“predictive maintenance in a white - boussias conferences · 17/10/2019 · predictive...
TRANSCRIPT
Whirlpool Corporation - Confidential
“Predictive Maintenance in a White Good industry: current status,
myths and facts”Pierluigi Petrali – Manufacturing R&D Manager
Whirlpool Corporation - Confidential
A LEADERSHIP POSITION
WHIRLPOOL CORPORATION
72 million products sold in more than 170 countries in the world
World’s leadingmajor home appliance company (NYSE:WHR)
Approximately
$21 billion in sales in 2018
92 000 employees
70 manufacturing and R&D centers
$1 billion investment in capital and R&D centers annually
Whirlpool Corporation - Confidential
OUR GLOBAL PORTFOLIO
*Whirlpool ownership of the Hotpoint brand in EMEA and Asia Pacific regions in not affiliated with the Hotpoint brand in the Americas
*
Whirlpool Corporation - Confidential
EMEAMelanoItaly
Comunanza Italy
CassinettaItaly
SienaItaly
NaplesItaly
IsithebeSouth Africa
YateUK
WroclawPoland
ŁódźPoland
RadomskoPoland
Lipetsk Russia
Poprad Slovakia
ManisaTurkey
Carinaro Italy
INDUSTRIAL FOOTPRINT
Whirlpool Corporation - Confidential
EMEA
BRANDS AND PRODUCT LEADERSHIP
Indesit Built-in Suite
Indesit Innex
Whirlpool SmartCook
Whirlpool SupremeCare
KitchenAid Black Steel Collection
KitchenAid Culinary Center
Whirlpool Corporation - ConfidentialWhirlpool Corporation - ConfidentialWhirlpool Corporation - Confidential
Maintenance in the perspective of World Class Manufacturing
Whirlpool Corporation - Confidential
Whirlpool Production System TECHNICAL PILLARS
7
Whirlpool Corporation - Confidential
Vision
➢ Obtain and maintain high machine efficiency by▪ Eliminating breakdowns▪ Reducing component replacement frequency / costs
Needs
➢ To maintain efficient, breakdown free machines➢ To continuously improve machine functionality
Objectives
➢ Zero Breakdowns➢ Increase the technical availability➢ Reduction of maintenance costs
INTRODUCTION TO PROFESSIONAL MAINTENANCE
8
Whirlpool Corporation - Confidential
7 STEPS OF PM PILLAR
STEP 2
STEP 3
STEP 4
STEP 5
STEP 6
STEP 7
STEP 1
STEP 0
Preliminaryactivities
REACTIVE PREVENTIVE PROACTIVE
Elimination and preventionof accelerated deterioration
Reverse deterioration (Breakdown Analysis)
Establishment of maintenance
standards
Countermeasures against weak points of
the machine and lengthened equipment
life
Build a predictive maintenance system (trend management)
(CBM)
Maintenance Cost Management
Establishment of a planned maintenance
system
Build a periodic maintenance system
(TBM)
9
Whirlpool Corporation - Confidential
Ma
inte
nan
ce M
ix
Key Machines (AA & A Class)Other Machines (B & C Class)
Autonomous Maintenance (Cleaning, Inspection, Lubrication, Refastening Activities)
Breakdown Maintenance
Time Based Maintenance
Corrective Maintenance
Condition Based Maintenance
TYPE OF MAINTENANCE AND TYPE OF MACHINES
10
Whirlpool Corporation - ConfidentialWhirlpool Corporation - ConfidentialWhirlpool Corporation - Confidential
Whirlpool approach to Industry 4.0
Whirlpool Corporation - Confidential
I4.0 TECHNOLOGIES MAJOR APPLICATIONS
CoreTechnological
Areas
Use Cases● Scouting: 81● Trial: 8● Pilot: 29● Deployment: 2
Augmented Reality
Human Centered
Automation
Data Visibility
Advanced Testing
Traceability
Training on critical workstations
Augmented Reality
Training on critical workstations
Virtual Reality
3D digitization for full line and equipment
Collaborative Automation
Cobots for panel screwing
Collaborative Automation
Mobile robots for material delivery
Visibility and Analytics
Cloud based KPI’s
Visibility and Analytics
Real Time Energy Management
Vision Systems
Smart Quality Control
Industrial Inspection
Drones for visual and thermal inspection
Real Time Location Systems
Forklifts and tuggers tracking
Tools at Suppliers Location
E Tooling. Remote performance monitoring
1
Whirlpool Corporation - Confidential
EU cooperation as Innovation Engine for I4.0
EU Research Projects
EIT Manufacturing
I4.0 CORE AREAS IMPLEMENTATION
Supplier
Scouting
● Contact/
evaluate/ engage
suppliers
● Identify/ develop
solutions
Trials
● Technologies try
out in the factory
○ Assess
technology
solution
applicability
○ Address
process/
organization
impact
Pilot
● First complete
implementation in
a factory narrow
context
○ costs/ benefits
identification
○ process/
organization
harmonization
Standardization
● After one/ two
completed pilot
extensions
● Use case ready
to be replicated
(TO BE):
○ Standardized
capabilities
○ Technical/
systems/
process/
organization
requirements
Deployment
● Gap analysis for
implementation
identified in each
factory:
○ Technical
○ Organizational
○ Opertional
● Handover
completed and
Local/ central
competence
available
● STD refinement
Rollout
● STD Solution
ready for roll
out
● Full roll out
costs known
● Prioritization
and
implementatio
n driven by
factories
Technology
Scouting
● Identify solutions
according to
knowledge intake
process
○ R&D
○ Competence
centers
○ Benchmarking
○ Trade shows
FactoryOE - Tech
Technical Readiness
Key Deliverables
First Pilot complete
Communication Events
OE - I4.0Main Accountabilities
Functional Readiness Cost/ Organization Readiness
1
21
Concepts
Core Areas
Technologies
Whirlpool Corporation - ConfidentialWhirlpool Corporation - ConfidentialWhirlpool Corporation - Confidential
Research projects on Predictive Maintenance:
General Information
06.11.18 16
DE
FR
I
E
I
TG
R
C
Y
▪ Topic FOF-09-2017
Novel design and predictive maintenance technologies for increased operating life of product system
▪ Duration 36 Months
▪ Start 01.09.2017
▪ End 31.08.2020
▪ Total costs € 6,248,367.50
▪ Max grant € 4,847,836.25
Whirlpool Plant under experimentation
• Clothes dryer drum production line
• Installed in Q1 2018 in Lodz factory in Poland
17.10.2019 17
Feeding Line StraightenerHydraulic Punching
Automatic Buffer
Seaming Flanging Spokes
AssemblyDrum Pre-Assembly
Flange Seaming
Dimensional Control
Unloading
UPTIME Deployment
17.10.2019 18
TO-BE Business Process
17.10.2019 19
Da
tafl
ow
Predictive Maintenance Order
Generation
Order Validation from
Maintenance
Supervisor
FMECA & Physical Measures To Be Monitored
20
• One objective of the FMECA is to address the monitoring of physical measures by means of sensors that could contribute to failure modes/causes prognosis and failure rate estimation
• Some parameters are already monitored by the PLC of the Substations – i.e. Part Gauge
• New physical measures have been identified and new sensors has been installed in March 2019 – i.e Equipment Gauge
Part Gauge Data preliminary analysis
17.10.2019 21Pierluigi PETRALI
• Anomaly detection
Tim
e-s
eri
es
Pre
dic
tio
n
WP5 Lessons Learnt (so far!)
17.10.2019 22
Need for high quality dataset
High quantity…
crucial
Collaboration
between Process
Experts and Data
Experts
2019-10-17 23
List of participants
Participant No * Participant organisation name Country
1 (Coordinator) COMAU S.p.A. IT
2 Finn-Power Oyj FI
3 VDL Weweler BV NL
4 WHIRLPOOL EMEA SpA IT
5 Kone Industrial Ltd FI
6 Engineering Ingegneria Informatica S.p.A. IT
7 OCULAVIS GmbH DE
8 SynArea Consultants S.r.l. IT
9 DELL EMC IE
10 Laboratory for Manufacturing Systems & Automation GR
11Fraunhofer Gesellschaft zur Förderung der angewandten
ForschungDE
12 VTT Technical Research Centre of Finland Ltd FI
13 TRIMEK S.A. ES
14 Politecnico Di Torino IT
▪ 14 partners, 7 EU member states, across Europe
WP5
2019-10-17 25
Overall Concept
Smart data collection
module
Analog Ethernet
USB
Data 4.0 Box
5G
Edge-computing
Condition monitoring
Data
management
AI-
machine
learning
Physics-
based
Data-driven
AR based tools for remote
assistance
Maintenance
personnel support
WP2
Remote control of
factoryData communications
(e.g. OPC-UA)
HW & SW integration
Communication
interfaces (e.g. 5G
interface)
Data analytics in cloud computing
(near real time application)
Plug-and-play cloud-based platform
AI CbM, Data analytics
Hybrid modelling
Planning of maintenance
WP3 WP4
Industrial needs, requirements, pilot cases definitionWP1
Demonstration activities
White goods industry
Metrology industry
Elevators production
Steel parts production
WP6
OCULAVIS
Finn-Power
VTTIPT
ENG
KONE
Robotics industry
2019-10-17 26
Demonstrator Description
Pressure sensors
Temperature sensors
Flowmeter
Pressure sensors
Temperature sensors
Flowmeter
Factory SQL Database
2019-10-17 27
WHR Testbed
Factory SQL Database
Pressure sensors
Temperature sensors
SCADA
SERENA Gateway
Linux Docker
- Breakdown- Repair Time
Future dataMechanical Sensors
SAP PM data
REST
Flowmeter
Alarms
2019-10-17 28
▪ Dataset available since April 2018
• Components flow, temperature, pressure
measured at Mixing Head
• Machine alarms
• More than 65000 cycles available
• SAP/PM activity
▪ Preliminary Analytics on data
▪ Testbed implementation
Work Progress
2019-10-17 29
▪ Data Analytics outcome:
• Analytics Methodology prepared with POLITO
• Mixing Head diagnosis PRETTY DIFFICULT with only pressure, temperatures and mass flow data
• Current data could more likely provide diagnosis on other components (Pumps, Valves, etc.)
▪ Change management
• Roles and skills gap has been identified
▪ OT: process engineers need for a more data driven knowledge of the physics behind
▪ IT: Data Architects need for a basic knowledge about Deep Learning Analytics
▪ Data Scientist role is missing.
Highlight of Significant Results
SERENA Intermediate Review Meeting │Turin │ Italy
Predictive Maintenance learning and challenges
● Identify the right data to be used using FMECA or other analytical tool● Act in advance: machine learning can require many months (years?) of data● Define a standard, semantic data model● Have an IT architecture to ease up data flow from sensor to storage to
applications● Provide mechanism to correlate different kind of data● Find a way in which process expert and data scientist can effectively
exchange information and knowledge● Prediction must gain expert trust● Enable the organization and business process to interact and work with
PdM
Whirlpool Corporation - Confidential
@WhirlpoolCorp #WhatMattersLinkedInFacebookWhirlpoolCorp.com
31
JOIN THE CONVERSATION