Summary of 2nd Fundamental Information Technology of the Next Generation (Real World Computing) Promotion Committee Meeting

The 2nd Fundamental Information Technology of the Next Generation (Real World Computing) Promotion Committee Meeting was held at the MITI on July 15, 1998 (Chairperson: Prof. Hidehiko Tanaka, University of Tokyo).

This committee was set up as a private advisory board for the Director-General of Machinery and Information Industries Bureau at MITI for discussing and determining basic policies and what should be done during the second phase of the RWC Project (FY1997-FY2001). In the first meeting held in 1997, the Master Plan for the Fundamental Information Technology of the Next Generation was produced.

In this second meeting, additions and modifications to the Implementation Policy of the Master Plan were made after reviewing the project, which is already one year into the second phase, the Operational Plan for fiscal 1998 was discussed, and the results of fiscal 1997 were reported. The summary of the session follow (see the last page for the list of Promotion Committee members).

Opening Speech

At the start of the session, an opening speech was made by Mr. Masanori Yoshikai, Councilor, Machinery and Information Industries Bureau, MITI.

[Abstract]

The RWC Project is now its second year in the second phase, and a major part (approximately one fifth) of the budget for information-related projects in MITI has been allocated to this project. This is expected to contribute to the economy and society and attract much attention from overseas countries.

Since it is important to accomplish technical innovation both timely and efficiently, this Promotion Committee should focus on evaluating the appropriateness of selected and/or decided research themes, the organization for R&D, and the achievements thereof, and reflect them in resource allocation and bring them to market. More weight will be placed on ideas such as speeding up the process of selecting research themes or putting greater bias on resource allocation, and how promising they are for industry. This project has already produced many research achievements and is highly evaluated by overseas academic organizations, so we should ensure that this trend continues and publicly announce the significance of the project.

During next fiscal year, which falls in the middle of the 5-year second phase, the intermediate evaluation will be conducted.

For this, it is important to review the project as appropriate, in order to appreciate the rapid progress in technologies and run this project properly.

Additions and Deletions to the Implementation Policy of the Master Plan The additions and deletions to the Implementation Policy of the Master Plan were explained by the responsible members and approved by the Committee. Main modifications are:

Enhancement of optical technologies to catch up with the progress in network technologies (2 research themes were added):

1) Optical Interconnection for Large Throughput Data Bus (Optical Interconnection Hitachi Laboratory)
2) Optical Interconnection for Low Power Consumption Subsystem (Optical Interconnection NEC Laboratory)

Tsukuba Research Center Made Their Research Achievements Available to Industry (2 Laboratories were added):

1) Parallel and Distributed Systems Sumikin Laboratory (Theme: Application of Parallel and Distributed Computing Technology for Server Systems)
2) Information Base Functions Toshiba Laboratory (Theme: A Self-Organizing Architecture for Adaptive Cognition of Situations)

Partial Change To Research Theme (Real World Intelligence Center)

The Language Integration Laboratory at this Center is now responsible for the research theme "Multimodal Common Format and Integrated Environment".

Reports on Latest Research Results and Topics

The research results were reported by Mr. Aoyagi, Executive Director of RWCP, based on press releases, including:

1) Software for Gesture-Speech Interaction (Multimodal Functions TRC Laboratory)
2) RWC PC Cluster II (Parallel and Distributed System Software TRC Laboratory)
3) Parallel Protein Information Analysis (PAPIA) System (Parallel Application TRC Laboratory)
4) Long Wavelength Semiconductor Laser for Interconnections (Optical Interconnection Fujitsu Laboratory)
5) Sign Language Translation System (Multimodal Functions Hitachi Laboratory)
6) Multimodal Interaction System (Real World Intelligence Research Center: Applied Vision Laboratory and Interactive Intermodal Integration Laboratory)
7) Evolvable LSI Chips which can Autonomously Reconfigure Hardware Structure (Real World Intelligence Center: Evolvable Systems Laboratory and Real World Adaptive Device NEC Laboratory)

FY1997 Operation Report and FY1998 Operation Plan The FY1997 operation report and FY1998 operation plan were explained by Dr. Shimada (Director of the Research Institute, RWCP) and approved by the Committee, including:

[Basic Policies]

Focus on demonstrations of the research achievements to promote Japan-originated trends (such as demonstration of MMPC at AAAI-98 and demonstration of PC Cluster at SC'98)
Focus on finding users of the research results (especially, support for finding users of TRCｹs MMPC, support to use PC Cluster at Los Alamos Research Center in USA, and transfer of PAPIA System to bio-industry)
Start the design of a common format to represent multimodal information in a unified manner, essential for enhancing systems for confirming novel functions in the Real World Intelligence field (Language Integration Laboratory was newly established)
Speed up the development of heterogeneous seamless computing systems for the next-generation 10Gbps optical LAN while nurturing a multivendor environment for the computer industry (Increase investment on development of optical interconnection implementation technology)

[Executive Plans by Technologies]
Real World Intelligence

1) Multimodal Function
Increase gesture vocabulary and improve functions for integrated PCs that understand speech/gesture, and improve the sign language recognition method and implement it on PCs for sign language translation systems. 2) Autonomous Learning Function
Develop methodology for different learning function technologies.
3) Self-Organizing Information Base Function
Develop and implement a methodology based on multimedia database technology and mobile robot environment learning technology.
4) Theory and Algorithm Bases
Continued from FY1997, study mainly basic algorithms.
5) Real World Adaptive Devices
Develop prototype systems for static adaptive devices, dynamic adaptive devices, and digital smart pixels.
6) Intellectual Resources for Research and Development
Create a common corpus of RWC database, mobile robot experiment database, real environment speech/audio database, and tagged text database.

Parallel and Distributed Computing

1) Multiprocessor Computing Technology
Prototype and study applications of key technologies for multigrain parallelization and scheduling.
2) Seamless Parallel and Distributed Computing Technology
PC Cluster System Technology
Develop and prototype communication functions, programming, library, and user environment.
Optical Interconnection Technology
Prototype optical modules and optical devices including switches.
3) Parallel Application Technology
Mainly prototyping. Enhance PAPIA System for industry use.

[Opinions and Requests - Highlights -]

The following opinions and requests were offered by the members of the Committee on the above Operation Plan:

Since this project is technology-development-intensive, it should encourage the generation of venture business. Also, the results should be those which can be used widely.
The members of this Committee from universities should actively serve as consultants in accomplishing the researches. The RWCP in turn should actively consult with universities for advice on researches.
The technology development should provide results in specific fields which are attracting attention, with centralized rather than isolated resources for researches or a clear statement of the year of completion.
It is important to strengthen ties among industries, universities, and governmental agencies to feed the results back to industries. To accomplish this, we should first understand how other advanced countries manage this.
The results should lead to subsequent projects. Besides key technologies, significant systems should be constructed by organizing these technologies in a standard form.
The developed software should be published early to obtain feedback. New technologies originated in Japan should be made available worldwide.
The final image of the project should be clarified to promote applications in industries.
The Master Plan states Real World Intelligence and Parallel and Distributed Computing as the core of the project. While Parallel and Distributed Computing has a clear goal, Real World Intelligence does not, only with the statement "Development of Elementary Technologies". The goal of Real World Intelligence may be applications and sophistication derived from distributed researches or integration of the whole (to be feasible 5 to 10 years later). The entire project should stipulate how and what Real World Intelligence technologies will be implemented and reflected into society to facilitate budgeting for it.