cwin16 tls-faurecia predictive maintenance

21
Maintenance Predictive ou comment le Big Data révolutionne les usines du futur AIE Suresnes, 26 Septembre 2016 Capgemini, Capgemini Consulting, Sogeti HT

Upload: capgemini

Post on 21-Jan-2018

439 views

Category:

Presentations & Public Speaking


3 download

TRANSCRIPT

Maintenance Predictive ou comment le

Big Data révolutionne les usines du futurAIE Suresnes, 26 Septembre 2016

Capgemini, Capgemini Consulting, Sogeti HT

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 2

Table of Contents

Enjeux, contexte et bénéfices

Solutions techniques Big Data

Applications IBM PMQ et Braincube

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 3

Manufacturing Intelligence => Braincube

Predictive Maintenance => PMQ

Faurecia Digital Enterprise Project

3- Prepare

Rapid

Scale-Up

2- Experiment

and Learn

1- Explore &

Design

FEB. 2015 SEPT. 2015

200 digital use cases

40 Proofs of concept

9 solutions

Deploy

40 sites

Deploy

40 sites

END. 2018

2016 2017 2018

Pilot

6 sites

Industrialize

Deploy

14 sites

A systemic approach, at the speed of light

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 4

Digital EnterpriseManufacturing Intelligence & Predictive Maintenance

Big data benefitsWhy do we implement Big Data initiative?

Improve productivity

OEE*, Improve production flows, stock, …

Optimization cost of energy, utilities, indirect cost

Accelerate run at rate (loss of raw material, FMC)

Run Plant respecting standards

Reduceproduct

quality issues

Reduce scrap

Anticipation of non-quality with alerts and recommendations

Reduce key equipment

issues

Minimize unscheduled downtime and breakdowns

Manage business opportunities such as insourcing capacity

Increased equipment life cycle

(*) OEE stand for Overall Equipment Effectiveness (« Taux de Rendement Synthétique » in French)

ManufacturingIntelligence

Monitor production process in real time

And make decisions based on data

PredictiveMaintenance

Predict potential breakdowns of a machine

through data analysis and historian

2 families of Big Data tools in Operations

Monitor & alert in real time production parameters

Display tuning information to the operator on the shop floor

Keep production line stability for all shifts

Benchmark plants

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 5

Table of Contents

Enjeux, contexte et bénéfices

Solutions techniques Big Data

Applications IBM PMQ et Braincube

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 6

Commencer par démontrer l’intérêt d’une architecture Big Data

au centre de la solution globale via un pilote

A pilot…

In time boxing (3 months on Big Insights environment with plants data)

Thru simulated flow in a first step and then connected to plants

Real-time data flows implementation, reusable for industrialization

Analytics : demo of some possibilities

Manufacturing

Intelligence

(Braincube)

Predictive

Maintenance

(IBM PMQ)

Plants Plants …sensors sensors

1

2

3

3

4 4

1

2

3

4

IBM Cloud/Hadoop infrastructures

One shot data initialization

Real time simulation alimentation

Direct real time alimentation

3

2

5

5 Analytics & discovery

Open Data, External

Data, etc.

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 7

Définir l’architecture Big Data cible en fonction des besoinsArchitecture Framework for Predictive Maintenance

Simplified Architecture Functions and Technologies

❶ Data ingestion of

Ticketing Data and

Traceability Data

❷ Data storage of Process Data,

Traceability Data and Ticketing

Data

Ticketing Data

Traceability Data

SAP logs Other Data

❸ Processing to calculate KPI’s,

traceability and graphs

preparation

❹Visualization of

KPI’s

Predictive Maintenance

(IBM PMQ)

Us

ag

e

Analytics Visualization API /

Drivers

Str

uc

tura

tio

n

Processing SQLNoSQL

Sto

rag

e

Hadoop

HDFS

Warehouse In memory

Ing

esti

on

BatchMicro Batch Real time

1

2

3

4

1

2

3

4

❶ Real time ingestion of

Process Data from

Plants

❷ In memory storage of Process Data

❸ Trans-coding for PMQ

and Braincube

❹ Publishing to PMQ with Kafka and

Braincube with HTTPs

Manufacturing Intelligence

(Braincube)

Process Data

Kafka

Kafka

Kafka

Big

Insig

hts

3

5

❺Data Discovery

❶ Batch layer ❶ Stream layer

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 8

Retour concrets et intérêts du Big Data

❶ Single point of entry

- reduce the load on PCo side

- distribute the process data to all analytical

components

❷ Storage capacities

- centralization of data in one place

- available for any type of request from MI/PM

❸ Analytics & discovery

- computing power for custom analytics

- direct analytical functions

❹ Data Publishing

- compatible with current & new partners

- custom data visualization

Manufacturing

Intelligence(Braincube)

Predictive

Maintenance(IBM PMQ)

PCOOther Data

Big Data

TraçaStratos

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 9

Quelques visualisations possibles des données dans HDFS

Ingestion

Plants

Monitoring

Storage

Processing

Visualization

Plants

Processing

Parts

Traceability

IT Ticketing

Flat filesExternal

DatabasesReal Time

Process Data

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 10

Table of Contents

Enjeux, contexte et bénéfices

Solutions techniques Big Data

Applications IBM PMQ et Braincube

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 11

Predictive Maintenance – Principe et mise en œuvre avec PMQ

Visualisation

& UsageData AnalysisData Storage & StructurationData Collect

7.5

5 min.

DATA COLLECTION DATA STRUCTURATION MODEL & ANALYSE DEPLOY & IMPROVEOBJECTIVES & DATA

IDENTIFICATION

Define clear objectives

Identify if relevant data are

available

Prepare Change

MIPM DEPLOYMENT

Industrial IS

Machines connected

Data collection

Secure & scalable

Data structuration

Data Lake

Analytics platform

Monitoring

Modeling

Dashboarding

Deployment

Adapt, optimize

Change management

1. Récupération des données du data lake en temps réel

2. Traitement sur intervalles puis mise à disposition d’un

modèle prédictif (algorithme)

3. Le modèle établit un score d’anomalies

4. Interprétation et décision

Machine learning : Détection d’anomalies corrélée à une

base d’apprentissage et de connaissances.

Performance: Disposer de modèles pertinents avec des

données significatives , d’un contexte métier et des process.

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 12

Predictive Maintenance – Illustration avec machine de Fine blanking

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 13

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 14

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 15

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 16

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 17

Into data: a concrete example on Big data for an automotive supplier

Data Driven Production

• Manufacturing

Intelligence

What we wanted to achieve with BIG DATA

Reduce scraps

Quickly investigate a production problem

19 Equipment

on the line

A measure

every 1s60 000 sin a production

day

220 days of

production

> 20 parameters

by equipment

followed in real time

X X5 Billions data available for analyse in 1

year of production

XX =

BRAINCUBE

Solution

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 18

How we can do that: Reduce scraps on dashboard

« It is not knowing what to do, it’s

doing what you know »

Anthony Robbins

2015 06 Scrap at the FRIMO

Manufacturing intelligence is

about undestanding what

makes your production green

and repeat it

Guides &

Rules

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 19

Braincube found a way to adjust production settings that reduce

scraps

Rule – RHD2

Lookint at only 2 parameters combined

(temperature galvano &

thickness) ...

1

: Good parts went from 96 to

98,2%

4

...we were 3,8%

time with a setting

that generate few

scraps...

The analytics say that we could be up to 40% time in

this favourable situation

4

And

save

M€ ! 5

...During the past 27 days...

32

Presentation Title | Date

Copyright © 2016 Capgemini and Sogeti. All rights reserved. 20

A collaborative plateform to share the production status in real time

FROM DATA TO

FACTS BASED

ACTIONS ON THE

PRODUCTION LINE

Manufacturing

Intelligence

Site manager, COO, BU manager

• Production line manager

• Quality manager

• Methods

• Process engineering

• Operator on the shop floor

www.capgemini.com

The information contained in this presentation is proprietary.

Copyright © 2016 Capgemini and Sogeti. All rights reserved.

Rightshore® is a trademark belonging to Capgemini.

www.sogeti.com

About Capgemini and Sogeti

With more than 180,000 people in over 40 countries, Capgemini is a

global leader in consulting, technology and outsourcing services. The

Group reported 2015 global revenues of EUR 11.9 billion. Together

with its clients, Capgemini creates and delivers business, technology

and digital solutions that fit their needs, enabling them to achieve

innovation and competitiveness. A deeply multicultural organization,

Capgemini has developed its own way of working, the Collaborative

Business Experience™, and draws on Rightshore®, its worldwide

delivery model.

Sogeti is a leading provider of technology and software testing,

specializing in Application, Infrastructure and Engineering

Services. Sogeti offers cutting-edge solutions around Testing,

Business Intelligence & Analytics, Mobile, Cloud and Cyber

Security. Sogeti brings together more than 23,000 professionals in

15 countries and has a strong local presence in over 100 locations

in Europe, USA and India. Sogeti is a wholly-owned subsidiary of

Cap Gemini S.A., listed on the Paris Stock Exchange.