differently to conventional ones. Or it might be that they would assume that AFL had tested the jacket system on the spiderweb design and found that it did not improve performance.

This was consistent with the evidence of the experts.

Mr Sutehall agreed in cross examination that the reader of the brochure would have assumed that AFL had tried the ribbed jacket on the IBR but found that it did not help or even that there was a problem with it:

Q. The reason the natural conclusion for the skilled team is to come to is that AFL must have thought that it was not worth bothering with ribs on IBR because there was a problem; it did not work well enough or it did not work at all. That is the natural conclusion to reach here, is it not, if you are trying to get to an explanation why it is not there?

A. Without the evidence, yes, you have to assume that.

Mr Debban was of a similar view. In the absence of any technical teaching as to how the ribbed design on the first page of the brochure improves blowing for those cables, the skilled team could have no expectation of success for IBRs. The obvious thing to do from the AFL brochure for IBR cables was to use the AFL non-ribbed design. The absence of any teaching in the brochure as to why the ribbed design might work (for non IBR cables) compared to the Patent also explains why Sterlite’s attempted insufficiency squeeze falls flat.

Even if the skilled team had wanted to try what AFL had apparently chosen not to do, it is also unclear what design they would adopt. The diagram on the last page of the brochure was said by Mr Debban to be representational and not something which could be scaled. Mr Sutehall agreed it was not an engineering drawing but was more inclined to use it is a starting point. Having re-read the evidence on this point and assessing the issue in the round, I think the skilled team would understand that the picture was merely a schematic and not meant to be scaled from. If they were really interested in the design, I think they would try to find the “pending” patent and follow that teaching, but there was no evidence about that. On this basis Mr Sutehall agreed that the skilled team would not know how to implement the combination of ribbed cable and IBR. So the skilled person is left with a design project and there is no clear starting point.

Even if the skilled person was inclined to scale from the non-IBR design on the last page of the brochure, and even if some of these designs had been shown by Dr Malburg to satisfy the claims (at 3.8-4.5mm diameter), there is certainly no basis for suggesting that the skilled person would make a cable of the correct size and design to fall within the recess parameters claimed. Even less that they would do so with a reasonable expectation of success.

There was a side issue arising from Mr Debban’s assumption about what AFL actually did in the market place before and after the priority date, which Sterlite sought to deal with using CEA material. This was a pre-priority AFL brochure containing a picture of a ribbed IBR cable. But this was not cited prior art and was not established to be CGK. It cannot advance the pleaded obviousness case and the reasoning I have outlined above does not depend on Mr Debban’s understanding about what AFL did.

In short, I find the arguments for obviousness over the AFL Brochure even less persuasive than those over Pedder. There is much less teaching in it as to why the skilled person should adopt a ribbed design, and there is positive reason not to apply that to an IBR (because AFL had not done so in the document). Absent hindsight, there was insufficient impetus to encourage the skilled person to try to apply the ribbed design to IBR, let alone with a reasonable expectation of success.

Finally, I need to deal with the pleaded collocation. This argument was advanced by the Defendant at a stage in the proceedings when it was not clear whether it would be accepted by the Claimants that IBR formed part of the CGK. The Defendant accepted in Closing that the collocation point was less important in the light of the acceptance that IBR was CGK and so both Pedder and AFL could be approached with IBR cables in mind. Nevertheless the point is still live and I must deal with it.

It was said by the Defendant that to the extent that it was not obvious to use IBR with a ribbed jacket (as I have found above in discussing AFL), then the ribbed jacket and IBR are separate and obvious inventions. Two citations were relied on for the disclosure of IBR, known as Spider Web and JP ‘747, both of which it was accepted disclose IBR.

The collocation approach is a way of permitting non cross-referenced pieces of prior art to be combined in circumstances where the rule against mosaicking would normally prevent this. The justification for it is that the claim in issue merely combines two obvious but non-interacting parts, and that there can be no invention in the skilled person doing the same, even if there is no single piece of prior art which suggests the combination. The keystone is that the two features must not influence each other. The EPO Guidelines refer to an absence of “synergy” between the features, but that rather depends what is meant by “synergy”. The example of a sausage machine is often used – the meat mixing part is separate from the part which places the product into the sausage skins, and can be just bolted on as a collocation.

The leading case is the House of Lords decision in Sabaf v MFI [2005] RPC 10. The Defendant relied on this for the proposition that it is necessary to demonstrate true synergy (in the strict scientific sense) between the two features to overcome a collocation plea i.e. the effect of the two parts combined needs to be more than merely additive. The Claimants relied on Illumina v Latvia [2021] RPC 12 for the explanation that true synergy is not required and all that is necessary to overcome a collocation attack is to demonstrate some sort of interaction between the relevant features.

In the present case the Defendant accepted that there was a relationship between the choice of IBR for the interior of the cable and the choice of protrusion design insofar as blowability characteristics are concerned. Mr Sutehall explained that IBRs are more difficult to blow than MLT cables. He also gave evidence that IBRs impart more rigidity than non-adhered fibres and less rigidity than a traditional ribbon, which will affect the air blowing characteristics. Despite this the Defendant argued that there was no synergism between the two and so a collocation attack was still permitted.

I reject the contention that Lord Hoffmann in Sabaf was laying down a principle that collocation arguments could only be avoided where there was synergy. See in particular his §26 where he refers not only to synergy but also to whether the features interact upon each other and contrasts this with where features perform independently:

If the two integers interact upon each other, if there is synergy between them, they constitute a single invention having a combined effect and one applies s.3 to the idea of combining them. If each integer “performs its own proper function independently of any of the others”, then each is for the purposes of s.3 a separate invention and it has to be applied to each one separately. That, in my opinion, is what Laddie J. meant by the law of collocation.

I consider that collocation arguments cannot apply whenever there is an interaction between the two features which is relevant to the invention, consistent with the observations of Birss J in Illumina. A collocation approach is only permitted where the features are truly independent, as in the example of the sausage machine.

In any event I do not see how the collocation argument can assist the Defendant in the present case. The reasons for rejecting the conventional obviousness case do not depend on the skilled person not knowing about IBR, and so the conventional obviousness case would not be improved by showing the skilled person Spider Web or JP ‘747.

I reject the idea that in the present invention combining the IBR with the ribbed outer sheath amounts to an obvious collocation.

This was the most difficult area of the case to determine.

It involves first arriving at an understanding of the correct test to carry out according to the claim, and then determining whether the claim is bad for uncertainty insufficiency. Only then is it possible to go on and decide if the Claimants had successfully established infringement of the Celesta 432F Sample Cable tested in the experiments. Then there is a final decision to be made as to whether the Claimants are permitted to and have in fact established whether there is infringement of the other cables in the Celesta range, excluding the 576F cables which were dropped by the Claimants during the trial.

I am well aware that the issues of construction and interpretation of the claim need to take place without an eye on the alleged infringement. I have sought to do that in the present case. However, I still believe that it is convenient to deal with all such issues in this final section of the judgment because they follow on from one another.

I should also make clear at the outset that the experiment protocol adopted by the Claimants in their attempt to prove infringement was unusual, in the sense that it was neither approved by their experts in advance, nor was it said to be what the skilled person reading the Patent would have done. Instead, it appears to have been concocted and carried out at the behest of Fujikura. It was characterised by the Claimants at trial as being “over-engineered”, yet still giving results which could be used to assess infringement. This further complicates the issues which I have to determine. I was told that the Mayne Pharma disclosure process underlying the experiments had been complied with but I was not pointed to any additional materials.

I have set out the claim in full above. There was no dispute that what it requires is to compare the cross-sectional area of a virtual optical fibre with the area of the actual cable in question. The issue is what characteristics to ascribe to the virtual curve. In particular, what does the skilled person do when the actual cable in question is neither circular nor elliptical, where the virtual shape being fitted requires a different central axis, where the closed curve is not purely tangential and/or does not contact each radial outer end of said protrusions.

The Defendant’s primary position was that the claim “was not in any way difficult to construe”. It said that the features must be interpreted strictly such that any of the four possible deviations identified above were not permitted. If it was wrong about this, it said that the claim suffered from uncertainty insufficiency, relying on the principle in summarised by Floyd LJ in Anan Kasei v Neochemicals and Oxides [2020] F.S.R. 8 at §25:

“25.

As Lewison LJ points out in his judgment, the objection to the claim in Kirin-Amgen [2005] R.P.C. 9 is not correctly described as “ambiguity”. The claim was conceptually uncertain. This type of insufficiency is far better described as “uncertainty”. The process of interpretation could not resolve the question of what uEPO the patentee had in mind for the necessary test. The consequent burden which this placed on the skilled person meant that the specification was insufficient. Jacob J gave an example in Milliken Denmark A/S v Walk Off Mats Ltd [1996] F.S.R. 292 at p.301 of a property which was required to be measured in the non-existent “Pinocchio units”. That would give rise to uncertainty in the Kirin-Amgen sense.”

The Claimants’ approach to construction focussed on some illustrations provided by the Defendant’s expert, Dr Malburg. The Claimant argued that the skilled person reading the claim in the light of the specification would understand that the middle of the three examples below was the virtual curve which should be applied:

As can be seen from the sketch, the figure on the left has straight lines between the protrusions whilst the figure on the right has individual radii drawn, neither of which is contemplated by the claim. The centre representation is not a strict circle but the Claimants submitted that it had a circular/elliptical shape within the meaning of the claim and touched tangentially the tip of each protrusion.

The starting point for the Claimants’ approach was the virtual curve shown in Figure 6 of the Patent which I have reproduced above. The justification for varying this to allow for something resembling the central figure above was that the skilled person would understand from the specification that the virtual curve was meant to be an approximation of the lip seal in the air pump. It was said that real world lip-seals would deform to follow the shape of the cable in a way approximated by the central sketch.

I think that the answer to these questions is obtained by construing the claim purposively, as I am required to do. I also bear in mind the practical approach of the notional skilled team.

To that end I return to the specification of the Patent. There is precious little in there to assist with how strictly the skilled person would understand the terms such as “circular” and “elliptical” in the claim to be interpreted.

[0017] does explain that “[w]hen the optical fiber cable 1 is non-circular in the transverse cross-sectional view, the central axis O is positioned at the center of the optical fiber cable 1”. This is not terribly illuminating as to how the centre is meant to be ascertained in a non-circular cable, but it does at least admit of the existence of non-circular cables.

I have already set out [0034] above. This acknowledges that the outer dimensions of the cable according to the invention may not be even, i.e. that some protrusions may stick out further than others. In such circumstances an average should be taken, as Mr Sutehall explained in his first report. At the same time he noted the lack of guidance as to how to ascertain the centre in a non-circular cable.

[0067] and [0068] provide some further commentary. [0067] explains that “the closed curve L in contact with the radial outer end of each protrusion is drawn”. [0068] goes on to say that “[t]he closed curve L is usually circular with the central axis O as the center. However, due to the deformation of the optical fiber cable, the closed curve L may have an elliptical shape”.

So the potential for deformation is expressly recognised, as is the potential for an uneven cable. The solution is to use an elliptical shape or to average the results.

This is all consistent with the skilled person’s CGK. The agreed statement quoted above acknowledges that it was not uncommon for an optical fibre cable to become ‘oval’. Yet as can be seen from the prior art, it was conventional to represent OFCs as being perfectly circular in trade literature.

Finally I note [0069]-[0070], quoted above, and the express references to sealability. There is no doubt that the reader would understand why they were being taught to control the cross-sectional area of the recesses. This is emphasised by the concluding sentence of [0072]. Although the closed curve L is introduced first, when the section of the specification is read as a whole I consider that the reader would understand what the closed curve was meant to represent, namely the contours of the seal.

The Claimants emphasised that the wording used in the specification and in the claims refers to a circular or elliptical “shape”, and not simply to “a circle” or “an ellipse”, which would be understood as having a more precise geometric meaning. I agree that the inclusion of the word “shape” supports a less strict approach and is consistent with what might be expected in a real-world situation.

Both Cable Experts addressed the issue. Mr Sutehall suggested that the patentee might have had in mind the metal dies from which the cables are extruded, or CAD drawings of such dies to calculate the area of the recesses. He described this as “an idealised cable design” compared to the finished cables – however, there is nothing in the specification to support this interpretation. He also explained how the diameter of an actual cable would be expected to differ from the nominal diameter of the die as a result of the “draw down ratio” which will vary from product to product.

For his part Mr Debban expressed the view that the reference to the closed curve being tangential to the radial outer ends of the protrusions in the claim was the equivalent to the reference to contact with the outer ends in [0067]. As far as the circular or elliptical shape is concerned centred on the central axis, he considered that this would be generally fulfilled when the protrusions were approximately the same height. He did not think that real life cables could ever be fitted with a perfect circle or ellipse.

This evidence from the experts is of questionable admissibility to the extent it goes strictly to the issue of construction. But I do think it supports the Claimants’ general submission that when read through the eyes of the practical team designing and installing cables, the Patent would be understood as describing an approximation to a real-world cable. As such, the skilled reader would not expect the patentee to have been proposing that the virtual curve needed to form a perfect circle or ellipse, or that it had to be perfectly tangential to the protrusions rather than touching them, or that the central axis had to be identical for each.

The Claimants placed significant weight on the fact that although there was a debate as to how to carry out the test in the claim, there was no suggestion that any of the results that had been recorded in the experiments of the case would have varied in such a way so as to render an apparently infringing cable non-infringing had the technique been carried out differently. In other words, the recess area of the alleged infringement landed squarely in the middle of the claimed range, so the arguments about clarity went nowhere.

That may be the reality in the present case, which makes it easier to argue that the quibbles are merely “puzzles at the edges of the claim”. But I was reminded by the Defendant of Arnold J’s observations in Sandvik v Kennametal [2012] RPC 23 at §164:

Although these effects will only make the difference between infringement and non-infringement for coatings which are reasonably close to the lower limit in integer [5], be it 1.3 or 1.5, in such circumstances it is impossible to say whether the product falls within the claim or not, because it is uncertain what the correct test is. Thus it is not merely a case of the claim having a fuzzy boundary. Accordingly, I consider that counsel for Sandvik was right to concede that in this event the Patent is insufficient.

Accordingly, the mere absence of a numerical example in the case which could be inside or outside the claim does not mean that I should conclude that there is no relevant ambiguity. However, the Defendant’s position would have been much stronger if such an example had existed.

Standing back and taking all of the above into account, I am not satisfied that the claim is bad for ambiguity. It is correct that the precise boundaries of the claim are not straightforward to define – for example, how circular in shape does the curve need to be? But many claims have puzzles at their edge. Despite this, the claim is capable of being construed and the test is capable of being implemented. A curve L needs to be drawn round each of the protrusions of the cable to form a circular or elliptical shape. This would be understood as approximating the lips of the seal S referred to in the specification of the Patent. Mr Sutehall agreed with this when Fig. 6 of the Patent was put to him.

Mr Sutehall also agreed that the middle one of Dr Malburg’s sketches, reproduced above, was the most representative of what would happen at the seal when blowing a real-life cable. He was then asked directly about the relationship between Dr Malburg’s central sketch and the teaching of the Patent:

I want to deal with the technical teaching in so far as it is. If you have one in the middle, the closed curve is doing what paragraph [0069] is telling you to do, which allows you to calculate what the hole through which the air can escape is. Do you agree with that?

A. I agree, yes.

Q. We know, I think we can agree that the patent is not concerned

with perfect circles or perfect ellipses, is it?

A. True, yes.

Q. It is concerned with real-world cables; yes?

A. Yes.

Q. As the judge rightly reminded me, construction is a matter for the judge, not for you or Dr. Malburg. But I think you would agree with me that the person would have firmly in mind two things when reading the patent: technically that he was being taught that he needed to calculate the sort of recesses between the seal and the cable that we see there, yes?

A. Yes.

Q. And it was teaching that he could do it in circumstances where the cable was not perfectly round. Do you agree with that?

A. But it does not mention the diameter, so the diameter will affect the recess area.

I have dealt with the diameter point above. Otherwise I take comfort from Mr Sutehall’s answers that the reader of the Patent would understand that virtual curve L should be drawn in a similar way to the middle one of Dr Malburg’s sketches.

As far as the individual points raised by the Defendant are concerned, in my judgment and adopting the purposive and practical approach of the particular skilled reader in this case, the curve only needs to be approximately circular/elliptical. The consequence of this is that the precise position of the central axis can vary. The curve should be drawn so that as far as possible it touches the protrusions – and “each” one, as the claim specifies. But it need not be purely tangential. It may be that where a particular protrusion is shorter than others the curve, if remaining circular or elliptical, has to miss the top of that protrusion. But that would reflect the real world situation of the lip of the seal. Contrary to the Defendant’s case, this does not amount to rewriting the claim – is simply an exercise in purposive construction taking into account the characteristics of this particular field. Any problems applying the remaining vagaries of that construction may yet come back to haunt the patentee in terms of proving infringement, but they are not so severe that the claim should be found to be incapable of being infringed, and therefore insufficient.

In chief and repeat experiments were carried out on a sample of the Defendant’s Celesta 423F cable. I have already touched upon the fact that the experiments were not designed by the Claimants’ experts in the conventional way.

The written evidence contained numerous disputes going to the way the experiments had been constructed. A number of these fell away at trial. By the time of closing the Defendant still maintained a number of arguments as to why the experiments were flawed. I will deal with the practical objections first before turning to whether, assuming the results are valid, the assessment of recess area met the requirements of the claim.

The Defendant objected to the way in which the Sample Cable cross section had been chosen. The point is that the profile of the cable is likely to vary along its length. It said that a single cross-section from a single cable at a single point in the drum was unlikely to be representative. Ms Jarvis accepted that it was not good scientific practice to take just one sample for an experiment such as the present. But I have nothing else to go on – only the result of the repeat (and any comparison with the result obtained in the non-witnessed experiment in chief). Had the Defendant wished to take a sample from elsewhere on the drum then it could have sought permission to do so or carried out its own experiments in chief or reply.

Similarly, the very exercise of cutting the cable for measurement could have an effect on its shape. This can be avoided to some extent by encasing the cable in resin to protect it, as Fujikura did when carrying out their experiments. Even then, the Defendant criticised the preparation of the sample, pointing to abrasions left on it as a result of the polishing process. It also suggested that the cross section had been tipped at a 10^o angle when mounted, creating an elliptical cross section. Overall it said that the appearance of the cable tested in the experiments differed from that shown in the body of the PPD (albeit that the resemblance to the figure in Annex 1 was much greater).

I do not think that there is anything in these points. The problem with all of these gripes is that the Defendant accepted that none of them could be shown to take the results of the experiments outside the range in the claims. Further, the Defendant did not challenge the evidence of Mr Heinze who explained how the areas of the recesses had been chosen and measured during the Repeats. No doubt it would have been better if multiple samples had been taken along the length of the cable. It is also possible that the polishing could have been carried out differently and that the cable could have been mounted without any tilt. But once again had the Defendant wanted to show that any of these points mattered, it could have carried out its own studies. After all, these were experiments on its own cables. No-one could know more about those than the Defendant.

The best the Defendant could do was to rely in Closing on a point about the measured diameter of the Sample Cable tested in the experiments. This was 13.52mm. However, according to the PPD the nominal diameter of the 432F cable tested was 12.7mm with a tolerance of 0.3mm. This showed, said the Defendant, that something had gone badly wrong with the experimental protocol.

This argument was itself problematic for the following reasons. Although the point was founded in Dr Malburg’s first report, it was only raised as follows:

146.

To undertake the measurements, the operator loaded a clean image of the cable cross-section (with no areas shaded) in the Keyence software and used the circle drawing function (called “Spc value”) in the Area Settings window to draw a circle having a diameter of 13.520 mm. (In contrast, I understand from Sterlite’s PPD that the 432F cable typically has a diameter of 12.7mm.^{5[Paragraph 43 of Sterlite’s Amended PPD]}).

So it was not positively suggested in evidence that this difference undermined the experiments. It was just a passing comment. Nor, as I point out below, was the point pleaded. Unsurprisingly given the way in which it was floated, the Claimants’ witnesses did not comment on it in reply.

It is fair to say that the point was mentioned in the Defendant’s Opening Skeleton, at §222 out of 254:

This choice cannot be justified from the disclosure of the Patent. It is also, inexplicably, approximately 1mm larger than the nominal outer diameter of the 432F cable (being 12.7mm^{174 [PPD ¶43]}).

It was then raised in the cross-examination of Ms Jarvis, the Measurement expert for the Claimants. I set out the passage in full:

Q. Okay. Let us just think about that diameter for a moment. We know from the PPD that the nominal diameter of this cable(indistinct) is 12.7 plus or minus 0.3 mm. I can show youthat, but you can take it from me.

A. Okay.

Q. That is the range of diameters of this cable. As far as we can see, no one took a Vernier caliper to the cable during the repeats. You did not witness that, did you?

A. No.

Q. So, one has to assume that cable diameter is somewhere between 12.4 and 13 mm according to the manufacturing tolerances; yes.

A. If that is what their PPD says.

Q. That is right. That is what if says; that is right. But, what has happened here is through an elaborate scheme of quadrant fitting with pairs of ellipses and outer and inner protrusions and so on and averaging, we have ended up with an arc diameter of 13.52, which is well outside the maximum possible size of the outer boundary of the cable. That isright, is it not?

A. That is what the PPD says; yes.

Q. What I cannot understand is why 13.52 mm is an appropriate arc diameter to fit to a cable of notional diameter 12.7 mm. Canyou help his Lordship with why that would be the correctdiameter of arc to use in this individual recess method?

A. As a scientist, we measure the diameter using that quadrant method and that was the result. So you could have used anominal diameter in this case, but we, as in I was justobserving, but AFL or Fujikura decided to use the measureddiameter for creating the arc.

THE JUDGE: Can you explain why it is 13.52 when it should be smaller, I think is the question?

A. That is what was measured. I do not know why it is different.

Q. Right. I think the next question is, does that suggest something is wrong with the number, 13.52?

A. I understand that.

Q. What is your answer to that?

A. I can only talk to what I observed in the experiment. I do not think anything as a result of the experiment would have created a diameter outside of the -- If it started as a nominal diameter of 12.7, I do not see anything in the experiment that would expand it to 13.52 or change it to13.52.

Q. Right. Do you mean that means that 13.52 must be wrong or are you saying that there is nothing wrong with that figure because you did not see anything to cause it to be wrong?

A. To me that just says whatever section the cable was taken fromwas not -- was 13.52 or close to 13.52 and not the nominaldiameter.

THE JUDGE: Okay.

MR. GAMSA: A different arc time would have produced different recess areas, would it not?

A. I assume so.

Q. Let us look at this another way. We know, we heard earlier, that recess areas are very sensible to small changes in protrusion height of the order of 0.08 mm. Did you hear that discussion I had with Mr. Debban?

A. Yes.

Q. Here, we are using a diameter which is getting on for 1 mm outside, excuse me, 0.5 mm outside the nominal range of the cable. Did it cause you concern that 13.52 mm was being used as the boundary arc diameter when you witnessed the repeats? Did that strike you as a problem at the time?

A.

No.

The cross-examiner then moved on.

As noted above, the Claimants elected not to cross-examine Sterlite’s PPD witness. But as they were not relying on the numbers in the PPD to prove infringement, but rather on the results of the Repeat, I am not sure it fell on them to do so, particularly given the rather throw-away manner in which the point had emerged in the evidence at all.

So I am left trying to resolve a dilemma in the evidence with no proposed explanation for the discrepancy. The evidence before me suggests that the diameter was correctly calculated in the experiments to be 13.52mm. This is said to be outside the Defendant’s own tolerances, but it was not suggested that the measurement was in error. I have no evidence before me about the Defendant’s own tolerances, or how that is measured and monitored. Further, the point was only explored in any detail with me for the first time in the Defendant’s Closing. As I expressed at the time, this is all rather unsatisfactory given the potential importance of the point.

Like most of the other complaints raised about the experiments and the evidence in support of infringement, it was well within the Defendant’s power to support the point with positive evidence of its own and/or its own experiments adduced at the proper time in these proceedings. It could have measured the diameter of its own cable, taken samples from different places in the drum or provided positive evidence from its fact or expert witnesses to explain why the results of the experiments could not be correct. I accept that the legal burden is on the Claimants to prove infringement, and it remains with them. But all the evidential material to undermine the Claimants’ case lay in the hands of the Defendant. It did not deploy any of this material.

Confining myself to the evidence which is in the case, the cross-examination of Ms Jarvis gave me no reason to doubt the results obtained in the experiments. I am unable to say why these results appear to differ from the nominal dimensions contained in the PPD. Applying the test set out in the claim as I have construed it requires the use of the diameter as it is actually observed in the witnessed repeat experiments. I have no basis to reject the experiments just because this differs from the number in the PPD.

Having accepted that none of the observed discrepancies in the experiments could make a material difference to the results, the Defendants took a different point about the size of the protrusions. It was said by the Defendants that only minor absolute differences in protrusion height would be sufficient to take the results obtained outside the claimed range, and that this demonstrated a flaw in the experiments. Thus, the Repeat Experiments established a protrusion height of 0.105mm, but if that was increased by 0.08mm to 0.185mm then this would have resulted in a finding of non-infringement. The difference is equivalent to the thickness of a standard piece of A4 paper. That may be, but that does not establish that errors of this magnitude were present in the sensitive measurement experiments being conducted. As noted above, Sterlite has an overall tolerance of 0.3mm in diameter for its 12.7mm cable, but this tolerance is not necessarily due to the protrusions, and could be due to the core or sheath. So this does not assist the Defendant either.

In summary, the Defendant made a valiant attempt to criticise the conduct of the experiments in order to suggest that the measurements were not safe. This fell flat, and I reject the criticism of the numbers. That does not mean to say that the technique used by the Claimants necessarily reflected the methodology required by the Patent. This is the penultimate topic with which I need to deal.

As noted above, the Claimants accepted that the experimental methodology was overengineered. The purpose of the methodology was to reach the same target as defined in the Patent, namely calculating the difference in area between a virtual curve surrounding the outside of the OFC and the actual OFC to give the area of the recesses. But the approach adopted to drawing the curve was more complicated than necessary, perhaps out of a (misplaced) desire for accuracy. I need to consider whether this disqualifies the approach altogether.

The method involved two steps. First the scans of the cable were analysed by the “quadrant method” to produce an overall diameter for the virtual curve. This method divided the cable up into quadrants and fitted an ellipse to each quadrant, before combining the four quadrants into a whole:

Further, two curves were fitted – touching the outermost protrusions and the innermost respectively. Both areas calculated from this fell within the claimed range, as of course did the average. The Claimants did not in fact rely on these figures for the area calculation, but did rely on the method to define the diameter of the cable of 13.52mm, discussed above.

The second step used the diameter of the virtual curve calculated from the first step to fit arcs of constant diameter of 13.52mm between neighbouring protrusions:

The area under the arc for each recess was then calculated to give a total recess cross-sectional area said to approximate Dr Malburg’s middle sketch reproduced above. The area in question can be seen below:

The Defendant criticised the output of the individual recess method for a number of reasons. It said that it was not a closed curve, was not tangent to the protrusions, missed some of the protrusions and did not share the same centre as the cable. All of these criticisms were put to the Claimants’ Measurement expert Ms. Jarvis, who agreed.

In my judgment the Claimants’ analysis does have flaws. This is partly because it has attempted to fit a virtual curve according to the claim in the Patent to a real world cable which does not allow a perfect fit. The analysis has also been over complicated by the methodology adopted by Fujikura. There is no suggestion in the Patent that the reader should adopt the approach utilised by Fujikura to draw the virtual curve and I find it highly unlikely that the skilled person would adopt exactly this approach. The position is not helped by the fact that Fujikura did not adduce any witness to explain why it had adopted the methodology it did.

However, the Patent does not specify any method to draw the curve. So what matters is whether the Fujikura approach results in a curve which would not be reached by conventional means and whether this means that the approach is so flawed that I should reject the claim for infringement of the 432F Sample Cable as not having been proven.

Mr Debban thought that a better method for calculating the virtual curve would have been to fit a spline curve to the protrusions. This is not taught in the Patent either. However Dr Malburg conceded that it would tend towards the same result as that achieved by the Fujikura analysis. Further, Dr Malburg accepted that changing the radius of curvature – which is all that the Claimants’ elaborate quadrant method generated – from 13mm to 13.5mm was “pretty small”. So a difference in the output of the methodology used to calculate the radius of the virtual curve would not appear to have made much difference to the calculation of recess area required by the claim.

In opening the Defendant preferred a minimum circumscribed circle approach, which had been suggested by Dr Malburg. This was not something that was pursued with any enthusiasm in cross examination or closing by the Defendant. Although Dr Malburg suggested that it would result in a measurement that fell outside the claimed range (and I have no reason to doubt this), when compared against the claim the approach has even more obvious flaws than the Claimants’ approach. This is because it uses a geometric circle, and not a “circular shape” as I have construed the claim to require. The following extract from Dr Malburg’s first report demonstrates why the results are so different from a curve that comes much closer to touching the protrusions, as I have also held that the claim requires:

So I am left in a position where I have to do the best I can on the material before me and decide whether the Claimants’ imperfect approach is good enough. I consider that it is significant that the Defendant was not able to support an alternative approach which resulted in the drawing of a curve which would make a material difference to the calculated recess cross-sectional area.

Of the criticisms made by Dr Malburg of the Claimants’ approach – the manual definition of the boundary, the placement of the ellipse and the fitting of the arc to each individual recess – none could be shown to make a material difference to the overall calculation of recess area. Ms Jarvis agreed.

So whilst it is possible to pick holes in the methodology employed, given that the overall aim – namely to calculate the area of the recesses – was the correct one, and that the Defendant has got nowhere near being able to suggest that any more reasonable alternative method would have resulted in a materially different result, I decline to dismiss the outcome of the experimental repeats. Even if the virtual curve had been closed throughout, this would have made very little difference to the overall result (a few pixels at the narrowest part). The same can be said for whether it was truly tangential or just touching, whether it touched all of the protrusions or just the majority of them, and whether it shared the same centre as the cable. None of these differences have been shown by the Defendant to make any material difference to the overall result.

Although Mr Sutehall was at pains to point out potential problems in the measurement techniques (e.g. presence of air bubbles, damaged protrusions, tilt) neither he nor Dr Malburg was able to quantify the difference that these issues might have made. Dr Malburg fairly accepted that he could have gone through the measurement steps taken my Mr Heinze and calculated a corrected area based on the alleged criticisms, but he had not done so. In the end I think this is fatal to the many criticisms made by the Defendant. If they had really mattered to the results, no doubt the Defendant would have demonstrated that.

Overall I consider that the Claimants have demonstrated on the balance of probabilities that the 432F Sample Cable has a recess area falling within the claimed range in the Patent. Although some of the decisions made by the Claimants in assembling their case on infringement remain baffling, I am satisfied that they did not make any material difference to the overall outcome.

The final question which I need to determine is whether, having established infringement of the Celesta 432F Sample Cable in the experiments, the Claimants can extend this finding to other cables alleged to infringe in the Defendant’s Celesta range. This involves a preliminary question of law/procedure before turning to the facts.

The preliminary question arises out of the Claimants’ Notice of Experiments, a concession made by the Claimants during a case management hearing before Bacon J dropping original fact (iii) in that Notice and the Defendant’s attempt to amend its pleadings as a result. I traversed much of this material at the PTR when I refused to allow the Defendant to amend its pleadings and I do not repeat the contents of that judgment here ([2025] EWHC 2066 (Pat)). No appeal was sought from that decision. At that hearing I also said that the Claimants would be held to the concession made at the hearing before Bacon J where they deleted attempted reliance on the experiments to prove the following (emphasis added):

“iii That when a transverse cross-sectional view is taken of anymaterial sample of Celesta 432F, the cross-section area of its recesses (as defined by Claim 1 of the Patent) will lie within a range of 1.3mm² or more and 4.8mm² or less.

The current state of the pleadings following the PTR is that the Claimants have pleaded as follows in §§4-5 of the Amended Particulars of Infringement:

4.

The Claimants rely on the 432F OFC [Sample Cable] as exemplifying the infringing Celesta Products as further set out in paragraph 5 below. If and to the extent that the Defendant contends that there are differences between the 432F OFC and other Celesta Products that are material to the issue on infringement of the Patent as sought to be unconditionally amended, it is requested to identify what they are.

5.

The 432F OFC is a product falling within the scope of protection of at least claims 1, 2, 4 and 5 of the Patent as sought to be unconditionally amended.

The Defendant responds as follows in §5 of its Re-Amended Defence and Counterclaim:

iv As to the second sentence of §4 PoC, the differences material to the issue of infringement between the 432F OFC and other Celesta Products containing fibre counts 144 to 576 will be addressed as necessary in the course of evidence and/or any PPD (in respect of the PPD, insofar as Sterlite has knowledge of the presence or absence of claim integers required by claims 1, 2, 4 and 5 of the Patent) in due course, including but not limited to the facts that (a) the 144F (variants 1 and 3), 288F (both variants), 432F and 576F Celesta Products are not suitable for insertion into a micro-duct by air-blowing, as required by the claims of the Patent as unconditionally sought to be amended, and (b) the diameters of the various types of Celesta Product differ from the Sample and 432F Celesta Product, meaning that all cross-sectional areas will be different as between those cables. Sterlite understands from paragraph 22 of the Second Witness Statement of Paul Harris dated 20 February 2024 that Fujikura’s pleaded case accuses dealings with only Celesta Products containing fibre counts 144 to 576 of infringement. Any other allegation of infringement must be properly pleaded in Fujikura’s PoI.

v §5 PoI is denied because the 432F OFC is not a product falling within the scope of protection of claims 1, 2, 4 or 5 of the Patent. Fujikura is required expressly to plead any other asserted claims and cannot rely on “at least” in §5 PoI as it offends CPR63PD4.1(1)(a). In the premises, Sterlite does not respond to Fujikura’s improperly pleaded assertions of infringement of other claims.

It can be seen that the Defendant expressly took points about micro-duct and diameter. It did not take a more general point that the Sample Cable was not representative of other 432F cables. It pleaded that any differences between the Sample Cable and any other cables in the Celesta range would be dealt with in the PPD or in expert evidence.

At trial the Defendant relied on the decision of Birss J in Electromagnetic Geoservices v Petroleum Geoservices [2016] EWHC 27 (Pat) [2016] FSR 25 at §21 for the proposition that the Claimants should be held to relying on the experiments only for the purpose set out in their notice:

…Part of the difficulty may have been highlighted by the submission of counsel that the most important fact to be admitted in the Notice was the first one, that when the procedure is followed, the results shown occur. Although Notices of Experiments routinely contain such a fact, to regard it as the most important is to misunderstand the function of that part of a Notice. That fact should normally be the least important fact to be admitted. CPR PD63 7.1 requires that a party seeking to establish any fact by experimental proof conducted for the purpose of litigation must (my emphasis) serve a notice “(1) stating the facts which the party seeks to establish”. This rule is vital for a number of reasons. It requires the party serving the notice to think precisely about what it is that the experiment is supposed to prove and what role it is to play in the case. It gives proper notice to the other party of what facts the experiment is supposed to prove. When the facts are properly set out it should also allow the court to understand what the experiment is for or at least should allow for that understanding to be arrived at fairly readily with further submissions. This in turn allows the court to retain control over the experimental evidence. The permission the court then gives to rely on the experimental evidence is permission to rely on it in order to prove those facts. If the party then changes its case and wishes to rely on the same experiment to prove something different, they may need permission to amend the Notice. That maintains some discipline in the proceedings and mitigates or avoids the problem which arose in Molnlycke Health Care AB v Brightwake Ltd [2011] EWHC 140 (Pat); [2011] F.S.R. 26 in which a party wanted to rely at trial on an electron micrograph which had been produced for one purpose for a very different purpose. Taking the approach that the expert’s reports explain what facts an experiment is supposed to prove in litigation is not adequate.

I have no problem with that principle. But one needs to look at the facts of that case too. In that case Birss J permitted the contents of the Notice of Experiments to be used to prove multiple parts of the case. His objection was to the order sought that the defendant be permitted to rely on the Notice without doing repeats. Birss J ordered that the claimant reply to the Notice, that repeats be conducted and that the claimant have permission to serve a notice in reply.

The circumstances of the present case are very different. The experiments have been repeated and witnessed. The Claimants are not trying to rely on the experiments per se to prove infringement of the other Celesta cables, including any other 432F cables. The experiments could never do that, as Bacon J pointed out. All the experiments can prove, and all the Claimants seek to prove with them, is that the Sample Cable has a recess area falling within the claims. There is then a separate exercise by which the Claimants take the recess area established for the Sample Cable, and allege that the Sample Cable is exemplary of 432F cables and/or representative of all Celesta cables, including scaling the recess area up or down using the information in the PPD and the ratios between the dies for that cable and the others alleged to infringe. These infringement issues are to be dealt with as a matter of evidence/on the PPD as the Defendant has pleaded.

The Claimants are not therefore using the experiments to prove infringement of the other cables, not even other 432F cables. That is why in my judgment the Claimants were correct to drop original fact (iii) from their Notice of Experiments. It is the plea that the Sample Cable exemplifies the other cables and/or the scaling exercise based on the PPD from the established recess area of the Sample Cable which is used to argue infringement of the remainder of the range. In response to that the Defendant is limited to running the points it has pleaded.

The Defendant has not pleaded that even if the Sample Cable infringes then other 432F cables it has sold do not. I have already dealt with the point above about the diameter of the 432F Sample Cable and the stated diameter of 432F cables in the PPD. I suspect this point gained added significance only after the Defendant abandoned its application to do “reply” experiments on other 432F samples before Bacon J and the outcome of the PTR refusing the amendment to the Defendant’s pleadings. It can be seen as an attempt to argue the same point a third time but I have rejected it on the evidence.

Of the points the Defendant has pleaded I have dealt with the micro-duct point as a matter of construction. The outstanding point is that cables with different fibre numbers to 432 will have different diameters and therefore different recess areas.

This takes me to the extrapolation exercise. The Claimants abandoned their case on the 576F cables in closing. In relation to the other cables, Mr Debban’s extrapolation using information about the die diameters together with an appropriate draw-down calculation based on the Sample Cable suggested that they too would all infringe. His evidence was that a 0.5mm difference in diameter produces only a 0.1mm² difference in recess cross-sectional area.

Mr Sutehall agreed in cross-examination that the scaling exercise was a reasonable one to have adopted. His only real criticism was to suggest that the calculations did not take into account the possible cooling effects following extrusion on different sized cables. Mr Debban considered that his draw-down calculations did take inherent account of this. However, even if there was anything in the point, there was no basis in the PPD to suggest that cooling effects would result in significantly different recess areas for different sized cables, as Mr Sutehall fairly accepted. Nor had he attempted to quantify the difference. So I reject this criticism of the extrapolation exercise.

The Defendant’s final riposte to the allegation of infringement across the range of Celesta cables was to accept that the Claimants’ expert Mr Debban may well have been right to assert that he expected that the recess area of many cables would fall within the range in the claim, regardless of the experimental evidence. The Defendant argued that this supported its case on obviousness, and the Patent could not be both valid and infringed.

But this only serves to highlight the problem the Defendant has found itself in. It was not suggested that there was any intermittently bonded OFC product on the market at the priority date which satisfied the claims. The only IBR cable in the cited prior art did not have ribs. The Defendant did not bring a positive case that any of its cables did not satisfy the claims. It sought instead to invalidate the Patent on grounds including lack of inventive step, but that attack has failed on the evidence.

It follows that the Patent is valid and infringed by the 432F Sample Cable. Based on this result there is also infringement by the Defendant’s other Celesta cables with fibre counts of 144, 288 and 432.