Thirty-five years ago this month, Japan’s Ministry for International Trade and Industry announced plans to build the world’s first fibered city. A two-way fiber-optic network called the Highly Interactive Optical Visual Information (Hi-OVIS) was far ahead of its time.
Hi-OVIS mobile system set up to broadcast a school event on a playing field.
Modern optical communications was born in a visionary era. Electronic technology grew explosively after World War II. Television sets went from novelties to commonplace in the 1950s, and, by the birth of the laser in 1960, Bell Labs had visions of transforming the telephone network to video. Social visionaries hoped that new communications technology would transform society.
Yet many visionaries complained that television had not lived up to its potential. In May 1961, the 35-year-old chairman of the Federal Communications Commission, Newton Minow, told the National Association of Broadcasters that commercial television had become a "vast wasteland." He didn't condemn all programs, but he said that too many of them were bad.
He was far from alone. When writer Isaac Asimov wrote about the tremendous potential bandwidth of laser communications in the August 1962 issue of The Magazine of Fantasy & Science Fiction, he mused: "Imagine what the keen minds of our entertainment industry could do if they realized they had a hundred million channels into which they could funnel new and undreamed-of varieties of trash. Maybe we should stop right now!"
Nobody stopped, of course, but people did look for alternatives. Television in the early 1960s was almost entirely a broadcast medium, with only three or four channels in most metropolitan areas. Only a few areas had cable, and those were "community antenna television" (CATV) systems, which used large antennas to pick up weak broadcast signals and distribute them locally. Promoters claimed CATV could offer more diverse programming, but skeptics scoffed at the idea of "pay television."
Dreams of wired cities
In the mid-1960s, a new vision for the future of telecommunications began emerging—the "wired city." Looking back, the concept was both bold and naive—linking homes, schools, libraries, businesses and factories with advanced communications to create a futuristic technological utopia. The wired city didn't get off the ground in the 1970s. Today we have the Internet.
Two-way video on Hi-OVIS home system.
The wired city concept evolved from Lyndon Johnson's "Great Society," which sought to revitalize troubled cities or build new ones. It was more of a social experiment than a technology demonstration. Its major goal was interactivity—to build local communities, facilitate public discussions, and let people "talk back to their televisions" rather than merely watch passively. Telephone service would be included, perhaps video instead of merely voice. Another idea was integrating meter reading, fire and burglar alarms, and remote controls for thermostats and ovens. By the early 1970s, visionaries also imagined that people would have access to news and information files stored on mainframe computers.
The U.S. government gave wired cities a couple of early boosts. In 1972, the FCC ruled that future large cable-TV networks should be two-way. That is, they should be able to transmit signals both to and from homes, although the technology had yet to be demonstrated. In 1974, the National Science Foundation sponsored two-way cable experiments in three small cities. However, the biggest experiment in two-way cable was a private venture—Warner Communications QUBE system in Columbus, Ohio. All those systems delivered multiple video channels to subscribers through branching tree networks of coaxial cable.
Fiber was still very new, but in 1972 Corning demonstrated germanium-doped fiber that was durable and had loss of only four decibels per kilometer. That led GTE Laboratories engineer John Fulenwider to suggest installing fiber in a communications system for a proposed new town development to avoid the need to bring 2,000 bulky coaxial cables together at central facilities. He proposed transmitting two color television channels, FM radio signals, digital data, two voice telephone lines and a single video telephone signal to homes through a total of three fibers using LEDs. A single return fiber would carry alarm and control signals and the return voice and video telephone signals. He described his fiber design at the International Wire and Cable Symposium in December 1972, but the developer went for a cheaper coax option.
In the end, the U.S. wired city projects faded away, and the FCC abandoned its two-way transmission requirement. Costs were high, political priorities shifted, the economy slumped, and the public showed little interest in two-way service. But the idea of the wired city survived overseas, particularly in Japan.
Hi-OVIS opening day ceremony.
The birth of Hi-OVIS
Two powerful Japanese ministries approached wired cities with different agendas. Seeking to promote Japanese industry, the Ministry for International Trade and Industry decided in 1971 that development of a "visual information system" could draw on Japan's strength in electronics and help the country move into information technology. The Ministry of Posts and Telecommunications focused on the traditional telephone industry and the nascent cable television industry. The two ministries began planning separate coax-based wired city demonstrations.
MPT designed and built a two-way tree-structured coaxial cable network for its wired city. Completed in January 1976, it served some 500 households in Tama until October 1980.
MITI created the Visual Information System Development Association in May 1972, partnering with a group of electronics and financial companies, and the following year it produced an ambitious catalog of 20 goals. When the oil crisis pinched budgets, VISDA narrowed its goals to television retransmission, a local video program service, video-on-demand, and data requests. Masahiro Kawahata was named VISDA managing director and chief engineer, and a site was chosen in Hikashi-Ikoma, an upscale new town near Osaka.
Meanwhile, MITI began sponsoring work on fiber-optic communications in 1973. The agency hoped to establish Japan as a major force in a young technology industry that did not require importing large amounts of expensive raw materials, like copper for coaxial cables. In March 1976, VISDA switched to fiber for its wired city, which became the Highly Interactive Optical Visual Information System, Hi-OVIS.
The installation of fiber in Hi-OVIS was a daring step. Hi-OVIS needed about 44 km of cable containing some 400 km of fiber to serve terminals in 158 homes and 10 public buildings such as schools and town offices. Picking fiber for Hi-OVIS avoided the cost of amplifiers that would have been needed on coax, but gambled that fiber would prove durable enough for outside installation. Bell Labs was in the midst of testing whether fiber hardware could withstand field conditions in ducts under a parking lot at its Norcross, Ga., plant.
Emperor Hirohoto tests the Hi-OVIS system.
With diode-laser reliability still questionable, Hi-OVIS engineers chose LED transmitters. That, in turn, led to a choice of fibers that may seem surprising today—step-index multimode plastic-clad silica with a 150-µm core, 350-µm cladding, and a coating that raised its outer diameter to 700 µm. Attenuation was less than 10 dB/km at the 830 nm LED wavelength.
That technology was less advanced than Bell Labs had chosen for the interoffice trunk system it was developing—graded-index multimode fiber with 50-µm core and 3 dB/km loss at 850 nm. Bell preferred gallium-arsenide diode laser transmitters but reserved LEDs as a backup. The cutting-edge of fiber transmission was moving to longer wavelengths. In 1976, Masaharu Horiguchi of Nippon Telegraph and Telephone and Hiroshi Osani of Fujikura Cable measured loss of 0.47 dB/km at 1.2 µm and then discovered the lower-loss 1.55 µm window.
However, Hi-OVIS did not need cutting-edge technology. Bell's goal was transmitting 45 megabits per second through 10 km of fiber. The longest link in Hi-OVIS was to a broadcast television antenna 4 km from the operation center; terminals ranged from 370 m to 2.27 km away, with the average fiber link 700 to 800 m. The required bandwidth was only 10 megahertz, with each downstream fiber carrying only one video and one audio channel in a 6-MHz band, and upstream fibers carrying the same service upstream, plus a 200 bit-per-second channel at 6.6 MHz for customer requests and control signals.
Installation of Hi-OVIS fiber-optic cable.
The Hi-OVIS network was a switched star configuration, with two fibers running between the operation center and each terminal. Each household selected one of the available program channels, which the operation center switched to the fiber serving that household. Branch cables strung overhead on poles that each contained 36 fibers and that split into subscriber cables with two fibers in them.
The switched-star configuration allowed Hi-OVIS to send any one of 30 input signals available at the control center to any of the 168 terminals on the network. Most programs could be transmitted to multiple subscribers simultaneously. That differed from the branched tree used in cable systems, which sends the same group of multiple video signals to many subscribers, who pick which signal to view on their set-top boxes.
The step-index multimode fibers used in Hi-OVIS couldn't match the bandwidth of the coax used in branched tree networks, but they were smaller and did not require amplifiers to span the length of the system. More important, the switched fiber network promised better support for the social experiments that were the key goals of Hi-OVIS.
A visionary project
Although fiber-optic technology was an essential part of Hi-OVIS, it was really a means to achieve social goals such as those of American wired city projects. In a series of articles and reports, Hi-OVIS managing director Masahiro Kawahata laid out four specific social goals with a visionary's enthusiasm:
"Establishing a new community in which the public can participate of their own accord," aiding each other and building a strong local community.
"Lifelong education" through face-to-face two-way telecommunications, helping participants to continue learning long after they finished formal schooling.
"Forming the safe local welfare society," to enhance the health and safety of the public, particularly children, the handicapped and the elderly. The hope was that the communication system would help build a community of people helping each other, enhancing medical care and fire prevention, and deterring crime.
"Establishing the initiative in the selection of information," helping people cope with the overwhelming flood of information so they could "regain independent thinking and creativity."
Kawahata at Hi-OVIS.
Kawahata's objectives show concern that television was becoming too powerful and too passivating a medium, drawing people away from the local community and leaving them isolated in their own homes. Hi-OVIS sought to change television so that it strengthened social bonds within local communities and served the public as more than a source of passive entertainment. His objectives are phrased in terms of Japanese society, but his concerns clearly paralleled those of wired city developers in the United States.
Realizing these objectives required a large investment in hardware for homes and the operation center, and in staff for the operation center.
Every subscriber terminal required a video camera, a microphone, a special control keyboard, and control electronics, as well as a color television that served as a display. To save money, homes received only black-and-white video cameras. Subscribers could enter requests on the 32-key terminal or contact the control center using the microphone and video camera.
The operation center included equipment for retransmitting nine broadcast television channels, six of them local and three distant, like an ordinary cable system, except that subscribers picked one for transmission through the fiber to their home.
More novel were information services that responded to subscriber requests. One generated still pictures assembled from images stored on microfiche and computer-generated characters. The microfiche provided reference information such as first-aid treatment, dictionary definitions and train schedules. The computer-generated characters reported changing information such as news and traffic. A second service allowed users to select videos from a small cassette library, which an automated system would play and transmit to their homes—a primitive form of video on demand.
The Hi-OVIS control room.
But the operation center's central role in the Hi-OVIS social experiment was generating and delivering local programming for the community. A full television studio could generate local programs and receive audio and video input from home terminals. That made it possible for people to generate programs in their own homes such as cooking demonstrations. Programs were designed to be interactive, whether discussing local affairs or providing instruction on how to speak English. The network also had two mobile facilities to produce programs on site—one a microbus carrying five crew members and the other a light van carrying two members. It added up to quite an operation, highlighting the importance of the social experiment.
One service was conspicuous by its absence—voice telephony, which fell under MITI's bureaucratic rival, the Ministry of Posts and Telecommunications. As in the United States, regulatory walls separated cable television and telephone services.
Hi-OVIS was turned on with considerable fanfare on 18 July 1978, and VISDA extensively documented the results.
For a one-of-a-kind system developed when standard products were few and far between, the Hi-OVIS fiber-optic system succeeded remarkably well. I don't know of any other noteworthy system that ever used 150-µm core step-index multimode fiber; it's never been a standard size.
Some challenges were significant. Installing the cables required fusion-splicing fibers at 1,100 points. No one was keeping track at the time, but that might have been a record for the most splices on a single project. The use of large-core fibers doubtless helped, and a report observed that "a lot remains to be improved in the future," including the need for highly trained splice technicians. Nonetheless, only one splice failed during the first years of operation covered by the report.
Video and electronic equipment at the operation center and in the terminals caused more problems than the optics. Transmission optics accounted for only 11 percent of equipment failures. Most of those failures were in the transmitters or receivers, not in the fibers. Except for the one bad splice, all the fiber failures were caused by objects hitting the cable. That's still the case today, with most cable failures blamed on "backhoe fade" caused by careless contractors.
The fiber-optic hardware remained in operation until Hi-OVIS closed at the end of March 1986. By then, it was far from the mainstream of fiber-optic development, which had turned to long-haul digital transmission at rates of hundreds of megabits per second through single-mode fibers. The final report on the Hi-OVIS project gives little detail on long-term performance, but it's safe to say that it was at least adequate.
The heart of the Hi-OVIS experiment was testing the social impact of the new community-oriented services, so VISDA carefully tracked how people used the system and surveyed them on their reactions.
List of locally produced programs—January 1979.
VISDA picked Higashi-Ikoma because it was a relatively young community with a heavy concentration of professionals and with a high level of spending on "information"—a group they expected to be particularly interested in participating in the experiment. A survey found that 58 percent of households were nuclear families: two parents and their children, with another 18 percent having a grandparent living with such a family. Virtually all the men worked, but 68 percent of women considered themselves "housewives." With someone home most of the time, such families were prime candidates for local interactive programs.
High production costs limited locally produced programs to about three hours a day in the first phases of the project; the table below shows a list of local productions in a typical week in January 1979. Opinion and "Hi-OVIS wide" programs sought to build community. The daytime programs presumably were aimed at housewives and children, with evening programs for working adults. Typically about 25 to 30 of the 158 households (15 to 20 percent) tuned in to Hi-OVIS local programs when they were running, but because they were broadcast only a small fraction of the day, subscribers spent many more hours a week tuned to the television channels carried by Hi-OVIS.
Hi-OVIS train timetable.
Participation in the two-way services came more slowly, despite strong encouragement from staff. Subscribers were particularly slow to go on camera and produce their own programs, complaining of shyness when asked by staff members. Many never did go on camera, but gradually they grew accustomed to the system, and it became a part of the community and a subject of conversation in itself. "With the advent of this new medium, community consciousness has been rising in the Higashi-Ikoma system," Kawahata wrote in 1980. Surveys after it was shut down show that residents missed it.
At first the system was a new toy, producing an initial surge of requests for still images, followed by a decline. Yet after a few months, subscribers still requested a few thousand still images a week, with more than half character-generated. The single most common request was to show train schedules, followed by a guide to Hi-OVIS. Videos shown at fixed times on a dedicated Hi-OVIS channel were requested a few thousand times a week.
A separate test of video-on-demand service from a small library of cassette tapes was initially overwhelmed with requests, but demand dropped quickly because only a handful of tape decks were available. Children quickly figured out how to make requests, and literally wore out tapes of some of their favorite cartoons, playing some more than 500 times. Program managers did not replace the worn-out tapes, thinking that children should watch more educational videos. Overall, men's favorite videos were golf lessons, and women's were cooking programs.
In its later years, Hi-OVIS shifted away from a visionary quest to build local communities toward more pragmatic tests of business prospects for the new services. Local programs were cut back to an hour a day, and some channels were leased to businesses that transmitted infomercials.
In retrospect, Hi-OVIS was a tremendously expensive experiment as well as a visionary one. In 1997, Kawahata estimated that the whole project cost about $80 million from inception to its completion in 1986. Cost breakdowns are not available, but the operation center, staff and local programming probably accounted for most of the budget. (MPT's Tama test cost an estimated $6 million.) At least during its early years, Hi-OVIS participants weren't charged for the services they received.
Fiber to the farm in Canada
Canada blazed a very different and far less expensive trail when it opened the world's second major fiber-to-the-home trail in June 1981. The goal was to see if fibers could bring advanced telecommunications services to highly productive agricultural areas. Canada's Department of Communications and the government-owned Manitoba Telephone picked 150 homes in and around Elie, Manitoba—a tiny town 50 km west of Winnipeg.
Elie, Manitoba, June 1982.
The Canadians kept it simple, spending $7.5 million (U.S.) to string fiber to carry telephone, cable television and videotext—a primitive information system that accessed data stored on mainframe computers at hundreds of bits per second. The data center was a trailer crammed with equipment parked in back of the Elie telephone office. The system had no studio for local programs, and outside of the videotex, it offered nothing an average suburbanite would not have taken for granted.
But it was a wonder in Elie, where residents had been sharing party phone lines among up to 10 families, and depending on tall antennas to pull in four snowy broadcast television channels from Winnipeg. It was a population in desperate need of better communications. They were delighted with private phone lines and a few American television channels. Children used the videotex system to play video games. A wheat farmer who cultivated more than three square miles used it to check the weather and prices for grain and hogs. Yet, like Hi-OVIS, Elie was ahead of its time, and it shut down a few years later because it was too expensive. A larger-scale test in the French resort city of Biarritz suffered the same fate.
Legacies of Hi-OVIS
The Japan Key Technology Center tried to follow up on Hi-OVIS with a project called Advanced Hi-OVIS or IBIS (for Interactive Business Information System). The Japanese government had created the center in 1985 to sustain the basic research that had been conducted by Nippon Telegraph and Telephone before its privatization that year. They put the center under the joint management of MITI and MPT, making it the logical organization to build on the legacy of Hi-OVIS, but this time they considered the economics.
Gone was the local programming, but the video on demand remained, with programs stored on laser disc as well as tape. The system also retained video cameras in homes, but for use in video conferences and videophones, not for local programs. Digital data were added at rates to 19.2 kilobits per second, state-of-the-art for dial-up modems, and companies were hooked into the system to provide information services. Future plans included transmitting high-definition television.
Sumitomo Electric, which had supplied fiber hardware for Hi-OVIS, developed a new generation of technology for IBIS. They retained the switched star structure, but chose 50-µm core graded-index fibers to carry signals. Downstream data were transmitted at 143 Mbit/s using 1.3 µm LEDs, which could span up to 3 km. Upstream voice, video and data were transmitted by an 0.85-µm diode laser that was wavelength-division multiplexed through the same fiber. The system began operating in Osaka with 155 subscribers in 1988.
With a $30 million budget, including development, IBIS needed several thousand subscribers to break even. To do that, they started monthly subscriber fees at $150 for data only and $400 for video and data, and they planned to raise rates to $250 to $550 per month. Instead, the system faded away, probably a casualty of the bursting of the Japanese economic malaise of the 1990s.
Looking back, the visionaries behind Hi-OVIS and the wired city experiments were right about some very important things. Technology could bring a vast variety of information to our homes and help us to choose what information we wanted. That's what the Internet and the global communications network do today. The problem was not the vision of an information network, but that the technology of the 1970s or 1980s was neither adequate nor affordable for the task.
What the visionaries didn't realize was that communities don't have to be local in geography. Thanks to telecommunications and the Internet, we can belong to communities that are not confined by geography. For example, the electronic edition of this magazine reaches the optics community around the world. In that sense, the visionaries have left a legacy that is reaching beyond their vision.
The visionaries behind Hi-OVIS also left a more tangible legacy. As of last May, Japan had 13 million households directly connected to fiber, plus another 6.6 million households in apartment buildings served by fiber. With more than twice as many households, the United States lags far behind, with fiber to only 6.5 million homes.
Jeff Hecht is a science and technology writer based in Auburndale, Mass., U.S.A.
References and Resources
>> T. Nakahara et al. "An Optical Fiber Video System," IEEE Transactions on Communications COM-26 7, 955-61, July 1978.
>> J. Yudeman et al. Teletext and Videotex in the United States, McGraw Hill, N.Y., 1982.
>> W.H. Dutton et al., eds. Wired Cities: Shaping the Future of Communications, G.K. Hall and Co., Boston, 1987.
>> New Media Development Association. A summary version of the Comprehensive Report on Hi-OVIS Project: Jul'78 to Mar'86, NMDA, Tokyo, 1988.
>> S. Takeuchi. "Optical Fiber Video Interactive System for IBIS," Proceedings, IEEE Region 10 Conference on Computer and Communication Systems, September 1990, Hong Kong.
>> J. Hecht. City of Light: The Story of Fiber Optics, Oxford, N.Y., 1999.