TRG's Commentary

ITV Infrastructure
Mechanics and Machinations

by Peter Cassidy, Eric Johansson and Rodney Thayer

This analysis was originally published in the April 2001 Digital Mogul Industry Report.

The development of an ITV infrastructure is one of the biggest technological proposals ever tendered. Think of it this way: the biggest interoperable machine on the planet is the telephone network. Through slow development, copious industrial cooperation through industrial standards groups and industry-state collusion and, most recently, deregulation and competition, the telephone network was born and raised a wholly interoperable medium with an attractive return on investment (ROI) proposition.

Our television infrastructure is the largest one-way electronic communications system in the world. Due to its provenance as a broadcast medium, it has been developed on a transfer pricing model (ad dollars for eyes) and technologically cultivated as a passive medium, some would argue, to gently feed commercial messages to somnambulatory, acquiescent viewers.

Naked television is useless for two-way communications. Before the ITV infrastructure matures, tens of billions will be expended in rolling out two-way wireline links to the home, in advanced digital Set-Top Boxes (STB), in two-way transmission media, in head end equipment, in home hardware, in ITV middleware and development tools by cable & satellite operators (MSOs) - in technology bought and installed by the consumer.

Fig.1

The ITV model has been stillborn many times - but not because the technology has failed. In fact, with every attempt, ITV has made great strides. Time Warner's QUBE system developed in Columbus, Ohio in the 1970s offered teletext, gaming - and even guitar lessons. In the mid-1980s, Knight-Ridder offered Viewtron that featured a kind of Minitel service, transmitting news and text-based entertainment through users' television sets. Time Warner, in the early 1990s, set up its Full Service Network (FSN) in Orlando, giving subscribers a full interactive menu, including TV shopping, news, interactive gaming, VOD and Pay-per-View. All were suspended because the MSO could not develop sufficient revenue to earn a return on the technology investment.

The Orlando trials by Time Warner have become, in some ways, the Waterloo of ITV deployments, a benchmark showing what is possible in an ITV service platform, but also provoking some discussion of what would be a reasonable deployment cost per household. The Orlando trial actually used a STB based on the Silicon Graphics Indy, a powerful workstation-level machine. At the time of the trial, estimates were that the in-home technology cost would add up to about $5000 per household. (This doesn't even include equipment and modifications at the head end.)

Today, the per-household figure being floated by some industry insiders and analysts is somewhere in the neighborhood of $500 to make ITV a viable proposition. Possible? Much more so today than ten years ago, when anyone rolling out an ITV platform would have had to build a lot of the technology from scratch - like a moon mission. But now there are competitive markets for media servers and middleware technologies, and for almost all of the technologies required for ITV deployment - save the STB which are still ruled by the Scientific Atlanta-Motorola duopoly. Standards bodies are rolling out specifications for many kinds of devices opening up competition even wider. These developments, coupled with the general trend toward falling prices of video, computing and transmission technology, conspire to proffer ITV platforms that can animate a viable business model. More like building a cinema complex or throwing up a local telephone central office than engineering a moon shot.

As proof of its development, ITV has already begun to speciate into rather well defined product/technology categories. Today, there are three primary genera, though they overlap in some basic functionality. The traditional terrestrial wireline MSOs, like AT&T Broadband, Time Warner, Charter and Cox, are leveraging their large networks to homes to provide enhanced entertainment packages as well as Internet applications such as e-mail, web browsing and instant messaging. DBS-based ITV systems like UltimateTV by Microsoft that combines DirecTV programming with the Internet functionality of its sun-setting WebTV service suite. Finally, the Personal Video Recorder (PVR) products that are rolling out now under TiVo and ReplayTV's brands are intriguing wildcards. While sold as sophisticated programmable recording devices, these machines are, under the hood, simply PCs, devices that can be - and are, reportedly, being turned into online video servers. The Napsterization of video, when it happens, may well be engendered by a television device.

Big Players Face - and Make -
Big Tech and Content Integration Issues

Some industry experts estimate the upgrade from basic cable service to full-blown ITV capability will entail a $70 billion-plus upgrade of U.S. cable systems. Essentially, it is the entertainment media equivalent of tearing down the telephone infrastructure and replacing it with videophones.

DBSes, terrestrially wired MSOs and PVR services all incur their own costs and security exposures for the service provider, content developers and for the consumer. Products in each of these operator categories support their own subscription and purchasing models for gear.

As well, rollout scenarios map closely to the technology's acquisition model and are driven, to a certain extent, by the ITV service providers alliances for content and connectivity, relationships that are shifting all the time. The biggest players in ITV have already made their moves, organizing technology platforms and content and applications plays. The establishment of standards in ITV technologies and the cultivation of ITV applications will depend largely on the power politics of the ITV operators, content companies, electronics companies and software/middleware providers.

AT&T Broadband has the largest wireline cable television network, the largest number of subscribers and ITV-enabled viewers. Meanwhile, though commanding a powerful buying position by dint of its size, the company needs to rationalize a network of many different technologies acquired when it bought out big cable concerns like MediaOne and TCI. In the latest iteration of its rollout plans, AT&T's was designing a system based on a Motorola DCT-5000 STB with Microsoft's WinCE operating system (OS), a Java-based engine from Sun Microsystems, TV Guide's ITV guide, a portal that would be the "first screen" from which subscribers would launch their viewing sessions. The industrial politics of this proposal are intriguing, given that an essential component of the infrastructure, Java, has been the subject of a long-running law suit between Sun Microsystems, Inc. and Microsoft.

Fig.2

ALO Time Warner is in the position of rationalizing its own pure ITV play with AOLTV, which brings AOL on-line services to TV viewers. Time Warner has been steadily building out its ITV infrastructure, planning to have its entire cable TV network ITV-enabled by year-end. Its focus thus far has been on VOD, building on its content catalogue, which includes Turner Classic Movies, HBO and Cinemax. With AOLTV, there are enormous opportunities to create blended programming experiences - yet the integration challenges are numerous. In particular, AOL's hugely popular instant messaging features can make viewing a sort of entertainment conference call that is fun. To get there, AOLTV will have to be engaged in two-way communications with the Scientific-Atlanta STBs that Time Warner uses and integrated with the PowerTV OS that controls it. (The company is also rolling out a satellite version by re-badging DirecTV's service.) Here the industrial and technical politics are intriguing but no less clear. AOL Time Warner is an investor in TiVo and the conglomerate has announced that it is developing a new STB that will bind TiVo functionality into it. DirecTV already has its own TiVo-enabled unit on the market.

Though the expansion of ITV through terrestrial cable continues apace, the technology that is making some surprising headway as a medium of choice for ITV services is DBS. The DBS subscriber base grew substantially last year, gaining around 3 million subscribers - up 29% from the year earlier.

Fig.3

Terrestrial cable now has 80% of the pay-TV market, down from 82% last year, while DBS now commands growing 15.4%. Some forty metropolitan markets now receive service from one or both DBS operators, with DirecTV active in thirty-eight markets, EchoStar in thirty-four and both competing in some thirty-two metropolitan markets. This expanding footprint has become a part of the strategies of many ITV competitors - including AOL Time Warner itself, which is developing its own DBS play with DirecTV.

Microsoft's UltimateTV, for example, has no terrestrial cable options at all. It rides a re-badged DirecTV service and bundles in a TiVo recorder and WebTV Internet application services to give the subscribers a hybrid system capable of a number of viewing profiles: passive watching, programmable time-shifting and Interactive programming (using the WebTV hooks embedded in some programming from Microsoft's partner NBC, such as Wheel of Fortune and Jeopardy). UltimateTV is the single most aggressive entry of the company into consumer electronics products, placing it, in many ways, head-to-head with AOL Time Warner. Though pricey, between the $450 hardware costs and subscription fees that can extend to more than $100.

Clearly, as far as carrier media go, Microsoft is laying its bet with DBS (though hedging through its investment in AT&T.) The company has sound reasons besides DBS' growth over the past year. DBS operators are shipping STBs into a fast growing segment - increasing from 15.7 million units in 1999 to 50 million by 2004, according to Cahners In-Stat Group. The new generation of DBS STBs allow for a great many new features (such as Internet applications) as well as doubling the number of content channels that can be maintained via new modulation and transmission standards.

The most surprising interloper in the ITV space, however, is the PVR. Two years ago, they were not even a category. Today there are more than 150,000 PVRs installed in the marketplace in the United States and Europe. Some are ReplayTV (which recently stopped selling to consumers in favor of licensing its technology), but the majority are TiVo-branded boxes. Jupiter Media Metrix predicts that by year's end, 2.2 million households will have a DVR, and by 2005, there will be 42 million households with the product. It is expected that PVRs will represent a mass market ITV platform by 2003, when 16.8% of U.S. households, or about 18 million, are expected to have one. A large part of ITV's appeal is its flexibility within a limited, yet attractive and familiar, application profile.

Conventional TV, ITV
and the Intrusion of the Digital Network Packet

Part of conventional television's inflexibility and the assumption of audience passivity lies in its technical origins, based in radio. It wasn't until after World War II that RCA (Radio Corporation of America) began mass production of television sets that allowed the medium to become a true part of the popular culture.

The genesis of cable television was in necessity. In the late 1940s, service operators rolled out coaxial networks to provide clear signals to television viewers out of range of the metropolitan broadcast areas. Soon hundreds of these so-called community antenna TV (CATV) operators followed.

About ten years ago, as fiber costs started to drop, cable operators began to install fiber closer to the home, greatly improving signal reliability, which is critical for two-way interactivity. As prices continued to drop, operators replaced co-axial cable with fiber, allowing for substantial digital traffic. The most ambitious cable operators soon realized there was a great deal more opportunity for profit-making on their customers than simply re-broadcasting content from distant stations.

The emergence of the Internet was attended and accelerated by crashing connectivity costs and an entirely new, flexible communications technology. A lot of the applications we see rolling out today on ITV are iterations of Internet applications for the ITV platform. Key to animating those applications is preserving a lot of the efficiencies delivered by common protocols and technical specifications that define the Internet. Some of that has been transferred for components of systems like, for example, UltimateTV, that can stand alone as a terminal to process e-mail and browse the Web.

Still, integrating applications like instant messaging and chat into workaday television programming presents the complications of technology and industrial alliances. If the boxes and subscription services won't interoperate easily - or at all - due to technology issues or industrial politics (say, you can't IM someone off your platform), those features may go unused and, with them, opportunities to engage the viewer. The same dynamic that opened up even AOL to the public Internet will open up the different ITV platforms over time. Still, the imposition of interoperability will create the channel by which content will flow around the larger ITV infrastructure.

For the nonce, viewers will have to make choices that will seal them into one or another fiefdom. Still, at the outset, the menu of services will be roughly consistent across most MSOs, with most serving up an Electronic Program Guide (EPG) and VOD services. A number of tech-intensive ITV applications will be rolled out over the next few years, at first, maintained by the MSOs themselves. Over time, as those successful prove durable revenue bearing applications, hosted services will no doubt spout to service the ITV operators. The most promising of these applications, in terms of providing true interactivity are as follows

VOD

(Video-on-Demand) and nVOD (near Video-on-Demand). What people in Orlando did enjoy in the FSN trial: movies when you want to see them. Vendors developing servers to store, manage and stream movies for ITV operators include Diva Systems, Concurrent Computer Corp., Seachange International, nCube, Intertainer and Alex Temex Multimedia.

Interactive Program Enhancements

It's TV with a Web page interpolated over it. Televised content is married to interactive applications that the viewer can call down and use. The best example out there is NBC's embedding of HTML hooks into the Vertical Blanking Interval for Jeopardy and Wheel of Fortune that WebTV units read, sending subscribers questions to effect a play-along version of the show.

Web Browsing

Who doesn't want it? Big players in this space are companies like WorldGate, Liberate and Microsoft. Aside from the STB and operating system issues, a mature application that can be readily deployed and leveraged, the ITV operators' first-screen advantage, offering sponsored and revenue-sharing links.

Fig.4

T-Commerce

An ITV variation on retail e-commerce. Subscribers can click on deals and buy things that are being advertised or placed in a program. See sweater on Britney. Click on sweater on Britney. Buy Britney sweater. There are a number of companies already developing these technologies and many more will be coming over from the Internet side as soon as revenue streams begin to look attractive. Big players in this space include companies like CommerceTV, RespondTV and Wink, and, among others, they are advancing this application.

The Walled Garden

Basically, it's the initial screen, offering sponsored services that are populated with sponsored services, among them e-mail, news, weather and local and national sports. Big players in programming Walled Garden services are Liberate, OpenTV and MicrosoftTV.

The cost of even making programming interactive can be a bear, depending on what the MSO wants to support. In effect, the MSO, with a full menu of ITV, effectively assumes not only the role of cable operator, but also an ISP and a transaction processor like any retailer or franchise. Much of this can be outsourced. But even then, these are tough businesses on their own, much less rolled into a single enterprise that has to be remotely managed.

Middleware Infrastructure Issues
from the Parlor on out to the Head End

Development content for television or for the Web are enterprises whose technical requirements are well known. Developing ITV content, delivering it, animating it on all this new hardware is going to become a complex and costly affair, given the competing standards and platforms. Cable television technologies we've grown up with in the parlor didn't change for a good long time. Four years ago, it was a big deal for the industry to add a little memory to STBs to facilitate the decoding of MPEG-2.z

Now, everyone subscribing to digital cable is going to be supplied with the equivalent of a powerful desktop PC to plop on top of their television. The new STBs have lots of processing power, memory, multiple tuners for recording and viewing, modems, special graphics chips, compression accelerators, hard-drives for storage and, in some proposals, slots for smart cards. Running all this hardware are real-time operating systems for STBs, the largest players for which are PowerTV, a division of STB maker Scientific Atlanta, Mentor Graphics, maker of VRTX, and Microsoft with its WindowsCE.

Only then we get to the applications layer, where the interactive programming has to be executing, have its hooks out to different devices as well as have viewer input coordinated. Here, middleware coordinates the viewing content with the interactive content, applications and viewing choices that define the experiences available to the viewers. Players to watch in this space are OpenTV, Liberate, Canal+, Wink and WorldGate Communications. Middleware is the tether that joins the programming and the technology that activate the interactive experience.

Fig.5

The proprietary nature of middleware technologies make necessary a great deal of expensive re-purposing of ITV content. To a certain extent, the ITV technology platform looks a lot like the PC world did in the mid-1980s, when the offerings from a number of makers were all running their own operating systems: PC-DOS, AppleDOS, CP/M and AmigaDOS. Other than the cost of supporting all these different technologies is the deeper trouble of interoperability between the hardware and from there, the software that is being developed for the different platforms. Unless this is addressed early on, viewers might end up suffering service outages in their ITV experiences. In the PC world, we call them crashes. To head off that eventuality, the ITV industry has begun organizing open standards to define technical specifications that will facilitate interoperability.

Cable Labs' OpenCable consortium has released finalizing its own specification - the Open Cable Application Platform (OCAP). OCAP defines a core presentation engine based on an HTML browser with extensions for interactive television. OCAP splits the project into an executable engine and a presentation engine. The former is for animating applications like clickable ads and other interactive applications. The latter uses HTML to map the applications being run to the proper positions on the screen. Cable Labs is trying to balance all the interests of the behemoths involved.

In part, the OCAP is trying to make up for some of the perceived shortcoming of the ATVEF (Advanced Television Enhancement Forum) specification, in particular, the omission of Java. JavaTV is an integral part of the OCAP specification. ATVEF left out Java, it is widely believed, because of Microsoft's power over the group and its deathless desire to limit the proliferation of Java and staunch the influence of its arch nemesis, Sun Microsystems.

If these standards bodies can exact the kind of authority that, say, the ITU exerts in the telephony space, or the IETF exercises in Internet technologies, there may be some hope for reduced interoperability friction when these systems are deployed. Those bodies only became powerful as they evolved with their industries. It seems much more likely that the initial ITV rollouts will be micro-fiefdoms beset with their own internal technical unrest. Coming to a living room near you: the Blue Screens of Death.

Disruptive Tech on the Way to ITVLand?

Of all the technologies defining the emerging ITV infrastructure, PVRs exhibit many of the characteristics of a disruptive technology. It can be controlled almost entirely by the consumer, it is open enough for experimentation and it is relatively inexpensive. A TiVo, when you get its cover off, is a tarted-up Linux-based computer. What PCs did for music through MP3s, TiVo boxes and its progeny could well do for digitized video.

Enthusiast engineers and third-party technology companies are providing hacks and gear to open these boxes up and join them to TCP/IP networks, essentially turning them into video servers that can stream content over the Internet. One company, 9th Tee, is already shipping a TiVoNET Ethernet Network Adapter Board. This product and the competitors that will follow are the equivalent of the CD-ROMs mounted in PCs in the mid-1990s that put music in the control of the user: the technology that closed the loop between a content genre and the larger Internet. With music, the PC was already networked, but needed a content acquisition device. With TiVO, the content acquisition mechanism is in place. Ethernet adapters give it the pipe to the rest of the world.

TiVo and its contemporaries also exemplify the "convergence" at work and the Faustian bargain it entails. Hack a TiVo with the right connectors, mount a keyboard and screen and you can call up a command prompt for Linux, ready for exploration. Toss in an Ethernet card and you're ready to roll content out onto a network - or the Internet.

Yet, understand this: TCP/IP network protocols and functionality, as well as storage capabilities, are creeping into most all of these home ITV technologies. Even the latest versions of the workaday STBs being introduced by Scientific Atlanta are mounted with Internet Protocol technology, modems and copious storage. The bottom line is that ITV gives content providers and broadcasters powerful new tools to engage the viewer - at the same time removing control of content from them and placing it in the hands of consumers. And we know what those busy hands did for music.

Wrapping Up

The art of using ITV technologies to capture viewers will be to create truly new experiences that exploit the technology to its fullest and within the context of the users' expectations to keep their attentions in real-time. Really fast, browse-able statistics and on-command replays for baseball viewers. Downloadable, viewable pages of conductors' scores that arrive in synch with the music during broadcasts of symphony performances. We have no doubt that we'll be amazed by the new content offerings.

The science of deploying ITV, however, will be in rolling out truly cost-recoverable technologies (no mini-computers sitting on the tube, thank you) that maintain control of the user experience to preserve the transfer-pricing model while, at the same time, animating other kinds of subscription, purchase and licensing models that will have to be developed to head-off the Napsterization of video content.

Systems that provide no alternatives to zap commercials out of the environment will leave broadcasters stranded. They have to deliver eye-balls to their advertisers. Engaging interactivity itself may help preserve that model. If a program becomes an event that viewers participate in directly - or through chats and instant messaging - the real-time necessity of being present and view the commercials will have a better chance of being preserved.

Interactive television's stickiness qualities have already been observed. Ed Graczyk, director of marketing for MicrosoftTV, told Electronic Media magazine this year that, "We have seen with WebTV that people who participate in interactive programming develop a deeper relationship with that programming."

Now all we need are business models that justify the costs and a sincere move towards interoperability, and maybe something grand will happen.

The authors are analysts and consultants with the international consulting collaborative TRG. Mr. Cassidy is TRG's director of research; Mr. Johansson, Director of Communications Infrastructure Development; and Mr. Thayer, Director of Information Security Architecture Consulting.