91-09/Bob.Japan.1 From: cyberoid@milton.u.washington.edu (Robert Jacobson) Subject: Japan Report, Part 1: May 1991, Fujitsu, NTT, Kanagawa Date: Thu, 19 Sep 1991 17:50:15 GMT Organization: Human Interface Technology Lab, Univ. of Wash., Seattle On two occasions this year, once in May and again in July, I was privileged to travel to Japan to study virtual worlds technology in that nation. Each trip was valuable in its own way, the first as a general introduction to Japan and in the field of electronics, the leading Japanese computer firm, Fujitsu; and the second to validate lessons learned on the first, including an appreciation of the situation in which Japan's leading virtual-worlds researchers find themselves. First, however, I would like to make some general comments. * * * Without question, Japan is a culture unique unto itself that (like all cultures) is best experienced from within. The tour books are helpful, and it is wise to read up before visiting, but somewhere along the line all explanations fail. The surface manifestations of Japanese culture -- the trains that invariably run on time and stop precisely, within inches of the indicated platform; the uniformity of clothing among urban workers; and the unfailing politeness of Japanese people in all situations -- soon give way to a deeper appreciation of what it means to be Japanese. I would not say that conformity is the most striking Japanese characteristic, though even the Japanese like to joke about how fads sweep the nation. More essential to "being Japanese" is a sensitivity to the appropriateness of behavior in different settings, and a concern for the well-being of individuals among the collectivity. I do not want to overdo this to the point of blindness or sentimentality. Truly, there are divisions between the wealthy and the less well-off (one seldom sees the poor, though they are present); some people walk while others ride in company cars; and the styles of food indulged in by cosmopolitan Japanese are many more than those enjoyed by the salt of the earth. But the pride that Japanese feel in being Japanese is impossible to miss and the care to do what is right, according to custom (even to the short-order cook who takes time preparing the traditional bento lunches or the department-store clerk wrapping a purchase), is pronounced. One story will suffice to sum up the essence of Japanese life: One day, in a busy train station downtown, our guide was purchasing train tickets for Tom Furness and me, in order to hasten our on-time arrival at an important appointment. In his haste, he bobbled the handful of change given to him by the ticket machine, spilling coins everywhere. Suddenly, everyone within hearing of the falling change stopped in his or her place, spotted a rolling coin, chased it down, carried it back to our guide, and place it in his palm, bowing as he or she did so. No words were exchanged; none were necessary. Anything less would have been impolite. One might also mention that 80 percent of private automobiles are painted white, the color of purity, not out of any metaphysical awareness but simply because cars painted white are easier to resell. On the other hand, taxis, destined for hard use and early retirement (the only used cars on the road are heirlooms), are garishly colored on the outside and linen-lined on the inside. Above all, as my Japanese friends remind me today, the appropriate word for all actions is -- patience. Things get done when it is right for them. Having set the tone for our visit -- we, the innocent gaijin (foreigners), for whom all gaffs are permitted but two (soiling the water in the soaking tub and wearing shoes on the tatami mats -- now embark on a short tour of Japan's virtual worlds community.... * * * May, Fujitsu Research Institute. Tom Furness, director of the HIT Lab, and I made our first trip to Japan in the still springlike but warming atmosphere of May. We were hosted by the Fujitsu Research Institute's Dr. Masahiro Kawahata. The Fujitsu Research Institute, a relatively new unit of Fujitsu Electronics, is considered to be Fujitsu's think-tank, a place where long-range planning for the entire corporation takes place. Dr. Kawahata, a former IBM engineer and director of the groundbreaking Hi-OVIS Japanese cable television experiment, is a managing director of FRI and one of those Japanese among whom groups congeal to undertake pioneering efforts. We were under two constraints on this trip. First, as guests of FRI, it would have been terribly impolite for Tom and me to visit with Fujitsu's competitors. Although we in other lands often see the Japanese society as monolithic, in fact there is fierce competition among Japanese corporations increasingly beyond the control of MITI, the once all-powerful Ministry of International Trade and Industry, and other mediating institutions. One may visit a company, its subsidiaries, and (with permission) the "family of companies" friendly to a firm, without creating too much fuss. (These "families" are different from the keiretsu, or large associations of commonly owned companies, for sometimes there is competition even within keiretsu and closeness among firms from different keiretsu doing business across keiretsu boundaries.) As a guest of a firm, however, one does not spend time with competitors ... and learning the terrain of business takes considerable time. So Tom and I stayed close to Fujitsu. However, we were enabled to visit with NTT, the national telephone company; and with MITI's various laboratories in Tsukuba Science City, as these are both customers and collaborators with Fujitsu. Our second constraint was that we not divulge confidences shared with us by Fujitsu, particularly concerning its activities in the area of virtual worlds. So my report on this first trip is necessarily sparse. In fact, our days were spent largely in the company of Dr. Kawahata and his associate, Mr. Masaru Hosoda, who accompanied us on all of our field trips. We never felt, however, that anyone was withholding information from us or being less than candid in describing their research or referring us to others. On returning to the U.S., our initial reports regarding Japanese accomplishments were considered with some skepticism, our inquisitors almost always wondering what was held back from us. I believe I can honestly say that we were as well-informed about current activity as we would be in an analogous university setting. About our hosts' future plans, however, I can say very little, as we were not made privy to them. Our tour of Fujitsu was almost overwhelming in terms of the many telecommunications and computer technology teams we met. In addition to meeting with FRI researchers, we also met with researchers from Fujitsu Laboratories, from which FRI was spun out, and with operating company scientists and engineers. A surprising number of these individuals were women, no less than we might see in an American engineering company, although I cannot speak to the respective roles of the researchers within these organizations. Only one woman researcher was the primary investigator of a research project, but again this is not much different from affairs in the U.S. Most of those with whom we spoke at Fujitsu were not involved directly in virtual-worlds research, but instead were working with "conventional" telecommunications and computer technologies. I add the parentheses to indicate that the work is on developments of technology that are cutting-edge in these fields. None was unrecognizable, but clearly Fujitsu has ambitions beyond its current market position. The two most interesting laboratories we visited at Fujitsu were those of Dr. Shuzo Morita and his colleague, Mr. Masanori Kakamoto, who head up Fujitsu Laboratories' well-publicized effort to develop virtual-worlds technology; and that of Mr. Kazutomo Fukuda and Osamu Imoto, who direct Fujitsu's HABITAT Project. Dr. Morita and Mr. Kakamoto have used the standard VPL technology, enhanced by the use of a neural network to instruct the DataGlove, to create an entrancing undersea world populated by friendly jellyfish and a menacing shark. One lures the jellyfish by means of a wave, Japanese-style (underhanded), and chases away the shark with a beam of light that emanates from the glove. The jellyfish give off cooing sounds not unlike the characters in a Japanese-produced children's cartoon. I was struck by the ad hoc presentation area in which we were shown this experiment, not too different from our own back at the HIT Lab. Dr. Morita's lab is striving toward a more standardized ensemble of equipment and greater comfort in their presentations. Also shown to us was the self-composing musical array put together by the woman engineer whom I mentioned earlier. This device "taught" itself to mimic and then expand upon Mozart's work. The HABITAT Project is an enhancement of the first Habitat designed for Commodore by Randy Farmer and purchased by Fujitsu, adapted to Japanese tastes, and now sold as a public telecommuni- cations service. For those who have not seen HABITAT, it is a sort of visual MUD, which a player, equipped with a Fujitsu FM Towns computer/CD-ROM player and modem, enters via the telephone lines. A player assumes identities, acquires trading tokens, and interacts with others who are similarly present in HABITAT at the time one calls. Several thousand Japanese currently use HABITAT, which has grown into many hundreds of "rooms," or separate environments. Tom and I were quick to note that the combination of virtual-worlds with HABITAT would make for a killer service, if the two could be made compatible. This had not escaped our Japanese hosts, but the task of elaborating HABITAT has been made difficult by the sheer immensity of upgrading and maintaining the current system. (We also suggested a build-it-yourself facility, something which had escaped our hosts' attention.) HABITAT is commercial, but it pro- bably still does not cover its costs; so it must be seen as primarily experimental at this stage. * * * Our other memorable visits on this first trip were to NTT and the MITI labs in Tsukuba. Regrettably, we did not get to see the ATR Laboratory, a real leading-edge spinoff of NTT, in Kyoto. NTT is the largest company in Japan, capitalized at many, many billions of dollars. Until recently, it was a monopoly, supplying local telephone service without competition throughout Japan. (Long-distance service is provided by KDD, still a monopoly, I believe.) Today, as a result of liberalization of the economy, NTT now faces competition on many fronts -- transmission, equipment, and so forth -- but still remains dominant across the board. However, as a result of the privatization of NTT, its budget is no longer certain. In fact, in the last year, NTT has suffered severe budget cuts, limiting its ability to fund basic R&D. Nevertheless, one cannot underestimate the variety of experiments taking place at NTT's Human Interface Laboratories, in Kanagawa, under the direction of Dr. Takahiko Kamae and Mr. Takaya Endo. Work on various interface variants is an ongoing preoccupation of these Laboratories, in both visual (Drs. Kobayashi, Suenaga and Ishii, and Mr. Hiraiwa) and acoustic environments (Drs. Furui and Koizumi, the latter from another NTT lab), and general human factors research (Mr. Suzuki) . However, we saw no work that could be called "virtual worlds" research per se, although with time some of the more interesting prototypes of facial recognition and groupwork technologies could find application in a virtual world. The emphasis on traditional teleconferencing, which I find so annoying among American telephone companies as if it were the end-all and be-all of telephony, was also pronounced at NTT, leaving me to wonder if the Japanese telecommunication vision of the future is so different from our own. (However, I do recommend a viewing of "Birthday," a short NTT videotape, which is perhaps the warmest evocation of the information age to date.) * * * Tsukuba Science City, a planned development intended to bring together researchers and developers from government laboratories and private firms, is something of a disappointment. It is not an unattractive place, but the spirit of technical excitement one finds at the most active (and admittedly, chaotically so) industrial parks in North America was not present. Instead, we found within the various laboratories scientists of a more academic bent who were struggling to bring forth Big Ideas with relatively small resources. This is not to say that the MITI labs are unproductive; far from it. Much interesting research is taking place there, and it is from the MITI laboratories -- primarily the Mechanical Engineering Laboratory (MEL) and the Electro-Technical Laboratory (ETL) -- that the pioneering telerobotics experiments of Drs. Tachi and Maeda, the tactile-field work of Mssrs. Fukui and Kotoku, and the successful optical networks experiments of my close friend, Dr. Ishihara, have sprung. With larger budgets and closer intercourse with industry, however, these scientists could do much more. In fact, Dr. Tachi, whose telepresence gear permits one to "enter" into the body of a video and virtual world-equipped robot body, recently left for Tokyo University, presumably in pursuit of that more intimate relationship with commerce. (His work is being continued at MEL by Dr. Maeda.) At Tsukuba University, a technical school close by the MITI laboratories, we met Dr. Hiroo Iwata, who with his students is conducting research in tactile environments. Dr. Iwata has tricked up two fine demonstrations, one a ramp on which one "walks" up stairs using roller-equipped shoes wired to sensometers; and an exoskeleton not unlike the Exos, which permits one to pick up and toss (not very far, however) a virtual ball. Dr. Iwata is one of the four Japanese scientists who were featured at the Nikkei Symposium on Artificial Worlds and Tele-Existence held in July, on which I next report. * * * Finally, I should mention the work being done by my dear friend Mr. Koji Sato and his protegee, Mr. Masao Kunishige, for the Kanagawa Prefectural Government. Kanagawa, the home of Yokohama, occupies a position second only to Tokyo (which it borders) as an economic center -- a situation it is trying to remedy by attracting more industry to itself. Mr. Kunishige now presides over Kanagawa Science Park, a large integrated industrial development to which many high-tech firms are moving. Mr. Sato is a researcher in industrial policy for Kanagawa and, lately, a student of virtual worlds technology, as well as an instructor at Chiba Institute of Technology, a prominent technical university in Kanagawa. * * * I have left out the wonderful time spent in restaurants with colleagues, and the adventures that novice travellers always have in strange places (easier to solve here, with friendly people, than in many other locales, however), in order to keep this report reasonably short. One week in Japan was all that Tom and I could afford, and we were completely saturated by the time we left the crowded Narita International Airport. I must not fail to mention, however, the rousing speech given by Dr. Kawahata to his researchers assembled from all of Fujitsu's relevant divisions. "For too long," he observed, "we in Japan have been relying on research done by others, improving on it but not generating it. Now, it's time for us to do our part, to add to the basic knowledge regarding artificial reality and to make *our* a contribution for human well-being." He was loudly applauded.... --