Separating the buzz from reality

You may not have believed it possible, but the phrase “interactive TV” has actually pulled ahead of multimedia in contention for Buzzword of the Decade.

Based on most of what you’ve seen written about interactive TV so far, you might assume that ITV takes multimedia technologies and puts them online, in a manner of speaking, allowing TV viewers to interact with programming in a slightly modified version of how people interact with their computers today. This may someday be true, and is in fact what some companies are banking on, but it certainly isn’t going to be here by Christmas.

In fact, since no architectural or regulatory framework exists for it yet, almost nothing about interactive TV is true today — except that a lot of companies have bet the farm that this vision of the future will come to pass. Even more than multimedia, interactive TV has become synonymous with big money, big players, big markets and a sunny future for us all based on ubiquitous digital technology.

Bringing order to chaos. Because so many have put so much at stake, the subject deserves deeper examination than it’s had to date. This two-part series will bring more of this complex subject to light and some much-needed structure to the discussion.

Part I, which you’re reading now, will attempt to define the various types of interactive TV as the phrase is used today. Part II, to be published in June, will discuss the deeper issues: the “private vs. public network” question, pay-per-view vs. open access to information, regulation and the “conduit vs. content” argument, security and privacy, using the HDTV model to set standards, deep-think about “dream applications” and commercial potential, and a nightmare scenario or two.

In addition, it will detail specific interactive TV efforts and deals, including the Time Warner Orlando project, Viacom’s Castro Valley project, the Microsoft-GI-Intel deal, the various combatants for the settop decoder market, and companies such as IBM and Silicon Graphics that are aiming to provide turnkey interactive systems.

Let’s hear from you, too. Since this topic is too big for any one person or industry to figure out alone, call or E-mail us with questions you’d like to see answered, companies, technologies and/or deals we need to know about, or your comments about where interactive TV is headed.

A caveat: announcements and deals are being made at such a furious pace that we cannot possibly include them all. We expect that this story will actually be a serial — or if past events are any indication, more likely a soap opera — that we will chronicle regularly.

INTERACTIVE TV IN FOUR LAYERS

Though the distinctions may be somewhat arbitrary based on how they’re implemented, interactive TV as it is discussed today mostly separates into four layers. At the bottom is what some like to call “fake interactive,” or dynamic broadcast television. The second is true interactive, analogous to multimedia as we know it today, and including true video on demand. Third is interactive services. Fourth is video telephony.

LEVEL I: DYNAMIC BROADCAST USES EXISTING PROGRAMMING

Fake interactive, or dynamic broadcast, is the only kind of so-called interactive TV that’s possible right now given today’s technology and regulatory infrastructure.

In its most basic implementation, programming remains in its current analog form. The illusion of interactivity is created by using a stream of control data running in parallel with the existing programming as it is delivered into the home.

A special unit, usually a controller box plus remote positioned atop the TV set, allows the viewer to dip into the data stream; based on viewer input, the control data can then manipulate or change what appears on the TV screen (thus the term “dynamic”).

This can be done in a variety of ways. In early 1992, the Federal Communications Commission under chairman Al Sikes, now head of new media and technology for the Hearst Corp. (and speaking at next month’s Digital World conference), allocated a section of the public airwaves for “interactive video and data services,” or IVDS.

Virginia-based TV Answer, an early interactive TV vendor that just signed a deal with the Public Broadcasting Service, uses IVDS radio signals as both feedback loop and the means to overlay information on existing programming. PBS will use the system to help its fundraising efforts, and to augment its educational programming (see “TV Answer begs the question,” Vol. 1, No. 9, p. 17).

Interactive Network Inc., based in Mountain View, CA (which just narrowly avoided financial disaster via a significant investment from cable giant TCI), uses radio signals for screen overlay and the existing phone network as the feedback loop for its system.

Direct broadcast satellite (DBS) systems in development today use a standard phone line as their feedback loop (though they will deliver digital video). Other cable-based systems are experimenting with using the coaxial cable itself to send control data upstream, much like input from a keyboard controls the screen of computers today.

The dynamic broadcast paradigm, no matter how it’s implemented, is perfectly suited to the first required application for interactive TV, and one that could make it wildly popular right out of the chute: an online TV programming guide that’s able not only to show previews and synopses, but is also likely to sport an easy VCR programming feature that does all the dirty work.

Beyond program guides, most systems in this paradigm are using the control data to provide screen overlays for sports and game-show play-along, for song lyrics atop music videos, or to allow viewers to request information from participating advertisers.

Where’s the revenue? Although vendors claim wild viewer enthusiasm for these systems (we’re skeptical), it remains to be seen how providing sports statistics or song lyrics, or even the chance to play along with Vanna White or Joe Montana, might drive a significant revenue stream.

An equally critical caveat is that intellectual property laws forbid screen overlays without express permission of the rights holder, generally barring the door for new product categories based on creative use of overlays. The potential for rights holders to demand a king’s ransom for an overlay are likely to make these types of applications more expensive than viewer demand warrants.

PRYING OPEN THE PAY-PER-VIEW MARKET

Moving to digital video delivery changes the scenario significantly. Today’s capability to fit six digital channels into the same bandwidth as one analog channel (with compression technology only expected to get better) will create more lucrative early opportunities for dynamic broadcast TV beyond the need for TV program guides.

The convenience offered by these primitive systems is likely to significantly open the pay-per-view market. Though the technology is not yet in place for true video on demand, channel compression will allow service providers to start a movie every five minutes if they control sufficient channel capacity.

Good for home shopping. Unlike today’s pay-per-view systems that require customers to place a phone call, dynamic broadcast systems can use their various types of response loops to do online billing authorization via remote control, sans human intervention, thus opening a new level of shopping and transaction-based services not before possible within a single medium. (The security and privacy issues raised by such transactions will be addressed in Part II.)

Some rudimentary multimedia-type programming may be possible with a dynamic broadcast system as well, using what’s called “streaming interactivity,” which uses channel compression to simultaneously broadcast a number of streams of related programming from which the viewer can select.

This concept was the basis for Warner New Media’s Megillah project a couple of years back (see Vol. 2, No. 6, p. 16). In one implementation, parallel streams of video were used to allow viewers to dip in and out of different camera angles of a live basketball game and get statistics on certain players; the game in progress was always available at a touch of the remote.

Another prototype was a discography on the band Fleetwood Mac, again using “live” action at a concert as the central stream. Fans could dip into parallel video tracks of different band members being interviewed about their history, high points of other concerts, best-selling records — more information than anyone is likely ever to want about Fleetwood Mac, but a fascinating approach to interacting with existing programming.

Similarly, the Vidéoway interactive system from Canadian cable system Groupe Vidéotron demonstrated an interactive news broadcast where viewers had the choice of watching a broadcast in its normal, serial fashion from start to finish, or using the remote to select a more indepth report on one of the headline stories of the day. The settop box acts as the traffic coordinator for which story stream the viewer selects.

Dynamic broadcast, especially using IVDS, may see a fair bit of action during the next couple of years, especially from terrestrial TV broadcasters who want to provide some degree of interactive programming without moving to a fully digital system, and who aren’t able or willing to use the cable network with its built-in, high-bandwidth response loop. (Don’t forget: cable may “pass” 95-plus percent of the homes in the U.S., but only 60 percent subscribe.) It’s a relatively cheap and easy technology to implement, and the FCC has already allocated bandwidth to enable these types of services.

LEVEL II: MULTIMEDIA AND VIDEO ON DEMAND

The next layer of interactive TV can be called “true interactivity,” and will enable viewers to initiate simple requests to view discrete media objects and receive responses based on those requests. This provides access to true video on demand, or VOD, and to online multimedia and/or video games. Unlike dynamic broadcast, which requires no investment in infrastructure (i.e., fiber or coax) and very little participation by existing broadcasters or cable system operators, achieving true interactivity requires investment and planning.

First of all, VOD and multimedia/games are bandwidth pigs and whether the service provider is a telephone or a cable company, they are very likely to require installation of some new conduit — either fiber from the cable headend to the neighborhood or home, or a new copper or fiber wire in the phone network — in order to provide adequate performance (“adequate” being significantly better than CD-ROM titles today).

Transaction processing. If games and movies will be pay-per-view and not subscription services (which is likely), they also require sophisticated online transaction processing systems, or OLTP, like those that banks, retailers and phone companies rely on today.

OLTP provides not only the meter, but also conducts the authentication process (i.e., I am who I say I am, and I can pay) and authorizes transactions between the service provider and the customer. Not insignificantly, OLTP systems also provide what’s known as “fault tolerance” to keep the network up and running under virtually any conditions.

As this is an all-digital TV system, a settop box will be necessary both to decode and decompress video and audio signals in real time, as well as to provide the process control for paging through online TV guides, interacting with multimedia titles and playing games. If vendors are serious about making the online game market a real one, then the settop box must also be able to render interactive 3D graphics in real time to provide realistic game play to the Nintendo generation.

Virtual VCR. It is in Level II of the interactive TV system that the “virtual VCR” concept can be implemented. Not only will viewers be able to pause, rewind or fast-forward a “broadcast” movie, but they’ll also be able to do what one service provider calls “reverse time shifting,” or paying extra for redelivery of programs they missed. (Hopefully this will also include yesterday’s broadcast of the “Kenneth Cole Shoe Hour” on the shopping network; more on that in Part II). Reverse time shifting could prove to be a financial bonanza for broadcasters who charge a premium to those of us who are too busy, lazy or forgetful to preprogram our VCRs.

LEVEL III: INTERACTIVE SERVICES, COMPLETE VIEWER CONTROL

Although VOD provides a new means of delivering (and charging for) existing, discrete entertainment titles and products, one of the most captivating promises of interactive TV is the transformation of the TV set into a window on the expanding world of information.

Ride ‘em, cowboy. In much the same way that online jockeys can find nearly any person or piece of information they want via online archives and information services, the viewer in this next level of interactive TV will be fully in charge of what shows up on the screen.

By the time this day has arrived, the number of channels delivered via cable, phone lines, satellite and cellular will likely number in the thousands and we will no longer be able, let alone willing, to get the information we need by consuming information passively.

An information profile. With the possible exception of watching headline news of the day, we’re likely to build and store profiles of the specific kinds of information we want to see delivered onscreen — what news topics we want to track, new movie and music releases, when and where certain items are on sale, who’s in town for book signings or concerts, etc.

A FUNDAMENTAL SHIFT IN APPLICATION OF TECHNOLOGY

This is a significant step past video on demand, as interactive services of this nature are much more akin to information publishing than they are to entertainment. (If you’re looking for a reason that Microsoft cofounder Paul Allen is interested in America Online, you just found it.)

In Level II, the “true interactive” system, it’s feasible that much of today’s delivery model would remain intact: virtually all the on-demand programming will originate with the cable service provider, who either owns the programming outright or has negotiated a royalty agreement with the rights holder.

As there are relatively few rights holders providing mass-market entertainment — Sony, MCA/Universal, Time Warner, Paramount, etc. — and all have existing distribution deals with cable operators, the only technology that really has to be implemented for Level II TV is an effective metering system (as outlined above) to track who is watching/playing what.

Creating a monster. But Level III’s provision of customized information products to individual consumers is a far more formidable task. In order to deliver information fast enough for consumer satisfaction, Level III TV would require the deployment of thousands or even millions of information servers (in some models, these information servers will serve as this decade’s analog to the personal computer software business in the 1980s), operating over a high-bandwidth, extraordinarily fast, fault-tolerant, open international network.

It would be highly unlikely that a search-and-retrieval system of this capability would be confined to one transport medium; in other words, this isn’t likely to be a “cable” or a “phone” network. What’s more likely is that consumers will subscribe to a service and expect that it be delivered over any existing network — cable, telephone, cellular, personal communication networks, direct broadcast satellite, or whatever else develops.

Expanding choices. Why? For one reason, entertainment and news programming, which is at the core of TV viewing today, will become only one of the services delivered “through the TV,” so to speak. Plenty of people will take advantage of the convenience of the television as information service center. But as a ubiquitous digital information network becomes more firmly entrenched, customers won’t care where the service originates, and will demand their information on whatever network serves the communications device at hand, whether that happens to be a powerful computer sitting atop (or built inside of) the TV set, a personal digital assistant, a desktop or laptop computer or a portable game machine.

REQUIREMENTS FOR FULL ON-DEMAND ARE AWESOME

The requirements to set up and maintain such a system are both costly and technically awesome:

• A standard architecture for full interconnection and interoperability between cable, wireless and telephone networks to deliver information.
• A standard for interoperability between televisions and computers, independent of display type and screen resolution.
• A standard for digital video compression that is open and not controlled by any system vendor.
• A standard method for querying the network for information.
• Highly sophisticated OLTP systems for billing, user authorization and authentication.
• Crackproof data security and integrity, to prevent theft of intellectual property and the misappropriation of customer information.
• Serious virus inoculation software (you think it’s bad now, just wait).
• A sophisticated data-tracking structure for billing and service inquiries.
• Agents to search out, purchase and retrieve information without human intervention.
• Storage and/or printing facilities for “time shift” viewing of information offline.
• “Media servers” with vast storage capacities.

LEVEL IV: THE FINAL FRONTIER — VIDEO TELEPHONY

With a network of these proportions in place, the final frontier for interactive TV is the capability to turn the television network(s), whatever they happen to be, into a fully symmetrical, two-way link: in other words, telephony and/or video teleconferencing.

It ain’t necessarily so. The three largest cable companies — TCI, Time Warner and Viacom — have made no secret of the fact that, as TW’s Geoff Holmes said at a recent conference, they believe they’ll be “the next phone company.” Although it’s not hard to believe the cable industry’s ability to plow forward in many other areas, telephony is a bit shakier, since developing a good telephone network is a nontrivial technical challenge; and telephony is, for many good reasons, a massively regulated business.

Operating under what’s left of their regulatory yokes, the regional Bell companies are working fast to get up to speed on video delivery, so they can provide better video telephones and start to sell quality videoconferencing services to businesses. Their job is relatively easy: get more bandwidth via whatever means necessary — fiber optics, better compression, etc. — so they can add video to the two-way data stream they’re already adept at transmitting and switching.

THERE’S A LOT TO KNOW ABOUT TWO-WAY DATACOM

The cable companies seem to believe that they can buy technology solutions to these problems and be in business providing full, two-way interactive services. They don’t yet know that telephony is a far more complex industry than theirs of broadcasting one-way, analog information. Money can buy expertise but it can’t buy culture, and cable’s lack of telephony culture might prove more expensive than the industry expects.

The cost of infrastructure. First of all, never underestimate the cost of infrastructure. Time Warner’s big-time digital cable system that’s being deployed in Orlando is fiber-based, which gives it plenty of bandwidth for video telephony. But two-way connectivity — especially of multiple data types — doesn’t come cheap.

Orlando will supposedly use AT&T’s asynchronous transfer mode (ATM) switches, which can handle two-way transmission of video, audio, text, voice, etc. This is exciting, as everyone believes that ATM is definitely the way of the future, but there are plenty of reports that ATM products aren’t quite ready for prime time even within the telco community. Anyone who remembers what happened with the Signaling System 7 software a few years back — a bug took down an entire long-distance telephone network for hours at a time — will realize that cost is not always monetary. One or two network crashes, when alternative services exist, are likely to dampen cable’s ardor for a fully switched, digital telephony network.

A LEGENDARY LACK OF THE SERVICE ETHIC

But the sorest point of cable’s transition to a telephone system is customer service, its lack of which is legend. Grumble as many of us do about the amount of hoop-jumping it takes to get a simple PBX installed, the telephone system in this country is remarkably reliable both in provision and reliability of service, and in its ability to bill customers accurately for services rendered.

This kind of reliability is the result of decades of infrastructure building and problem solving for analog voice services, and more than a decade in the switched digital data market. Cable has experience only in unswitched, uni-directional, analog video, and thus is fooling itself if it believes it can “be the phone company” any time soon. Expect to see some competition for local service using wireless equipment, but cable isn’t likely to be a strong contender for telephony — video or otherwise — for quite a while.

IT’S ACTUALLY NOT INTERACTIVE TV AT ALL

So it comes down to this: What’s been passing in the national press for “interactive TV” isn’t really TV at all. Once we move past Level II, the notion of TV as we know it today becomes almost quaint. With the entry of interactive services, we are really talking more about a ubiquitous digital information network than simply about moving pictures. In this context, the “TV” actually diminishes in importance, becoming one of many in an expanding universe of communications devices and information receptors.

With that in mind, we encourage you to broaden the way you think about interactive TV. As you’ll read in Part II, there are many compelling reasons to be sure that no single network or industry is allowed to corner the market, especially not during its gestation period. We need to be inclusive, not exclusive, and shape a network that takes advantage of the strengths of cable, of phone companies, of cellular, of satellite, and that provides access to as many users and entrepreneurs as possible. If we do it right, we can say we helped shape the most significant advancement ever in human communication. Let’s not squander the opportunity.

Denise Caruso

Next month: Time Warner’s Orlando project, Viacom in Castro Valley, what Silicon Graphics is up to, the various combatants for the settop box market, the commercial and social potential (both negative and positive) of digital interactive TV, and more. Contact dcaruso on MCI Mail, AppleLink or America Online with input, responses, requests, etc.