Play...Fast Forward...Rewind...Pause U.S. Firms Want to Wire America for Two-Way Tv, But Their Systems Are Not Yet Ready for Prime

By Philip Elmer-DeWitt

You're watching the movie Dracula on Bell Atlantic's Stargazer video-on- demand service, and it's working like a charm. It's part of the phone company's two-way interactive-TV system, which is designed to demonstrate how tomorrow's couch potatoes will order movies and TV shows from a menu of listings and make them start and stop whenever they want.

But hold on. Suddenly the doorbell rings, or you want to go to the bathroom. What now? Unfortunately, Stargazer's pause button is still under construction. To stop a movie in midscene, you have to pick up a telephone handset, dial into the Stargazer control center, punch in a series of codes, and wait as long as half a minute for the order to be processed.

This is hardly the smooth navigating you've been led to expect on the so- called information superhighway. Neither Stargazer's shortcomings, however, nor the collapse of Bell Atlantic's proposed merger with Tele-Communications Inc. in March seems to have dampened the enthusiasm of the company -- or the rest of the telecommunications industry -- for the superhighway venture. Stargazer, which expects to serve up to 1,000 customers by early next year, is just one of four Bell Atlantic pilot projects, and it joins several dozen similar tests being patched together across the U.S.

The faith behind these trials is being sorely tested this spring. Most of them are running behind schedule, many don't work at all, and none are ready for prime time. Not only is the basic technology snarled, but the road ahead is cluttered with legal and regulatory obstacles. Says Mitch Ratcliffe, editor of the newsletter Digital Media: "I think consumers are going to be unimpressed for a decade or more."

The interactive-TV pilot projects play a critical role in the race to build the communications networks of the future. It is here that the rhetoric about the info highway comes face-to-face with the realities of engineering. And what the trials have been showing, say industry observers, is that the top executives who commissioned them have run ahead of their engineers. "It's a constant theme," says Steve Krause, a technology expert at SRI International. "When these projects finally get down to the implementation level, reality hits."

Reality seems to be hitting all over. The TV trial that Viacom was supposed to launch in Castro Valley, California, next month has been postponed until the end of the year. Ameritech, which had ambitious plans to deliver interactive services to 200,000 homes in the Midwest by year's end, is waiting for government approval. US West says it is still set to bring video to 60,000 telephone customers in Omaha, Nebraska, by December but admits that "some of the more speculative video services" may have to wait.

No company has risked more to bridge the gap between promise and reality than Time Warner. Early last year the company announced with great fanfare the launch of an ambitious experiment called the Full Service Network. Starting in April 1994, according to the original schedule, an area serving 4,000 Time Warner Cable subscribers in Orlando, Florida, would begin to enjoy the benefits of interactive TV: hundreds of movies and top-rated TV shows delivered when the viewers wanted them, video games they could play by themselves or with friends on the other side of town, and video malls where they could shop for goods and services by remote control.

April has come and gone, and those Orlandans are still watching TV the old- fashioned way. "Expectations were way overheated," admits Jim Chiddix, Time Warner Cable's technology chief. In early March, the start of the trial ! was postponed until the fourth quarter of 1994. Some outside observers predict the company will be hard pressed to make that deadline as well.

Time Warner's frustrations may be greater because its ambitions are higher. Its Orlando project was designed to be a fully functional model of a system that could eventually be rolled out across the country. So-called market trials, by contrast, are often little more than electronic Potemkin villages patched together on a personal computer or run from a back room by people doing the work by hand.

Other pilot projects are primarily technology trials. In GTE's TV trial in Cerritos, California, for example, engineers tested a video-on-demand system using VCRs preloaded with tapes that allowed viewers to pause, fast-forward and rewind from remote locations. But to keep the technology and cost manageable, the service was made available only to two homes and a pair of elementary schools.

The ultimate goal of all these trials is to build what engineers call a switched, broadband network. From the consumers' point of view, this is basically a TV set connected to something that works like the phone system. The wires in this network have to be fat enough (in terms of information capacity, or bandwidth) to carry TV signals. The network must also have switches and software flexible enough to allow movies to be shuttled back and forth without a break in the action, even if thousands of viewers want to see them at different times. And it must have screens sharp enough to display text as well as video.

All these technologies exist, but they have never been put together. In order to assemble the Orlando trial, Time Warner had to sign up more than half a dozen outside vendors, some of them bitter rivals. For example, Silicon Graphics, which is building and programming the huge disk-drive systems known as video servers, and AT&T, which is making the network's switches, are competing elsewhere in the race to create the best video servers. The danger is that if the two companies decide in the future they cannot share key proprietary technology on the Orlando project, Time Warner may find that its system components are incompatible.

Storage space on the disk drives could also pose a real problem. Time Warner engineers have had to buy more and more computer-disk storage to handle all the new programs and services that the company wants to offer. At first the engineers thought they could store hundreds of movies on a so-called terabyte file server (which can hold the equivalent of 1 million floppy disks) and still have room to spare. Now they realize they will need a server twice that big. Although the prices of these components will eventually come down, right now the start-up cost of Time Warner's Orlando project is reported to be about $5,000 a household. Says Gary Arlen, a cable-TV consultant based in Bethesda, Maryland: "You need to sell a lot of showings of Wayne's World to justify that capital expenditure per site."

Difficult as it is to get the hardware to work, the software may be even more perplexing. Creating the system that will yoke together a bunch of Orlandos, each designed by a different cable-TV or telephone company, is expected to be a nightmare. Even more critical, say some experts, is the software that presents viewers with a menu of offerings. If the new TV controls are much more cumbersome than channel dials or up-and-down buttons, viewers may simply refuse to use them.

Even after the technological problems are ironed out, legal impediments will remain. One of the toughest has to do with syndication rights. It's one thing to talk about making TV sitcoms like Murphy Brown and Seinfeld available to customers who missed them at their regular broadcast hours, but quite another to negotiate a deal with the Hollywood agents and unions that represent the shows' creators. "How are you going to compensate the actors, the directors, the grips?" asks Ratcliffe. "Lucille Ball still earns money for her shows, and she's been dead for years."

THE GOVERNMENT'S ROLE IS STILL A mystery. Left to their own devices, cable operators are likely to concentrate on the richest neighborhoods of the most densely populated regions -- a practice known as cream skimming. If they were required, as phone companies are, to provide service to everybody, their development plans might look very different. So far the Clinton Administration has concentrated on getting schools, libraries and hospitals hooked up and has said little about how to bring the technology to the rest of the country.

What really matters, of course, is what people want to see on their TV sets and how much they are willing to pay. The preliminary results in this regard have not been encouraging. Video-on-demand is billed as the best reason to get interactive television. But TCI reports that subscribers using its video-on- demand test trial in Colorado are ordering an average of 2.5 movies a | month. That's probably fewer videos than many U.S. families rent each week.

Some analysts think that by focusing on TV viewers, the companies building these systems may have overlooked a more promising market: the millions of computer users who are already playing games, exchanging mail and entertaining themselves on the computer networks. Although a switched, broadband network could serve both computer users and television viewers, cable-TV operators in particular seem reluctant to allow computer owners to plug in. The cable operators, contends Michael Godwin of the Electronic Frontier Foundation, a public-interest group involved in electronic communications issues, "have a couch-potato vision of the future."

One player who has a deeply vested interest in interactive services is Bill Gates, chairman of Microsoft, the largest personal-computer software company. Not surprisingly, Gates has dismissed the current crop of TV trials as "crummy" and "embarrassing." Says he: "There's not a single pilot that's been done that justifies the huge financial stake I and others have just intuitively decided to invest in this." Gates remains convinced that interactive TV is inevitable, however. He just wants it to appear on devices running his company's operating system, and he has 500 employees working on making that happen.

But the computer networks may be no more ready than the TV networks to handle the freight of the information superhighway. Today's personal computers are too low-powered -- and the modems that connect them to the phone lines too slow -- to transmit and process video signals in real time, as they are broadcast. Even if everybody were to replace their PCs with the new, more powerful models coming into the market, someone would still have to build an electronic highway fast and wide enough to carry the traffic.

Once that switched, broadband network is built, it won't matter much what people plug into it -- TVs, PCs or some device that hasn't been invented. Like that of the telephone system before it, the power of the information highway will come from the new ways it allows people to connect, not with machines but with each other. And for that privilege, even the most stubborn couch potato might agree to get wired.

With reporting by David S. Jackson/San Francisco and Suneel Ratan/Washington