r&d

The TRON Project, 1995

One of the most talked-about Japanese research initiatives of the late 1980s, the TRON Project has received relatively little media attention in recent years. This has led many to believe the project is no longer being actively pursued, but that is far from the truth. In November, Computing Japan visited Professor Ken Sakamura, the leader of the project, at his lab in the University of Tokyo to talk about the current state of the TRON Project.

by Steven Myers

Tron (the Real-Time Operating-System Nucleus) is a large-scale, long-term effort aimed at creating a completely new, all-encompassing computer architecture for the "computerized society" of the future -- one in which all tools, appliances, and other objects are interconnected by computer networks. The main focus of the project is to provide a real-time environment for the large number of distributed computers embedded in these "intelligent objects." This type of environment is referred to by the TRON developers as a highly functionally distributed system (HFDS).

The TRON project was initiated in 1984 by Professor Ken Sakamura of the University of Tokyo, and it is funded entirely by an industrial consortium called the TRON Association. The project generated a fair amount of controversy in its early years. It was perceived by many computer professionals in the US to represent an attempt at implementing unique Japanese standards that could become worldwide industry standards if adopted by the large Japanese electronics companies. Many TRON supporters, on the other hand, felt that their efforts were being misconstrued in the English-language press, leading to more misunderstanding.

For whatever reason, it seems that a variety of factors unrelated to the technical merits of the project have contributed to resistance outside of Japan. The purpose of this article is not to judge the motives or social significance of the effort, but rather to simply introduce the basic conceptual and technical content of the TRON project and provide an update on its progress.

The TRON subprojects

From the beginning, the TRON project has included five fundamental subprojects, each of which focuses on one component of the entire HFDS picture. These subprojects include CHIP (TRON-specification VLSI microprocessor architecture), ITRON (real-time operating system specifications for embedded systems), BTRON (an architecture for PCs and workstations), CTRON (interface specifications for communications systems), and MTRON (network operating system architecture to handle interconnection and control of the above systems). A sixth subproject, called "TRON Electronic Equipment HMI," has also been undertaken with the goal of formulating the TRON Human-Machine Interface specifications.

Four of these projects -- CHIP, ITRON, BTRON, and CTRON-- have been highly developed. Each of these subprojects has already produced both detailed specifications and products. MTRON is still in the basic research stage, but it is becoming more active now that some of the necessary components from the other subprojects are available.

ITRON (Industrial TRON) is a specification for a real-time, embedded-system operating system for use in a wide range of applications, ranging from robotics and automotive systems to consumer products such as VCRs. The actual implementation of ITRON specifications is carried out independently on each type of processor, and many of Japan's semiconductor manufacturers have implemented these specifications for a variety of chips appearing in all types of commercial products.

The BTRON (Business TRON) subproject is aimed at exploring human-machine interaction and the incorporation of the HMI (human-machine interface) into the operating system (OS) specification and hardware design of personal computers and workstations. So far, three sets of formal BTRON specifications have been developed, each intended for a different class of hardware. These specifications are BTRON1, designed for a 16-bit CPU and relatively small-scale hardware; BTRON2, meant for 32-bit CPUs and more powerful hardware; and BTRON3, designed to support a loosely coupled distributed environment. Several OSes conforming to the BTRON specifications have been implemented -- most notably by Matsushita Electric and Personal Media -- and in 1994, a BTRON1-specification OS capable of running on DOS/V computers appeared on the market.

The CTRON (Central/Communication TRON) operating system interfaces are geared specifically toward switching and advanced telecommunication networks. Recently, special emphasis has been placed on the application of CTRON to multimedia communication services, investigating the real-time exchange of image and sound data over networks.

The figure on page 14 shows how the different TRON subprojects are integrated to form the entire HFDS. As shown in the diagram, the CHIP (TRON VLSI) CPU is ideally suited and optimized for running ITRON-based embedded system OSes and BTRON PC/workstation OSes. It is important to note, however, that this is not required; virtually any part of the TRON architecture is capable of interacting with existing architectures.

The Sakamura lab

The visionary and driving force behind the TRON project is Professor Ken Sakamura of the University of Tokyo, an engaging and energetic computer scientist/architect who has managed to attract huge support for TRON from Japanese industry and academia. Computing Japan visited Professor Sakamura at his University of Tokyo lab in mid-November, one week before the Twelfth TRON Project International Symposium was scheduled to begin in Tokyo.

He began by presenting a basic introduction to the project and some of the "application subprojects" designed to test actual HFDS implementations in order to find and break down potential problems in the TRON architecture. The best known of these application subprojects is the TRON-concept Intelligent House, which incorporates over 1,000 TRON-based computer systems. Located in Tokyo's Nishi-Azabu district, the house was used for three years of experiments, ending in 1992. Currently, design work is nearing completion on the TRON-concept Intelligent Building, with construction scheduled to begin soon.

Prof. Sakamura spoke strongly of the need to move away from the "desktop PC" computing paradigm -- with its screen and keyboard interface -- to other, more natural interaction methods. This, he says, is the essential idea behind TRON: to create an environment for ubiquitous computers that is natural and unobtrusive for the users. The sharp increase in PDA (personal digital assistant) use over the past year is a step in this direction, says Sakamura, but considering the current media hype and flashy marketing techniques surrounding desktop PCs in Japan, he acknowledges that traditional desktop computers could remain popular for a while.

TRON is quietly but effectively making its way into industry products. According to Sakamura, there are millions of copies of OSes based on ITRON specifications in use, and countless electronics products now incorporate TRON-based processors. Sakamura points out that because TRON is basically about creating standards, and much of the resulting technology is imbedded within commercial products, the technology has made its spread without much of the media fanfare that is currently being lavished on PC systems and software.

There has not been much written about TRON in either the Japanese or foreign press in recent years, leading many to believe that the project has diminished in scope and significance. Not so, says Sakamura, pointing once again to the fact that, while other factors were involved, current media focus has shifted to multimedia and the Internet. In reality, he says, the TRON project continues to gain momentum as internal components based on TRON specifications quietly gain widespread acceptance in Japanese products and systems.

Alive and well

Media reports to the contrary, not only is the TRON project still alive, it is flourishing. This year's TRON conference promises to be one of the largest ever, with the majority of the papers focusing on actual implementation issues.

The TRON subprojects have produced a wealth of hard, practical technology, the elements of which are appearing with ever-increasing frequency in commercial products. Simply put, among scientists, researchers, and technologists with an interest in computing standards and architectures, the TRON project deserves serious attention.ç




(c) Copyright 1996 by Computing Japan magazine