is today’s information technology smart enough for a smart world?

23
| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions Is today’s Information Technology smart enough for a smart world? M2M Summit 2016 - Düsseldorf Joachim Hoernle Bull BES Business and Enterprise Systems

Upload: m2m-alliance-ev

Post on 16-Apr-2017

16 views

Category:

Technology


0 download

TRANSCRIPT

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Is today’s Information Technology smart enough for a smart world?

M2M Summit 2016 - Düsseldorf

Joachim Hoernle

Bull BES Business and Enterprise Systems

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Today’s agenda

2

▶ From Smart X,

▶ Smart Systems,

▶ Smart Data Integration to

▶ Smart Factory: ScaleIT

▶ Q&A

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

BES Business Portfolio

3

Focus Areas and Expertise

IT operations, IT operational safety, IoT Management and Data Integration

– Management

– Monitoring

Business Modell

– Off the shelf software solutions

– Custom solutions

– Respective services

• Consulting and

• Implementation projects

• Trainings

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Smart Is The New Green

4

Smart Factory Smart Home

Smart Grid

Smart Cities

Smart Material Smart Health

In future literally every - thing will be smart. In future literally every - thing will be smart.

Smart X

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Is There Any Smart Definition?

5

▶ There are several definitions of “smart” floating around.

▶ Typically Smart Systems / Objects – have some sort of intelligence, the ability to learn

and to deal with or understand situations especially if they are complex, non-standard or problematic.

– some kind of interaction between the smart object or system and the ambience, environment or physical context.

– are pervasive and ubiquotous.

– things or systems have some kind of autonomous behavior.

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

How to transforms a Thing to Smart Thing?

6

And many other aspects - Identity / Discovery - Security - Lifecylce - Usage data - ...

Communication Communication Communication Communication

Self Mgmt.

„Intelligence“ „Intelligence“

Knowledge Base

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Smart Things require Smart Data

7

Meta Meta Data

C C C C

C C C C

C C

C C

C C C C

C C C C

Smart Smart Data

• Time • Location • Accuracy • Value range • Vendor • ...

Data

Data describing the context • Process • Order • Lot • ...

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Smart Systems

8

▶ Smart systems typically consist of diverse components which are related to the basic capabilities of the system:

• Sensors for signal acquisition

• Actuators that perform or trigger the required action

• Some kind of knowledge base

• Networking to transmitting information and decision and instruction to the command-and-control unit

• Power Storage and Energy Management

• …

▶ In addition there are some capabilities which are mandatory

– Integration / information integration / data integration

• Low scale integration – addressed by Smart Systems Integration and similar approaches

• Large scale integration – currently in the clouds

– Management

• Operations management for the smart world

– Monitoring and control, security, identity, network management

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Typical Solutions

9

Cloud

Backend

Data Souces D

C

Solution 1

D

C

Solution 2

D

C

Solution 3

D

BE

Solution 4

D

C

Solution 5

D

C

Solution 6

C

BE

M Model

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Classic Data Integration

10

▶ Old hat: Data integration is an established discipline in IT since many years

▶ Classic data integration is an approach which is typical in enterprise IT.

▶ Strong repository and database focus

▶ Objectives

– Ability to cope with complexity and with inconsistencies at various levels

• Reduce the number of i/fs - provide uniform access to data from multiple sources

• Integrated system illusion

– Facilitate re-use

– Ensure interoperability and provide independence from

• data source specific aspects such as interfaces or hardware: technical DI

• specific representation of information: syntactical DI

• from specific schemes: structural DI

• from specific contextual information: semantic DI

▶ Many different approaches, technologies and tools such as e.g.

– EAI – Enterprise Application Integration

– ETL – Extract, Transform, Load

– EII – Enterprise Information Integration

– ESB – Enterprise Service Bus

– MOM – Message Oriented Middleware

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Classic Data Integration Issues

11

▶ Classic data integration is often complex, cumbersome and costly. Mainly because of the complexity people tend to “divide and conquer”

▶ Therefore data integration technologies are often limited to a small subset of data sources.

▶ Many steps for cleaning, enrichment, matching and fusion of data have to be performed manually.

▶ Often people do not distinguish between different types of benefits

– Benefits for technology : ease of IS management or creation of IS

– Benefits for end-user: use of concepts and terminologies from the end user domain

▶ The bad news is: there are not many good example for successful integration initiatives in IT especially if the subject is large, heterogeneous, complex, polymorphic and dynamic as it is in the “smart world”.

▶ To address the requirements of digitalization initiatives it is not enough to focus on a subset of aspects of data integration (technical, syntactical, structural or semantic integration) or to provided powerful but scattered integration approaches or just technology. Smart system dealing with smart data require an

– holistic, meta data aware and model based data integrative approach focusing on the end user domain and includes an

– integration architecture and

– provides the appropriate tools and facilities.

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Smart Data Integration

12

▶ Smart data integration focusing on digitalization initiatives has additional and different requirements.

▶ For instance it is important to support and facilitate the collaboration of experts from different domains e.g. electrical engineering or software engineering. Experts tend to use different tools, which are well suited for their specific purpose, but usually do not provide sufficient mechanisms for cooperation with other engineering tools. Especially cross domain integration is both, critical and problematic.

Integration of the conceptual / engineering models Integration of the conceptual / engineering models

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Smart Data Integration

13

▶ In addition to generic requirements related to data integration digitalization initiatives requires the following data integration capabilities:

– Multi mode modeling

• Support for different models at the conceptual focusing on the same of similar domain

– Consistent and pervasive integration from the shop floor up to the level of the engineering tools or management

• technical,

• syntactical,

• structural and semantic integration

– Meta Data Management based on a standardized meta model

• Including lifecycle management of models and meta data

– Mapping and binding facilities

– Abstraction, aggregation and enrichment of information

– End user suitability of the modeling tools

• The major focus is the engineering domain not the IT domain

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Data Integration Requirements in a Smart World

14

Heterogeneity Heterogeneity

Extensibility Extensibility

Holistic Approach Holistic Approach

Real Time Real Time

Low Effort Low Effort

Data

In

teg

ratio

n Scalability Scalability

Req. Classic DI Req. Classic DI

End User Enabled End User Enabled

Plug & Work

M2M

Predictive Maintenance

Lot 1 Production

Data and Meta Data Data and Meta Data

Scenarios / Use Cases Requirements

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Data Integration for Digitalization in Production

15

Mech. Engineering Mech. Engineering

Shopfloor Managemen

t Managemen

t

Elec. Engineering

IT Engineering Data

Integration

Design Design

Planning Planning

Engineering Engineering

Production Production

Service Service

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

ScaleIT

16

▶ The ScaleIt project is an „Industrie 4.0“ project funded by German government (BMBF).

▶ The focus is to provide an architecture and components of a scaling ICT for increasing productivity in mechatronics manufacturing.

▶ https://scale-it.org/

▶ Project partner

– Sick AG

– Zeiss 3D AG

– RoodMicrotec GmbH

– Smart HMI GmbH

– Ondics GmbH

– FEINMETALL GmbH

– digiraster GmbH

– Bull / Atos GmbH

– University Stuttgart

– Fraunhofer Institute IAO

– Karlsruhe Institute of Technology

– microTEC Südwest e.V.

▶ Scalability in terms of the number of components or smart systems but also in terms of technologies, approaches and standards

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Data Integration Focus

17

Knowledge

Knowledge

Information Information

Data Data

Meta Data

Meta Data

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

From Data Integration to Data-Morphosis

18

Data Acquisition

EAI/ESB

ETL

Syntactic Integration

XML Technologien

Semantic Framework

Semantic Consolidation

Ontologies, SPARQL, RDF

Semantic Framework

Data Information Knowledge

Analytics

Rule based Systems

Industrial Intelligence

Insights

Semantic Integration

Syntactical Integration

Technical Integration

Holistic Model

• Semantic Annotations

• Binding • Mapping - Abstraction - Aggregation - Enrichment

Meta Data Repository - Dependencies - Relationships

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Data Integration Architecture

19

Semantic and Structural Data Integration

•Engineering models, standards ...

•Based on ontologies focusing on the end user domain or standards

Semantic and Structural Data Integration

•Engineering models, standards ...

•Based on ontologies focusing on the end user domain or standards

Syntactical Data Integration

•Independence from formats and language

•JSON, CSV, DSLs, AutomationML, ...

Syntactical Data Integration

•Independence from formats and language

•JSON, CSV, DSLs, AutomationML, ...

Technical Data Integration

•Independence from interfaces and respective technologies

•OPCUA, MQTT, COAP, fieldbusses, ...

Technical Data Integration

•Independence from interfaces and respective technologies

•OPCUA, MQTT, COAP, fieldbusses, ...

IT IT

Data sources on the shopfloor or in IT

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Engineering Model

20

Example: Conceptual Engineering Model

Technical Object

References Meta Data

• Unit • Accuracy • Location • ...

• Process • Organization • Order • Customer • Product • Product

component • IT System • Failure • …

• Sensor • Sensor Node • Machine • ...

Semantic Annotation

Semantic Annotation

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

Federation Concept

21

Technical Data Integration

OPC UA Ontology

MQTT Ontology

COAP Ontology

Vendor Specific Ontology

• Discovery • Identity Management • Property Management

Syntactical Data Integration

Adapter

MES / ERP

Ontology

| 05-10-2016 | Joachim Hoernle | © Bull/AtoS Bull/AtoS | Bull Software Solutions

The Big Picture

22

Modell Mgmt.

Technical Data Integration

OPC UA Ontology

MQTT Ontologie

COAP Ontology

Vendor Vendor specific

Ontology

Syntactical Data Integration Utility

Ontologies

Mapping Ontologies

Semantic Data Integration

FMEA Ontologie

Engineering Modell

Abstract Abstract Sensor

Ontology

User Ontology

Representation Layer

Testing Ontology

Ontology Ontology encapsulating

a DSL

xxxML Ontology

Concrete Concrete Sensor

Ontology

Security Ontology

Mgmt. Models

Atos, the Atos logo, Atos Consulting, Atos Worldgrid, Worldline, BlueKiwi, Bull, Canopy the Open Cloud Company, Yunano, Zero Email, Zero Email Certified and The Zero Email Company are registered trademarks of the Atos group. February 2015. © 2015 Atos. Confidential information owned by Atos, to be used by the recipient only. This document, or any part of it, may not be reproduced, copied, circulated and/or distributed nor quoted without prior written approval from Atos.

01-08-2016

Thanks

For more information please contact:

M+ 49 170 34 26 975

[email protected]