Rovi Guides, Inc v Virgin Media Ltd & Ors

This is an appeal from a judgment of Mr John Baldwin QC sitting as a deputy judge of the Chancery Division in the Patents Court, and from his consequent order revoking European Patent (UK) 1,327,209 (“the patent”). When the matter was before him, the registered proprietor of the patent was United Video Properties, Inc. and that company together with Rovi Solutions Corporation were the claimants in the action. By a subsequent consent order Rovi Solutions Corporation has ceased to be a party to this appeal and United Video Properties, Inc has merged into Rovi Guides, Inc which is now the registered proprietor and sole appellant. In this judgment I will use the term Rovi to refer to Rovi Guides, Inc and its predecessor.

At the trial Rovi alleged that the defendants (together “Virgin”) had infringed the patent. In the course of the proceedings Rovi made what is described as an unconditional application to amend the claims of the patent. This meant that the validity of the patent stood or fell with the validity of the amended claims. Where I refer to a claim in this judgment I am to be taken as referring to the proposed amended claims.

“Video on demand” or “VOD” refers to the ability of a user to play requested films or other recorded content at a time requested by him. “Live TV” or “Live feed” refers to ordinary broadcast or cable TV services where the time of transmission is chosen by the provider. The invention which the patent protected was concerned with allowing a user to stop watching VOD or live TV and resume watching elsewhere, for example in a different room, on a different apparatus. This feature or “functionality” was referred to as “relocation”.

The judge had to consider two categories of claim in the patent: the “VOD relocation claims” and the “live feed relocation claims”. The VOD relocation claims were alleged to be infringed by Virgin. All the VOD relocation claims were held invalid by the judge for obviousness over a single item of prior art referred to as DAVIC (Footnote: ¹) and common general knowledge, and there is no appeal from his conclusion in this respect. The consequence of this finding of invalidity is, of course, that Virgin do not infringe the patent.

Virgin also maintained that the live feed relocation claims were invalid. The judge held that these claims were invalid because they were similarly obvious over DAVIC and common general knowledge.

Rovi is dissatisfied with the judge’s conclusion that their live feed relocation claims are invalid, and appeals to this court. The judge himself gave permission to appeal.

The court has the power to revoke a patent on the application of any person if the invention to which it relates is not a patentable invention (section 72(1)(a) Patents Act 1977 “the Act”). One of the grounds on which an invention is not patentable is if it does not involve an inventive step (section 1(1)(b) of the Act). Section 3 of the Act provides:

“An invention shall be taken to involve an inventive step if it would not be obvious to a person skilled in the art, having regard to any matter which forms part of the state of the art…”

The courts have advocated a structured approach to decision-making on questions of obviousness. In Pozzoli SpA v BDMO SA [2007] FSR 872 at [72] Jacob LJ, with whom Mummery and Keene LJJ agreed, set out the following steps as being involved in the exercise:

“(1)

(a) Identify the notional person skilled in the art; (b) Identify the relevant common general knowledge of that person;

(2)

Identify the inventive concept of the claim in question or if that cannot readily be done, construe it;

(3)

Identify what, if any, differences exist between the matter cited as forming part of the “state of the art” and the inventive concept of the claim or the claim as construed;

(4)

Viewed without any knowledge of the alleged invention as claimed, do those differences constitute steps which would have been obvious to the person skilled in the art or do they require any degree of invention?”

The first three steps set up the necessary mental scaffolding required to approach the only statutory question, which is included in the fourth step. Also included in the fourth step is a warning against the dangers of hindsight and “reconstruction a posteriori”, emphasised in many authorities: see e.g. per Moulton LJ in British Westinghouse v Braulik [1910] 27 RPC 209 at 230. Also very well known is Lord Diplock’s warning in Technograph Printed Circuits v Mills & Rockley (Electronics) [1972] RPC 346, 362 about “step by step” cross-examination of the patentee’s expert. The question for the court is not whether the invention looks, or can by skilful cross-examination be made to look, obvious today, given one’s knowledge of what the invention is and any additional wisdom acquired since the priority date of the patent. The question is what was obvious to a person skilled in the art at the priority date of the patent.

Whether an invention, viewed in this way, does or does not involve an inventive step is an evaluative judgment based on a number of different factors. Kitchin J (as he then was) put it in this way in Generics v Lundbeck [2007] RPC 32 at [74] in a passage which has received subsequent approval at the highest level:

“The question of obviousness must be considered on the facts of each case. The court must consider the weight to be attached to any particular factor in the light of all the relevant circumstances. These may include such matters as the motive to find a solution to the problem the patent addresses, the number and extent of the possible avenues of research, the effort involved in pursuing them and the expectation of success.”

That then is what needs to be done. If a trial judge has followed this approach and made no error of fact or principle, then this court will be in the highest degree reluctant to interfere with his conclusions. In Biogen v Medeva [1997] RPC 1 at [45] Lord Hoffmann explained the rationale behind this approach and concluded:

“Where the application of a legal standard such as negligence or obviousness involves no question of principle but is simply a matter of degree, an appellate court should be very cautious in differing from the judge's evaluation.”

The invention is concerned with set-top boxes or STBs and in particular with allowing the use of multiple STBs, for example within the same household, so as to allow content which is being viewed on one STB to be paused and resumed on another, for example in a different location in the house.

VOD describes a service where the service provider maintains copies of films on its server which the customer can purchase for a fee. When the customer purchases a film he can use the film as if he were operating a video cassette recorder, in the sense that he can pause, re-wind and fast forward the film. So far as VOD is concerned the patent claims, in effect, a VOD service which has a relocate feature, so that the customer can pause the film and resume watching in his home or elsewhere. This invention is the subject of the VOD relocation claims, which are obvious.

The patent also describes and claims the idea of using this relocate feature with a live television feed. As described, when a viewer requests to pause his viewing of a live television program which he is watching, the system will make a recording of the program going forwards from that point. The recording will then be available for viewing later by that viewer, without missing any of it, via a second STB. This invention is the subject of the live feed relocation claims, which the judge held to be obvious.

The only relevant claim, claim 22, is a lengthy one. It contains features which demand that the system is configured to allow the user to “freeze delivery” of a media on demand service (which term is used, slightly unusually, as including live TV). The media on demand service is delivered from a remote server (referred to in the industry as “the head-end”). When the user requests the delivery to be frozen the system must allow delivery to be resumed from a second STB. The important feature for the purposes of this appeal is that which was referred to in argument as “feature J” and reads as follows:

“if the media is a live feed the remote media-on demand server begins recording the media when the request to freeze delivery is received from the first [STB].”

In summary therefore when the user presses “pause” or “relocate” whilst watching live TV, the claimed system begins recording. It is important to note two things. Firstly, the recording is required to take place at the head end, not on a digital medium within the STB. Secondly the fact that recording begins only when the user presses pause means that a re-wind to an earlier point in the programme will not be possible. It will only be possible to re-wind to the point at which the pause button is pressed.

There was no dispute that this is what claim 22 meant.

The judge recorded that it was common ground that the skilled person, the person to whom the patent is addressed, was a person or a team working in the cable, satellite, terrestrial TV or telecoms industries and interested in the development of VOD systems. The judge also recorded that the patent was “an ideas patent” in that the invention lay in having the ideas prescribed in the various claims, not in any perceived difficulty of enablement once one had had those ideas. He then made some careful findings about the common general knowledge:

“12.1

the way that cable systems, OTA [over the air] broadcast systems and telecoms systems operated to deliver content. Thus:

12.1.1

cable networks deliver content to subscribers over channels in a private network, with content being delivered simultaneously to all STBs in the network from a central transmission facility, referred to as the head-end. It was common for networks to include a 'return path' so that each STB could communicate with the head end and this enabled interactive functions such as VOD. The content of any channel may contain entitlement messages which, although delivered like all content to all STBs in the network, are such that only a particular STB is capable of decoding data transmitted on that channel. Such messages enable the cable provider to stream, e.g. a film, down a channel on the network and restrict the ability to view the film to a single STB, being that one where the subscriber has requested and paid for the film. Although there is no actual point to point connection in a cable network, the use of entitlement messages permits what was referred to in evidence as a logical point to point transmission of content.

12.1.2

OTA broadcast systems deliver content over the air to viewers with antennae. Content may be free to air or encrypted, in which case the viewer needs a decryption device. There is no return path unless one is provided in another way, e.g. by a phone line, in which case the systems work in a similar way to cable.

12.1.3

Telecoms systems deliver content via copper pair cable or fibre optics. The infrastructure was set up as a telephone network and two-way connectivity is established via point to point wire and switches. That infrastructure was known to be suitable for transmitting television content. A VOD programme is delivered on a single point to point connection that is set up between the provider and the STB requesting the content.

12.2

VOD and other interactive services could be provided over cable networks, and that, in a VOD system, the recorded material is stored at the head-end. Individual STBs can access that stored material and the user can play, pause, rewind etc., as with a VCR.

12.3

Storage at the head-end in a VOD cable system necessitates the head-end being able to identify the point at which content began to be recorded, was paused or is to be resumed.

12.4

The use of subscriber management software at the head-end to cater for multiple cable STB households.

12.5

In connection with new functionality proposed to be added to a cable system, there was a debate in the industry as to whether the new aspects should be added to the user's STB or, centrally, at the head end. This debate was well known, with proponents on either side.

12.6

Servers in networked systems were useful for providing services to users. In the context of interactive VOD services, server requirements were generally split between an audio-visual streaming device (storing and streaming the content) and a controller server (managing the interaction with the user to enable other functionalities such as selection of content).

12.7

The concepts of layered software and distributed systems in service-providing systems. Thus system requirements were split into software layers (categorising requirements into common operation levels so that they were more easily identifiable) which enabled service providers to provide services, including VOD services, to customers in a more cost effective and standardised way.

12.8

The concept of sessions which could exist at various layers. The main session, or user session, is the one created when a user starts to use a service on a VOD system and closes when the user finishes using the service. A subsidiary session or connection session is set up which consists of the network connection between the user equipment and the service source^[. The two different sessions are not necessarily set up and closed down at the same time, the possibility existing for a user session to be maintained after a connection session has been closed.

12.9

Personal Video Recorders (PVR), how they worked and the functionality they provided. This included time shifting, pausing live TV and later resuming from the same point, and rewinding live TV for replay. The notion of pausing live TV was a hot topic at the priority date.

12.10

Cable system providers were in competition with PVR providers and were interested in offering PVR functionality to their customers. The industry debate previously mentioned included a debate as to whether it was preferable to record content and provide functionality from the head-end or a local STB.”

Mr Abrahams had a quibble about the judge’s paragraph 12.3. But this is restricted to VOD and the DAVIC prior art clearly teaches that, in VOD, the service provider can keep track of what the user is doing, including bookmarking the point at which the user “left” a movie so that he can re-start at that exact point.

The judge heard expert evidence from two experts, a Mr Hoarty for Rovi and Mr Kerr for Virgin. Both experts were well qualified to give expert evidence on the issues facing the judge, but the judge found Mr Kerr to be a “very impressive and knowledgeable witness who was able to explain concepts clearly and succinctly”, whereas Mr Hoarty “was not very good at explaining things and not very good at answering questions”. Mr Hoarty had also given inconsistent answers (which the judge proceeded to exemplify), and gave the judge the impression that he was arguing or asserting Rovi’s position rather than trying to explain the reasons for his views.

The judge next considered what the disclosure of DAVIC would have revealed to the skilled person. At [40] he concluded that DAVIC teaches the relocation feature of the patent. Given that the VOD claims were built upon a conventional VOD service or immaterial variations of it, this finding was fatal to the VOD relocation claims.

The judge then turned to the live feed relocation claims. The judge’s reasoning on these claims is contained in paragraphs 41 to 45 of his judgment. I summarise his reasoning as follows:

The grounds of appeal start from the proposition that the step from DAVIC to the invention of the patent required four steps. These were: (1) deciding to add PVR functionality to their system, including live pause; (2) deciding to do that at the head-end; (3) deciding to provide live pause by commencing recording when pause is pressed and (4) deciding to enable relocation and to use the live pause functionality for the purpose of the relocation. With this in mind the grounds of appeal can be summarised as follows:

Mr James Abrahams who appeared on the appeal for Rovi with Ms Isabel Jamal, accepted the judge’s findings about the common general knowledge, subject to the quibble I have mentioned. However he emphasised one aspect, namely that the known PVR devices produced by Tivo and Replay Networks operated as follows. The STB contained a hard drive on which programmes could be recorded. When the user tuned to a channel to watch live TV the STB would automatically start recording that channel – it did not wait until the user pressed pause. The user could press “pause”, which would pause the live TV, just as they could pause a video tape. However, in contrast to the invention of the patent, these known devices offered a facility to re-wind to the point at which the channel was selected, and not merely to the point at which the user pressed “pause”. This facility is described as “instant replay”. It allows the user not only to pause and come back where he left off viewing, but also to review the immediately preceding sections of the broadcast. So a football fan can instantly replay a goal, or someone watching a drama can replay a passage of dialogue which they did not hear or understand.

Taking DAVIC as the starting point, it was important to recognise that the delayed broadcast disclosure related to the recording of whole programmes. Indeed the ability to have whole programmes available to other users was the, or the main, motivation for providing the functionality at the head-end.

With these matters in mind, Mr Abrahams accepts that step 1, the step of deciding to implement PVR functionality and live pause was obvious. PVR functionality also included, however, instant replay, and that required recording from (at the latest) the moment the user tunes to a channel.

As to step 2, Mr Abrahams accepts that he cannot now challenge the judge’s finding that it was obvious to put PVR functionality at the head-end. However, the motivation for doing this was to allow multiple users to access a single recording (in contrast to using the STB where it would be available to one user only). It followed from this that the whole programme must be recorded.

Mr Abrahams submits that, in these circumstances step 3 (decide to implement recording by pressing pause) cannot be obvious, because it goes directly contrary to steps 1 and 2 once the motivation for those steps is considered:

Given the judge’s findings about the comparative qualities of the expert witnesses, Mr Abrahams wisely relied solely on the evidence of Mr Kerr, Virgin’s expert. In support of the point made at (c) in the previous paragraph he drew our attention to an article by Dr Walter S. Ciciora entitled “V.O.D. is H.O.T.!” exhibited by Mr Kerr, and in which he reports on a meeting of consulting telecoms engineers at around the priority date. The article includes the following:

“The second hottest topic at the meeting was the significance of the Personal Video Recorder (PVR) for cable. The two main brands of PVRs are TIVO and Replay TV. Two views emerged on how to best apply this concept to cable. One view advocated adding PVR functionality to the cable settop box. Another view held that the hardware investment should be in the head end and not in the home (sound familiar?).

Since massive hard drives have become so inexpensive, PVR functionality could be provided by cable using VOD technology. Subscribers could be rented capacity on the bank of hard drives to store the programs they find interesting. Because multiple users can access the same copy, the amount of storage is substantially less than would be required in cable set-top boxes or in consumer-owned PVRs. Total hard drive capacity investment would be less if it was shared at the headend than if it was installed in each of the subscriber locations. Consumers would be pleased to avoid another set-top box and its costs. As hard drive technology continues its advance, consumers would keep pace without the worry of obsolescence. This approach can be both a revenue opportunity and another sustainable competitive advantage over other multichannel video providers who have no choice but to put the PVR functionality in the set-top box.”

Mr Abrahams emphasises the part of this article which speaks of multiple users renting the same copy.

Mr Abrahams also submitted that the judge had failed to conduct a proper Pozzoli analysis. Whilst such an analysis is not compulsory, the failure to conduct it had led the judge into error. In considering step 3 the judge had overlooked feature J and failed to consider the implications of the logic which had enabled Virgin to overcome steps 1 and 2. He went as far as to submit that the judge’s conclusion was one which no reasonable judge could reach.

Mr James Mellor QC who appeared with Mr Andrew Lykiardopoulos QC for Virgin answers these points as follows:

Mr Abrahams devoted little if any of his argument to step 4, the relocation feature. He was right to do so given that DAVIC’s teaching about this feature is generic, and one would expect any functionality incorporated at the head end to be capable of being made available at a second location. It is of course essential to consider the totality of the claimed invention when assessing obviousness, but I do not think that the re-location feature could save the live feed relocation claims if feature J is held to be obvious.

I therefore consider first the question of whether it is contrary to DAVIC’s teaching to record anything other than whole programmes. The relevant applications in DAVIC are movies on demand, broadcast (or live TV) and delayed broadcast. The suggested system which DAVIC describes does not require the implementation of all applications, or indeed all features within each application. In section 9, concerned with movies on demand, DAVIC teaches that the user has “a network-delivered service that offers the functionality of the home VCR”. This includes amongst all the other features, pause, fast forward and fast re-wind. Under the broadcast application there is no mention of recording. Under the delayed broadcast application (i) the user selects a program to be broadcast later, or (ii) the user can inspect what programs are stored on the server even if they were not ordered by that user, or (iii) the user can request a start at any time after a recording has begun. This last facility, means that if a user arrives late for a program he wants to watch, provided it is being recorded, he can ask to watch from an earlier time, for example the beginning of the programme, even if he has made no request.

Mr Mellor accepts that in the specific applications discussed in DAVIC it would be absurd to record anything other than whole programmes. But his case was not based on a skilled person slavishly implementing delayed broadcast. His case was that the skilled person would consider implementing “pause live TV” in a DAVIC system. The fact that the specific applications taught in DAVIC required whole programmes to be recorded did not mean that in other contexts they would have to be as well. He pointed out that whilst Mr Kerr had accepted that whole programmes were necessary for the specific applications, he did “not see anything within this [i.e. DAVIC] which says that fragments of programmes are not recorded”.

I do not think, therefore, that recording parts of programmes is contrary to the teaching of DAVIC for these purposes.

I consider next whether Mr Abrahams is right that there is an inherent contradiction between the motivation for providing functionality at the head-end and the patent’s feature J. Mr Abrahams took us to the cross-examination of Mr Kerr in which this contradiction was put to him. Mr Kerr accepted that one lost the benefit of being able to share programmes between users if the system only started recording when the user pressed pause. But Mr Kerr was not prepared to go the further mile with Mr Abrahams and accept that there would be no purpose in providing this functionality at the head-end. Thus he would not accept that a system which recorded at the STB was the easiest way to implement a system which recorded only fragments. It was one way to do it, but required hardware to do it in every STB. They were both technical options.

Moreover there were other arguments of which the skilled person would be aware as to why things should be done at the head-end. The Ciciora article mentions some of these, such as the ability to have personalised capacity on the head-end server for users’ personal use. Mr Mellor drew our attention to other evidence before the judge, such as Mr Kerr’s evidence on how keen the service providers were to keep the profit margins for themselves. Mr Hoarty accepted in cross-examination that one of the advantages of placing PVR technology in the head-end was that upgrading it was cheaper than upgrading thousands of STBs. This passage of cross-examination is no doubt why the judge recorded at paragraph 21 of his judgment that there “were good arguments for locating the functionality at the head-end”. They were plainly not limited to the provision of whole programmes for sharing by other users.

I would therefore reject Mr Abrahams’ argument that there is an inherent contradiction between the motivation for providing functionality at the head-end and the patent’s feature J, which requires recordings of parts of programs. There was a lively and informed debate as to where the functionality was to be provided. Sharing whole programmes was but one argument in the overall controversy.

Mr Abrahams’ most powerful point was that a skilled person considering incorporating PVR functionality into the head-end would want to incorporate full PVR functionality. This would include not merely “pause live TV” but also instant replay. Following that logic would not lead the skilled person to feature J, because that feature requires recording to begin only when pause is pressed. It rules out instant replay.

Mr Abrahams pointed out that neither Mr Kerr nor Virgin’s advisors had initially appreciated how the digital systems on the market which offered PVR functionality, TiVo and Replay Networks, actually worked. They had assumed, incorrectly, that those systems started to record when pause was pressed. If they had been correct on that point, one can see that an argument to incorporate that functionality in the same way into a system which used the head-end would be a very powerful one indeed. Mr Abrahams submitted that, without that assistance from the common general knowledge, the argument showed all the hallmarks of hindsight. The idea of the pause button being used to initiate recording was not to be found in the prior art. Mr Kerr’s evidence in his witness statement that the feature J approach was technically preferable was therefore given on a false basis.

We were taken at length by Mr Abrahams through the cross-examination of Mr Kerr. In addition, because Virgin had indicated by a respondent’s notice that they intended to rely on the evidence of Mr Hoarty, we were also taken in some detail by both counsel through his cross-examination as well. I hope I will be forgiven if I do not repeat the process in this judgment, because it seems to me that the overall effect of Mr Kerr’s evidence was that he held and thereafter maintained the view that “pause live TV” was an independent and more important aspect of the possible functionality than instant replay or re-wind. His evidence is well summarised in the following answer:

“The reason why I am stating this is because I honestly did not remember, and these things were of interest to me in 1999, the fact that they [i.e. TiVo and Replay Networks] recorded all the time even when you were not asking for it. The one liner was “Hit pause on live TV”. It was not “Go back on live TV”.”

Mr Abrahams asked us to dismiss this and other answers on the grounds that they were hindsight. I do not see how we could do so. The judge was entitled to rely on this evidence.

I am not quite so convinced that the judge could have drawn much support on this precise point from the cross-examination of Mr Hoarty, given that Mr Hoarty was very resistant to the idea that the skilled person would have thought of recording only from the press of the pause button at all. He did however accept in the end that the point from which one recorded had a direct impact on the resources necessary to implement a feature. That, as it seems to me, would militate in favour of a minimalist approach, i.e. recording from the press of the pause button.

I would conclude that the judge had material to enable him to hold that the implementation of a system with feature J was obvious.

Finally I would turn to Mr Abrahams’ criticisms of the judgment. It is true that the judge does not spell out in as much detail as I have done the reasons for rejecting the arguments which Mr Abrahams presented to him and which he has repeated on this appeal. But I am not persuaded that his judgment is wrong. The judge was plainly aware of the Pozzoli guidelines. Mr Kerr’s evidence on these claims had been structured by reference to those guidelines, particularly pointing out the difference introduced by feature J. Whatever initial errors had been made by Mr Kerr, the judge was plainly aware how the TiVo and Replay networks machines worked. He recognised that the excitement they created was, in particular, about pause live TV.

It is of course necessary to guard constantly against the dangers of hindsight. The judge carefully considered Mr Abrahams’ argument that if all this had been obvious functionality to include at the head-end, the highly skilled industry experts who wrote DAVIC would have been expected to mention it. The judge recorded however that the art had moved on in the period between the writing of DAVIC and the priority date. By then “pause live TV” was a must have feature. That provided a good enough explanation for DAVIC’s silence on this topic.

Most importantly of all, the judge was immersed in these and other matters throughout a lengthy trial spanning seven days. Although a judge must state his conclusions and brief reasons for reaching his conclusion, he is not required to deal with every argument and evidential dispute which is debated in the course of the trial: see per Lewison LJ in FAGE UK Limited and another v Chobani UK Limited and another [2014] EWCA Civ 5 at [114] to [117]. Having heard the arguments on this appeal, I am not persuaded that the judge made any error which it would be open to this court to correct.

I would accordingly dismiss the appeal.

I agree.

I agree also.