TRONSHOW '99 got off to an enthusiastic start on Wednesday, March 10, 1999, when a standing room only crowd of approximately 250 people showed up for a special lecture on design by TRON Project Leader Ken Sakamura, who is also an architect of houses, buildings, and even cities. During his lecture, Prof. Sakamura described the contents of TRON Design, a book especially published to mark the occasion of the 15th anniversary of the founding of the TRON Project. TRON Design is a photo collection of computer components, computer systems, furniture, and structures that the Project Leader himself designed, including the TRON Intelligent House that was actually built and used as a test bed for experiments in computerized living.
Toward the end of his lecture, Prof. Sakamura mentioned his recent trip to Europe, where he gave a presentation on the TRON Architecture to French enthusiasts of the GNU/Linux freeware movement. GNU/Linux, he said, is extremely popular in Europe, since Europe does not want to become dependent on American-made operating systems. He also mentioned "concept design," which he believes the TRON Project has succeeded at; and he said that he would like to do a network language similar to, but not exactly the same as XML. The TRON Project Leader's interests are thus moving from the development of basic technologies, such as microprocessors and operating systems, to higher level technologies, such as middleware and content description, which are important for the creation of networks.
The most noteworthy feature of this year's TRON show was a series of workshops, where attendees could sit down and listen to representatives of various exhibitors describe the TRON-based and/or TRON-related technologies and products. During the three-day the following technologies and products were explained at the workshops.
Across from the seminar hall where the workshops were held, a theater show was presented in which a live actor in a home setting on a stage demonstrated the TRON-concept of computerized living based on cooperation among intelligent objects. The demonstration showed how audio-visual equipment, air conditioning systems, illumination lamps, and home electronic appliances such as microwave ovens can be controlled via preprogramming and/or a wrist mounted controller for the comfort and convenience of humans. This demonstration drew the attention of a lot of film crews, who spent considerable time filming the demonstration.
In the exhibit area, there were exhibits by the companies and organizations that gave presentations in the workshops. Aplix Corporation, the leading proponent of JTRON, displayed three products: "JBlend," an operating system based on the JTRON1.0 specification that can execute both Java and native ITRON programs; "Jaccas," a distributed network architecture based on Jini and Point of Sales (POS) codes that allows computerized devices to be linked together without a server or server administration; and "BitSpirit Internet," a World Wide Web browser for embedded use. Metroworks Corporation displayed information concerning its "CodeWarrior" development environment for NEC's V800 microprocessor series and MIPS embedded systems, plus "Pipeline," a toolkit for developing graphical user interfaces for embedded systems.
Newly developed TRON-based products were also on display in the exhibit area. Personal Media displayed its new flagship products: the BTRON-specification "B-right/V" and JTRON-specification "J-right/V" operating systems for IBM-PC/AT compatibles. The company also had on display its BTRON-based "1B/V3" and ITRON-based "I-right/S" operating system, which have been on the market for some time. The Sakamura Laboratory of the University of Tokyo, which is supporting the BTRON architecture with freeware, had on display three software application it developed: a basic Web browser, an e-mail application, and "SmartPoint," a portable groupware program that supports presentations and discussions. ITRON fans also had something to look at. Yokogawa Digital Computer demonstrated its ADViCE in-circuit emulator and microVIEW-G debugger, which support real-time operating system software development on a wide variety of CISC and RISC microprocessors.
In addition to commercial and research organizations, non-profit organizations also had exhibits on display. There were exhibits by the TRON Association, TRON Enableware Research Group, the BTRON Technical Committee, the TRON HMI Technical Committee, and the TRON Fan Forum, a TRON user group! Yes, there are TRON users, and their exhibit was about an experimental system for a new type of TRON Intelligent House called "G-House." G-House is based on a new type of ITRON-based "intelligent object" called "G-Cube," a robot module that can be instructed to move about the house and perform various tasks for the occupants.
TRON Web was invited to attend a press chat session that Professor Sakamura had with members of the press on the afternoon of March 10. During the chat, the TRON Project Leader, who has had to carry the burdens of both design and popularization of the TRON Architecture on his back, spoke from the point of view of an old technological warrior about the TRON Project. He said he launched the TRON Project in 1982 because of the humiliating IBM incident in which Japanese companies were lured into industrial spying. Because of that, it became obvious that Japan needed its own microprocessor and operating system technology; and an open architecture was important to spread the technology. One of the outstanding achievements was the Gmicro 500 microprocessor, which was faster than Intel's first Pentium chip.
Unfortunately, the TRON Project was not able to take the initiative in the creation of the digital society, Prof. Sakamura said. However, he said, it has made many contributions and wants to make more. For example, Bill Joy of Sun Microsystems Inc. who invented the Jini network operating system for linking Java-based appliances together said he got the idea for this from Prof. Sakamura. The BTRON-specification operating system with its superb multilingual functionality is also another, he noted. It's the only operating system that can handle multilingual processing properly, a function that is also needed for electronic books. Moreover, Java and Jini also need BTRON to run quickly, he added.
During the chat, Prof. Sakamura showed TRON Web a copy of the March 8, 1999, issue of the French computer magazine Programmez! In it, there is an interview with the TRON Project Leader, who gave a presentation on TRON at a Linux convention in Paris. And what is one of the things the editors chose to illustrate the text with? None other than a picture of TRON Web's entrance page! Maybe the project is finally coming out of the long tunnel of darkness after all!!
Vive Programmez! Programmez pour l'architecture BTRON, s'il vous plaît.
Keynote Speech
The 15th TRON Project International Symposium began with a keynote speech by TRON Project leader Ken Sakamura, who pointed out that the TRON Project has been around for 15 years--an incredibly long period of time for a Japanese technology project that is not a national project. Unfortunately, many aspects of the TRON Project, in particular the BTRON-specification operating system, haven't been well received in the land of its birth. In fact, Prof. Sakamura plainly stated that this was because Japan is negative about new things. He said it would have been easier to do something new in the U.S., and many people told him just that.
However, his keynote speech was not all lament. He noted that the project has succeeded in producing both specifications for creating both operating systems and microprocessors. It also helped pioneer a concept that has become common sense today--that a computer architecture has to be "open." As to why the BTRON-specification operating system has not become very popular, he noted that whereas the designs of ITRON, CTRON, and the TRON VLSI CPU were conservative, BTRON is actually based on revolutionary design concepts, such as hypertext at the operating system level. The latest versions of Microsoft Corporation's and Apple Computer Inc.'s operating systems enable users to create HyperText Markup Language (HTML) links, but this is not the same as having a hypertext/hypermedia filing system at the operating system level.
Prof. Sakamura does not believe that the BTRON-specification operating system is a failure. Rather, he considers it a new class of computer, which he likes to refer to as a "communication machine." He said that the concept of the personal computer--a term he called a "marketing term"--is based on the mainframe computer idea of a "general purpose computer," which is old and thus should be avoided. In comparison to a personal computer, he said the communication machine is a new design concept for a computer with limited functionality to be used in open networks where communication is based on scripting languages and standardized data formats. Accordingly, it would cost about a tenth of what a personal computer would cost.
The next steps for the TRON Project are standardizing data formats and figuring out how to deal with the various types of data contents that will be transferred across the Internet in the future, Prof. Sakamura said. This involves creating a new method of describing data attributes; that can be done, for example, by tagging someone's name in a manner that allows it to be interpreted as the "name of a human." A good multilingual computer is essential in such an environment, he noted. In keeping with these goals, Prof. Sakamura said three new application projects have been launched: (1) an electronic commerce project, (2) a museum project, and (3) an electronic book project.
Summing up, Prof. Sakamura said he believes that the TRON Project has succeeded at the "concept level," where he believes there is nothing to reconsider. However, he stated that when he launched the TRON Project, he underestimated the political and marketing obstacles, such as attempts by the U.S. government to derail the TRON Project with threats of unilateral trade sanctions against Japan and the lack of enthusiasm among Japanese electronics companies for marketing new technologies.
Hiroaki Takada, who is a professor at Toyohashi University of Technology and also the head of the ITRON Technical Committee, gave the symposium's first subproject presentation on the state of the ITRON subproject. Prof. Hiroaki said that one of the reasons the ITRON subproject has been an outstanding success is that Japan makes a lot of embedded systems, and hence it needed a standard real-time operating system for such applications. As a result, ITRON has become a de facto standard in Japan, accounting for 30% of all systems on a design basis. It is especially widely used in consumer appliances and communications equipment. Moreover, ITRON is starting to draw interest overseas, particularly in the U.S. For that reason, a U.S. branch of the ITRON Technical Committee is scheduled to be established and commence its activities in 1999.
The good news about ITRON is that the micro-ITRON4.0 specification is 98% decided, and that the documentation is almost finished, Prof. Takada said. The micro-ITRON4.0 specification is different from the micro-ITRON3.0 specification in that its minimum subset is actually smaller than the minimum subset for micro-ITRON3.0, and yet its full set is larger than micro-ITRON3.0's extended level. This allows the new specification to be used with a variety of standardized profiles, such as an automotive profile. The micro-ITRON4.0 specification is also aimed at larger types of embedded systems, and thus it includes support for software parts, in addition to standardized configuration methods with configuration steps. The provision of software parts is the theme of the JCG Project. It includes the JTRON2.0 specification, CORBA for embedded systems, and a GUI package for embedded systems, which will be released to the public as freeware in 1999. A separate technical committee has already drawn up an embedded TCP/IP specification for ITRON.
One of the areas in which the ITRON subproject has encountered problems is in the area of debugging environments, which in turn has affected software development environments and tools. The main problem is that there are various versions of micro-ITRON-specification kernels, and up to now it has been a practice to make debuggers, in-circuit emulators, and the like individually compatible with the respective micro-ITRON kernels. Accordingly, there is a strong need for standard debugging interfaces that will allow debugging environments to be easily matched to the numerous micro-ITRON-specification kernels. An additional benefit is that once the standardized debugging interfaces have been drawn up, the micro-ITRON-specification kernels will benefit from better development environments, which should allow the use of ITRON to be expanded to the types of applications that have been the preserve of proprietary real-time kernels and development environments up to now.
JTRON
When one mentions Sun Microsystems Inc.'s Java or Jini in the U.S. or Europe, ITRON and the TRON Project do not come to mind. However, in Japan, ITRON-specification kernels are the engines for running Java-based appliances; and Jini, which is designed for dynamically networking Java-based appliances together, is the software embodiment of the ultimate goal of the TRON Project--the TRON Hypernetwork. This is referred to as the Highly Functional Distributed System (HFDS) in TRON techical circles. In his presentation on the JTRON subproject, Mr. Shinobu Nemoto of Aplix Corporation, explained the background of the Java on ITRON (JTRON) project, the latest developments in the project, and products that have been implemented on the basis of JTRON specifications. The first JTRON specification, JTRON1.0, appeared in December 1997 and was aimed at merging the highly portable Java networking programming language with widely used, high-performance ITRON-specification real-time operating system. JTRON1.0 was quickly followed by the JTRON2.0 specification in October 1998.
Both Aplix Corporation and Personal Media Corporation have developed operating systems based on JTRON. Aplix has developed "JBlend," which runs on microprocessors made by NEC, Hitachi, Toshiba, Fujitsu, and Intel; and Personal Media has developed "I-right/S" and "J-right/V," which run on SPARC-based Sun workstations and Intel-based IBM-PC/AT compatibles, respectively. Commercial products based on these JTRON-specification operating systems have also appeared. The most famous is a small handheld computer/remote controller called BossaNova, which was developed by PFU Ltd. and is currently undergoing trials. In addition, Matsushita Graphic Communications Systems Inc. has developed an Internet-compatible color facsimile machine; and a car navigation system is also said to be in the works.
In order to network together JTRON-based appliances, Aplix has developed a new Jini-based networking operating system called "Jaccas," which stands for "Java Controllable Consumer Appliances System." The basic concept behind Jaccas is the separation of the program for controlling an appliance (a Java Applet stored within the appliance) and its execution environment (a JTRON-based remote controller such as BossaNova). Although Jaccas is based on Jini, it is not the same. The main difference is that Jini requires a server, which will prevent it from spreading into the Japanese family environment, Mr. Nemoto said. Moreover, Jini cannot handle multiple air conditioners, multiple video cassette recorders, etc., he added. Accordingly, what Aplix has done is to use Point of Sales (POS) codes, which are almost in universal use in Japan, for identifying appliances to be controlled via Jaccas, including multiple units of the same type.
BTRON
The BTRON subproject is still awaiting the results of the GT Mincho project at the University of Tokyo in order to be able to upgrade the BTRON-specification operating system to full multilingual capability with an unlimited Japanese character set. However, the BTRON subproject has not come to a complete standstill. In January, a new venture capital firm called Sennet Inc. was launched under the leadership of a former Mitsubishi Electric vice-president to pursue opportunities centering around the BTRON subarchitecture. The company is planning to establish a subsidiary in Silicon Valley, and it has also succeeded in gaining capital participation from both North American, European, and East Asian sources. Moreover, as Mr. Akira Matsui pointed out in his presentation on the current status of the BTRON subproject, the operating system continues to be applied to new tasks. For example, the micro-BTRON operating system for PDAs, which is already used by insurance agents, the general public, and museum visitors, has now been applied to an order entry system by Seiko Instruments Inc. This is proof that the micro-BTRON operating system is a splendid basis on which to base all sorts of miniaturized hardware applications.
So what's holding up the GT Mincho project? As Prof. Sakamura has pointed out many times in the past, and as Prof. Noboru Koshizuka also of the University of Tokyo pointed out once again in a workshop during TRONSHOW '99, you cannot just dump a huge character set of 50,000 or 100,000 kanji (Chinese characters) in a computer system and ship it off to customers. Character sets of this size need complex databases with various types of information, such as information to limit the number of characters to be processed (e.g., for young children) or to explain the differences in the uses of characters that are very similar. Unfortunately, as Prof. Koshizuka pointed out, there are differences of opinion among scholars on how to use these similar kanji. Another problem with a large Japanese character set is that it affects the kana-to-kanji conversion engine that converts character strings in the Japanese hiragana syllabary into kanji, since the conversion dictionary becomes considerably larger. Prof. Koshizuka said researchers at the University of Tokyo are currently experimenting with mechanisms to dynamically set characters depending on the selections a user makes. Another innovation in the works is a kanji database on an Internet server that will allow BTRON users to download specialized characters as they need them, he said.
TRON HMI Research
Prof. Noboru Koshizuka also gave presentations on the TRON HMI (Human-Machine Interface) subproject both in a workshop at TRONSHOW '99 and at the symposium. Prof. Koshizuka said it is easier to create an HMI specification than to get people to follow it. In fact, he has done experiments with his students, and as a result he has found that they do not follow it. This points out the need for a development environment based on a reference model, he said. However, the TRON HMI guidelines themselves are available at present in both Japanese and English. They were published in Japanese as Toron hyu-man intafue-su hyoujunn hando bukku by Personal Media Corporation in 1993, and in English as TRON Human Interface Specifications by the TRON Association in the same year. In addition, a revised edition of the Japanese version was published in 1996 by Personal Media Corporation.
The main features of the TRON HMI are: (1) loose standardization, which means no set look and feel; (2) standardization of abstract HMI layers; (3) consistency between the Graphical User Interface (GUI) elements and Solid User Interface (SUI) elements; (4) normalization based on Enableware for the disabled; (5) multiculturalism based on multilingual functionality; (6) safety provisions; and (7) multimedia functionality, such as a speech based interface. The GUI/SUI elements, incidentally, cover everything from the standard personal computer screen and its peripheral nput devices to the tiny screens and knobs and switches of office automation equipment and numerically controlled factory devices. Thus TRON HMI guidelines are intended for standardizing human-machine interfaces throughout human society, not just those of a particular piece of equipment or a certain field, such as home or business applications.
The TRON HMI model that lies between the human end user and the equipment is a multilayered structure that consists of primitive operations and primitive parts--the letter of which are used to make compound parts--that interface with the human end user. Below this layer lie operation sequences and layout rules that interface with function calling rules, which directly call up the equipment functions. This HMI model is analogous to human language in which there is a lexical layer--consisting of words and idioms--that interfaces with grammatical rules, which in turn interface with semantic rules that call up human thought processes.
The TRON HMI guidelines are actually the "glue" that will allow all the parts of the TRON Hypernetwork to come together and function properly for the human end users who live and work within it. Thus getting companies to follow these guidelines is an extremely important step in the creation of the computerized society of the 21st century.
CTRON
The CTRON subproject, like the ITRON subproject, has been a very successful subproject of the TRON Project. CTRON is currently used in ISDN switching systems, in ATMs and intrabank transfer systems, and in PBXs throughout Japan. What that means is that thanks to CTRON just about everyone in Japan is a user of the TRON Architecture. The manager of the CTRON Technical Committee, Mr. Toshikazu Ohkubo of NTT Software Corporation, gave a presentation on the next two goals of the CTRON: (1) the creation of a Multimedia Network service Platform (MNP) to enable CTRON to support multimedia network services when optical fiber connections are run to every home in Japan; and (2) transmitting the fruits of the CTRON subproject overseas, in particular to China.
As everyone using the Internet and the World Wide Web knows, new types of multimedia network services, such as data streaming, are now in their embryonic state. The biggest problem at present is that very few users at the client end of the network have access to high-speed connections to properly utilize them. However, another problem is that network servers are not optimized for supporting multimedia service content. Mr. Ohkubo said that at the server end of the network real-time performance has to be raised, management functions have to be optimized to set and change network resources, resource management agents have to be made mobile using Java, and service requirements and network requirements have to be matched through a translation function.
To achieve this on the CTRON-based MNP architecture, Mr. Ohkubo said a three-layer interface architecture has been built above the Basic CTRON operating system. This consists of the Basic OS Interface, the Distribution Extended Interface, and the Java Multimedia service Application Program Interface (JMAPI) Interface. In addition, new terminal/server agent functions have been developed for use between multimedia terminals and servers. The lower levels of this architecture use existing CTRON technology, such as the basic operating system and the general extended operating system (Fujitsu Ltd.'s ROSEC). Thus only the upper layers of the MNP and the terminal/server agent functions are actually being constructed.
As for transmitting CTRON to China, Mr. Ohkubo said that the translation of CTRON documentation from Japanese is well under way in Shanghai, where a Chinese TRON Association has been established. Moreover, even the documentation concerning the MNP architecture is being translated, so China will probably be the second nation in the world to benefit from the fruits of the MNP project.
TRON VLSI CPU
Nowadays we don't usually hear much about the TRON VLSI CPU, and the topic is almost never mentioned at TRON Project international symposiums. However, this year we got an earful of comments from Prof. Ken Sakamura on what it was like designing the world's first open 32-bit microprocessor architecture. Prof. Sakamura said that the impetus to develop an open 32-bit microprocessor was that the companies participating in the TRON Project wanted something to run the ITRON-specification operating system on. There were also other factors, such as the IBM industrial spying scandal and the general perception that Japanese electronics companies couldn't do anything on their own except copy other people's ideas. And so in 1985, the TRON VLSI CPU subproject was begun.
Unfortunately, developing the microprocessor architecture wasn't that easy due to conflicting requirements. First, there was the problem of a key design parameter, i.e., how bytes should be ordered in the words the microprocessor was to process. Should they be big endian (with the most significant bits on the left) or should they be little endian (with least significant bits on the left). The semiconductor manufacturers wanted the former to maintain compatibility with existing mainframes, while Prof. Sakamura wanted the latter due to the variable length of TRON VLSI CPU instructions. The former won out, which created data compatibility problems within TRON Application Data-bus (TAD) when BTRON-specification operating systems were implemented on Intel microprocessors, which are little endian. On top of that there was a strong desire initially to use TRON VLSI CPUs in a wide variety of applications, such as personal computers, workstations, and minicomputers. That required a general purpose design. And then there were implementation problems--implementations appeared that differed from the TRON VLSI CPU specification.
But Prof. Sakamura is not unhappy with the results of his efforts. He noted that there were nine implementations of the TRON VLSI CPU architecture. One of them, the Gmicro 500, appeared in 1993, the same year the first Pentium appeared. Although they are both similar in design, at 66MHz the Gmicro 500 is capable of 130 MIPS, while the Pentium is only capable of 112 MIPS, making it 16 percent faster than the Intel chip. The Gmicro 500 along with the Gmicro 300, incidentally, are widely used in Nippon Telegraph and Telephone Corp.'s switching exchanges, and the Gmicro 200 was used in Japan's "Kiku 7" experimental satellite. Gmicro 100s and several other TRON VLSI CPU implementations are widely used in embedded applications. Perhaps best of all from Prof. Sakamura's point of view is the fact that the International Organization for Standardization (ISO) selected the TRON VLSI CPU system bus as an international standard for supporting multiprocessor fault tolerant systems. It is officially known as ISO/IEC 14576 (STbus), and it is the first international standard to come from Japan in this field.
Electronic Commerce and the Disabled
Everyone knows that the age of electronic commerce has arrived, but very few people have ever thought about how it will affect the disabled, in particular the visually disabled. That topic was taken up as the theme of this year's TRON Enableware symposium, which is officially known the TRON Electronic Prosthetics Symposium (TEPS). This year's symposium was unusual in that in addition to members of the disabled community, three representatives from organizations currently in the process of implementing electronic commerce were also on hand. These were: Mr. Takanori Ninomiya, secretariat of the Debit Card Promotiom Association; Mr. Juichiro Hirohata, chief of Seibu Department Store's Customer Planning Department; and Mr. Toru Takahashi, chief of the Ministry of Posts and Telecommunications Postal Savings Bureau's Operation Planning Section.
The TEPS '99 program began with reports by two members of the visually disabled community, Mr. Sadao Hasegawa and Mr. Tetsuya Fukui. Mr. Hasegawa, who is a well known specialist on Braille, outlined the problems that the visually disabled have when using debit cards. The first is that the visually disabled have a hard time using touch panels, and hence they would like to use keypads for inputting numbers into card reading terminals. The second is that even when there are keypads, there is no standard layout; both calculator and telephone layouts exist. Moreover, another problem is the location of the correction and enter keys, which vary according to machine. And finally, there is the problem of audio feedback to confirm input. To remedy these problems, Mr. Hasegawa made the following proposals:
Mr. Fukui said that automatic teller machines (ATMs) are difficult for the visually disabled due to their design, not the handicap of the visually disabled persons using them. Like Mr. Hasegawa, he favors the telephone keypad layout, and he noted that the Ministry of Posts and Telecommunications put Braille on their ATMs 15 years ago. This has enabled the visually disabled to utilize any of their ATMs throughout Japan. He said the debit card association is now listening to their problems, although he stressed the need for guidelines at the national level to solve this problem. This is something that has already taken place in the U.S., where there are laws that have come into effect to enable access by the handicapped. He also showed how easy it is to provide voice feedback for the disabled. He held a small calculator aloft, input some numbers, and the calculator, which was equipped with a voice output chip, read them off.
Mr. Fukui's presentation was followed by that of Mr. Yoshitake Misaki, who is a trainer of the visually disabled in Hachioji, Tokyo. As an example of ignoring the visually disabled when designing equipment, he introduced the example of the JR East's ticket vending machines, which have low contrast touch panels screens that are particularly difficult for the visually disabled to use. They have been criticized in the press, and discussions are now under way with management to make changes. His presentation was followed by that of Ms. Eiko Tachimatsu, who works with the mentally handicapped and described the problems the mentally handicapped have have using ticket vending machines and the like in society. Such disabled persons cannot handle input operations that require a lot of time or require complicated input procedures. Thus they need to be able to use non-mechanical means when traveling to school on their own.
After the members of the disabled community gave their presentations, there was a panel discussion moderated by TRON Project Leader Ken Sakamura who mentioned that the Americans with Disabilities Act requires the U.S. government to purchase equipment with provisions for the handicapped. Mr. Fukui strongly supported the idea of the Japanese government getting involved by enacting such legislation. Mr. Hasegawa brought up the topic of standards, Japan Industrial Standards (JIS) standards in particular. According to JIS standards, there is no fixed place for "zero" on keypads. Accordingly, there are eight or nine different types input panels the handicapped must interface with when those with/without touch panels are taken into consideration. However. Mr. Misaki noted that it would be impossible to change all the touch panels at once due to the costs involved, so it must be done over time. Prof. Sakamura noted that mass production can reduce such costs; he also took time to introduce the "TRON Human Interface Standard Handbook," which includes exactly the guidelines that Mr. Fukui and Mr. Hasegawa talked about. However, Mr. Fukui also pointed out that some people want to talk to people in the train station, and that in the true sense multimedia means not just touch panels, but also various types of interfaces for input.
The first panel session was followed by presentations by the representatives of electronic commerce organizations. Mr. Ninomiya of the Debit Card Promotion Association demonstrated a debit card and two card readers, one handheld and the other on a stationary mount. He said he plans to transmit the needs of visually disabled people to his association so that they can be incorporated into guidelines that will be drawn up in October 1999. Mr. Hirotake of Seibu Department Store gave the second presentation. He said he believes everyone will be using debit card service in the future, including even minors and the unemployed. He said debit card readers at Seibu have had transparent Braille seals put on top of input keys for use by the visually disabled. Mr. Takahashi of the Ministry of Posts and Telecommunications, which runs banking operations out its nationwide network of 24,600 post offices, said the reason that there have been inadequate preparations for the visually disabled is because of two problems. The first is that people did not know; the second is that even though some people may have known, they were unable to respond due to competitive pressures. He said not knowing was probably the biggest problem, and that he would like to respond as soon as possible. He also noted that there were a surprisingly wide range of opinions among the disabled themselves.
After the three representatives gave their respective presentations, they held a panel discussion among themselves, which was moderated by Prof. Ken Sakamura. Mr. Hirotake said he wants to responds to the needs of the disabled from here on. Prof. Sakamura noted that if you try to do something after the machines are made it is expensive, but if you do it from the beginning it is not expensive. Mr. Ninomiya said that there are various committees at the debit card association, but the association will definitely handle this problem. However, certain things such as voice output will take time, since there are security considerations involved. Mr. Takahashi said there is a limit to how well it can be solved. Prof. Sakamura responded saying he believes the American approach, in which it has been made a law that you cannot sell something to the national government unless it is handicapped accessible, is a good one. However, Mr. Hirotake retorted that competitive forces will force companies to help the handicapped.
TEPS '99 ended with a combined panel session in which all of those who made presentations took part. This also was moderated by Prof. Ken Sakamura. Mr. Hasegawa, the first to speak, noted that a lot of people don't know about the problems of the visually handicapped, but he also pointed out that the Ministry of Posts and Telecommunications helped out at the trials in the city of Omiya to the north of Tokyo. Mr. Fukui said that people's perceptions on the issue are different. For example, as to the issue of voice output, he said he believes voice output capability should be cheap. Moreover, even though current guidelines call for non-voice output of PINs due to security concerns, if a clerk has to help a visually disabled person confirm his/her PIN, then security will likewise be compromised.
Mr. Ninomiya, clearly rattled by his direct contact with the visually disabled, said the problems involved are not something you can understand through reading a written explanation or a textbook. He said you have to close you eyes to figure out and understand what the visually disabled are saying. Mr. Hirotake said that there are things that can be done to help the visually disabled, such as unifying the key layout of the input panels on the terminals. He said he would immediately write a report to his CEO with suggestions as to how to do something concrete to help. Prof. Sakamura, obviously impressed with the efforts of the U.S. government's efforts to help the disabled, said that having a law such as the ADA would not be such a bad thing. He also said there is a need to look into a voice input chip.
The greatest moment of the the Enableware symposium, however, came right at the end. When the moderator asked the members of the audience if they had any questions, two young Japanese electronics engineers stood up in succession and began asking a long series of questions as to how it would be possible to help the disabled. This was a truly remarkably development. After over a decade of Enableware symposiums, the TRON Project had finally brought together the disabled, a government official, industry leaders, and Japan's up and coming engineers. What a remarkable development! The TRON Project was serving exactly as the technological catalyst it was intended to be!!