Systems Approach to Computing Dependability In and Out of Hitachi

Systems Approach to Computing Dependability In and Out of Hitachi Concept, Applications and Perspective

IFIP-WCC2004 Topic 3 In honor of Emeritus Professor Algirdas Avizienis

Hirokazu Ihara Hiro Systems Laboratory [email protected]

Motohisa Funabashi Systems Development Laboratory, Hitachi, Ltd. [email protected] 1 Sculpture by Ryoji Goto

Inherent Property of Control Computing Compared to Enterprise Computing (a) Reliability is prior to its performance and cost. (b) Environment conditions are very rigorous (ambient temperature, humidity, vibration, dust, etc.) (c)Event-driven or data-driven processing from the machines (d) Time-critical (e) The turnkey or the solution business (f) Periodically to adapt their functions by changing demands Two Business groups for Enterprise and Control Computing has been operated in Hitachi Ltd. 2

Evolution of Computing Systems Requirement

System Structure 1960s 1960s

Performance

Expandability & Versatility for Manufacturer

Centralization 1970s

Feature Time-sharing Tight Connection

Shutdown

Distribution

Function-sharing Strict Interface

Partial Cutoff Extendability Dependability in Operation

1980s

Decentralization 2000s

On-line Metamorphosis

Autonomous Function Receiver Initiative

Autonomous Reconfiguration

Multi Layered Decentralization

Emergent Function Multi-modal Initiative 3 3

Our Target of System Concept Autonomous controllability

Hierarchical Distributed

Autonomous Decentralized

Fault tolerance

Cost performance Coordinability

System performance Autonomous Coordinability

Fault avoidance

Centralized

Controllability

Functional Distributed 4

Dual Structure System of COMTRAC COMputer aided TRAffic Control System input DSC (Dual System Controller) Constant value Computer A

Register

Register

Comparator Interrupt

Computer B

Interrupt

Judgment AND

System output 5

Symmetrical Structure of COMTRAC Dual –Duplex System Structure

DSC;dual system controller

DSC System input

Computer A

STR

Computer B

DSC

DSC Computer C

STR;status register

System output 6

Present Dependable COMTRAC Network of Shinkansen has been developed line by line since 1964 and operated by COMTRAC since 1972 without any system failure.

Akitata Line

Sanyo Line Hakata Kumamoto

Yamagata Line Akita Jyoetsu Line Yamagata Okayama Osaka OsakaNagoya Niigata Nagoya Nagano Nagano

Kagoshima Kyushu Line 2004

Tokaido Line 1972

Hachinohe Morioka Fukushima

Tokyo Tohoku Line

Nagano Line

7

Principle Recognition of Autonomous Decentralization Concept (ADC) (a)

Always includes inactive (temporary faulty, complementary or spare) parts (b) Always changes its conditions and states among operation, metabolism, generation and growth (plus or minus) (c) Always changes its objectives to the goal by alternatives selection, optimization and daptation (d) Always keeps accomplishing its objectives almost completely This observations are opposite recognition against that system should be complete and stable. 8

Properties of ADC Uniformity of structure

Equality among subsystems

Locality of information

Definition of ADS (a) Autonomous controllability: if any subsystem fails, the other survivors can manage themselves (b) Autonomous coordinability: if any subsystem fails, the other survivors can coordinate their individual objectives among themselves.

9

Virtual Data Field by Broadcast VDF (Virtual Data Field) ACPs distinguish their necessary information by Content Codes

Broadcasting networks ACP

ACP

ACP

Autonomous control processor(ACP)

ACP

Subsystems

Elements of Typical Information Format Flag Flag

Content CC Code

Sender Sender Control Control ID Code ID code

Data/ Data or Command Command

RC

Flag 10

Connection of Autonomous Decentralized Systems a2

ACP

L1

b1

a3

b2

c1 b3

a1 c2

ACP c3

L2

Analogy of Nerve

Analogy of Brain Tightly coupled Hexagonal connection

Loosely coupled Double loop network ACP

ACP Sculpture by Ryoji Goto

11

Autonomous decentralized Transport Operation control and information System (ATOS) introduced since 1996 in Tokyo Metropolitan Wide Area Jinbohara 2,200 computers

Kuroiso Omiya

Ikebukuro Kofu

Tachikawa

Shinjyuku

Hatori Ueno

Ochanomizu

Chiba

Tokyo Atami

Ofuna

Yokohama

Shinagawa

Kawasaki Yokosuka

19 lines 300 stations

6,500 trains Total line 1,200km 12

Our Target of Remained Area of Computing Degree of structuring high Non-linear Strong Non -linear Complex Problems Problem Problem Solving Systems Solving Systems Search Engine abstract Formula, Axiom, Law

Human Brain & Intelligence

segmentation linearization -

Syntactics - oriented Semantics -oriented

-specialization --domain

Common/Specific Knowledge Perception, Understanding Selection, Analogy, Association

Linear or Simple-non linear Problem Procedural Processing systems

Analytical Engine specific Ad hoc, Matching

Knowledge Conquer Intelligent Information Processing Systems

poor

Inference Engine

13

Evolutional Practice of AI Technology Attempts in early 1980s Diagnosis of Nuclear Power Plant Quality Control of Semiconductor Manufacturing Project Management of Oversea Construction of Fossil Power Plant

EUREKA (Real-time Inference Engine)

Fuzzy Expert System for Subway Operation (1987) Management Methodology for Uncertainty

Real Applications Developed in 1980s Operation Guidance of Blast Furnace and Public Utility Plants Generation Assistance of Train Operation Diagrams Construction Planning and Project Management Support Assessment Support of Financial Investment Blast Furnace

14

14

Knowledge Acquisition Technology for AI Systems Gray Box Approach Bare Expertise

Knowledge Refinement By Neural Network

Applications for instance Consistent Expertise

Control Systems for Public Utility Plants Financial Trading Systems Moving Average

Data Warehouse

Black Box Approach Objective of Acquisition

Acquired Knowledge

Data Mining Prediction Model

 CRM (Customer Relationship Management) Systems in Telecommunication Operators  Quality Control Systems in LSI Manufacturing

15

Hierarchical Systems of Human-beings

Vertical Distribution

Disperse Hierarchical Layers and Equal Diversification

Nation City Community Family

EADC

Individual Organ

ADC Cell

Close Variety

Uniform Horizontal Distribution

16

Definition of Autonomous Observability for EADC Autonomous Observability: if any change occurs in the surrounding of the subsystem, subsystems can observe it immediately EADC is characterized by Autonomous controlability Autonomous coordinability Autonomous observability 17

Extended ADS Attached Autonomous Observability to ADS controllability autonomous

Two dimensional observability

Coordinability

Pre-install Off line

Heuristic autonomous coordinability

autonomous observability controllability

On line Four dimensional (Space & Time)

18

Emerging Players for Autonomous Observability Antenna Built-in Type RFID

Sensor Network Devices

µ-chip 2.45GHz RF antenna

0 .4

mm

0

m .4

LSI(Wireless + Controller)

m

Prototype

µ-TAS (Micro Total Analysis System) for Health Monitoring

Sensor 10mm

1mm

Vibration based Generator

Micro Channel

Micro Separation Devices

R. Amirtharajah et al. of MIT (1999) 19 19

Super Distributed Objects: AYA(context Aware & Yet Another service)  Ad hoc Collaboration according to the Context  Collaboration among AV, Facility, and OA Systems with Autonomous Observability Voice Video Listening via Mobile in case of Outdoor UPnP World

Object Model: AYA PC Voice Video Video

PC

Server Functions -Monitoring -Content Delivery

Air Sensor Conditioner

Viewing via TV in case of Home HDLA World

Quiet and Energy Saving in case of Sleeping ECHONET World

ᗀ࿈䠌…

Mobile World

OA World

20 20

Autonomous Group Management for ITS based on Autonomous Observability Reliable Data Transmission for High Speed Cars through DSRC (Dedicated Short Range Communications) by Autonomous Group Formation of Base Stations ABS

ITS: Intelligent Transport Systems

ABS ABS

Service Providers Data Field

ABS

ABS

ABS

ABS

Autonomous Group Formation Service Control Autonomous Service Delivery

Dynamically Formed ABS (Autonomous Base Station) Group according to Car Position

21

Open Development and Operation Infrastructure Overview of Emergent Synthesis Environment (ODOI) Domain engineers

requirement

development

operation

Intelligent engineering system Life Cycle system Dynamic management system Simulation

Simulator

System knowledge

Intelligent collaboration tools

Life cycle system model Model view

Integration,Analysis,Optimization Information Base Systems engineers

Outside expert/ Citizen 22

Aims of Open Development & Operation Infrastructure (ODOI) For project members to retrieve, accumulate, suggest, simulate, validate and synthesize comprehensive information and knowledge for the phases of plan, design, validation and operation For project members and experts outside to communicate, exchange, discuss and interact recursively For people interested in including citizens to watch recursively the on-going process and proposed new idea or suggestion For developed systems to have autonomous operation supported by ODOE 23

Sample Scenario for Yaoyorozu Society in 2010

Health Diagnosis and Advice

Augmented Memory Services

My Remembrance Memorizing Service

Pervasive Information Society (Yaoyorozu) in 2010 (Annotated World by Small Smart Devices)

P2P Exchange

Reminder Service

3rd Party Service

Your Remembrance

Memorizing Every Personal Health Conditions, Perceptions, and Social Interactions (Conversation, Mail, Phone, …)

Memory of Lights

Memory of Roads

Memory of Benches Small Smart Devices

24

Yaoyorozu in Japanese Myth “Yaoyorozu” (= eight million) means countless in number, particularly in the phrase “Yaoyorozu no Kami-gami, or “eight million gods”, indicating that gods live not only in the many shrines, but in trees and stones, in the sky and water, constantly surrounding and protecting us. For our project, we use “Yaoyorozu” in place of Ubiquitous for the heterogenious intelligent devices surrounding and protecting us.

Eight Million Gods in Japanese Myth

Home & Community

Information & Media Production

Secure and Affluent Society by Convergence of IT, Bio, and Nano

Material Production

Public Facilities

25

The Yaoyorozu Project Project Organization

Humanity (Lifestyle, Ethics) Systems(TransDisciplinary Science & Technology) Engineering (IT)

Research Goal - Desirable Institutional Systems & Core Technology for Ubiquitous (Yaoyorozu) Information Society in 2010 - Research Methodology for Multi-disciplines (August 2002-March 2005)

Supported by Special Coordination Funds for Promoting Science and Technology by MEXT (Ministry of Education, Culture, Sports, Science and Technology, Japan)

26

Our Research Attitude

㻤㼏㼗㼋㼒㼘㼊㼋㻃㼒㼘㼕㻃㼏㼌㼉㼈㼗㼌㼐㼈㻃㼐㼄㼜㻃㼑㼒㼗㻃㼖㼓㼄㼑㻃㼄 㼋㼘㼑㼇㼕㼈㼇㻃㼜㼈㼄㼕㼖㻏 㼚㼈㻃㼋㼄㼙㼈㻃㼆㼒㼑㼆㼈㼕㼑㼖㻃㼒㼉㻃㼄㻃㼗㼋㼒㼘㼖㼄㼑㼇㻃㼜㼈㼄㼕㼖

Calligraphy by Mr. ODAIRA Namihei (Founder of Hitachi, Ltd.) 27 Sculpture by Ryoji Goto