Preventive Maintenance

Principles
SPL 7.2
Scott Couzens, LFM 06
Scott Hiroshige, LFM 06
Erik Smith, LFM 03 Intel Corporation
Presentation for:
ESD.60 Lean/Six Sigma Systems
MIT Leaders for Manufacturing Program (LFM)
Summer 2004
These materials were developed as part of MIT's ESD.60 course on "Lean/Six Sigma Systems." In some cases,
the materials were produced by the lead instructor, Joel Cutcher-Gershenfeld, and in some cases by student teams
working with LFM alumni/ae. Where the materials were developed by student teams, additional inputs from the
faculty and from the technical instructor, Chris Musso, are reflected in some of the text or in an appendix

Special Thanks to:

Intel Employees:
Jonathan Matthews
David Latham
Eli Sorenson
Danny Miller

DNS Employees:
Roger Nuffer
David Villareal
Marcus Hunsaker

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 2

Part IV: Disconnects

Part V: Conclusion

Overview
Session Design (20-30 min.)

Learning Objectives
Familiarity with the different
types of maintenance
activities
Appreciation of the benefits of
preventive maintenance
Understanding of lean
principles for designing a
preventive maintenance
schedule
Awareness of specific
challenges to implementing
preventive maintenance

Part I: Introduction and Learning

Objectives (1-2 min.)
Part II: Key Concept or Principle
Defined and Explained (3-5 min.)
Part III: Exercise or Activity
Based on Field Data that
Illustrates the Concept or
Principle (7-10 min.)
Part IV: Common Disconnects,
Relevant Measures of Success,
and Potential Action
Assignment(s) to Apply Lessons
Learned (7-10 min.)
Part V: Evaluation and
Concluding Comments (2-3 min.)

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 3

Part IV: Disconnects

Part V: Conclusion

Types of Maintenance
Breakdown Maintenance:
Waiting until equipment fails before repairing or servicing it

Preventive Maintenance (PM):

(Time-based or run-based) Periodically inspecting, servicing,
cleaning, or replacing parts to prevent sudden failure
(Predictive) On-line monitoring of equipment in order to use
important/expensive parts to the limit of their serviceable life

Corrective or Predictive Maintenance:

Improving equipment and its components so that preventive
maintenance can be carried out reliably

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 4

Part IV: Disconnects

Part V: Conclusion

Benefits of Preventive Maintenance

the cost of breakdown maintenance is usually much
greater than preventive maintenance. 1
Preventive maintenance
Keeps equipment in good condition to prevent large problems
Extends the useful life of equipment
Finds small problems before they become big ones
Is an excellent training tool for technicians
Helps eliminate rework/scrap and reduces process variability
Keeps equipment safer
Parts stocking levels can be optimized
Greatly reduces unplanned downtime

http://www.prenhall.com/divisions/bp/app/russellcd/PROTECT/CHAPTERS/CHAP15/HEAD01.HTM

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 5

Part IV: Disconnects

Part V: Conclusion

The Manufacturing Game

Similar to the Beer Game
Simulates a typical plant with three roles:
Operations Manager
Maintenance Manager
Spare Parts Stores Manager

Each round, participants make decisions such as:

Which equipment to take down for PMs
How to allocate maintenance resources
How many spare parts to order

Revenue, cost, output, uptime, inventory are recorded

Source: Business Dynamics: Systems Thinking and Modeling for a Complex World,
Sterman, John D., 2000.
[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 6

Part IV: Disconnects

Part V: Conclusion

The Manufacturing Game Results

Teams who follow a cost-minimization strategy (reactive
maintenance policies) are able to keep costs low for a
while. However, as defects build up they find their
uptime falling and costs rising.
Teams who follow a preventive maintenance strategy
initially find higher costs and reduced uptime as
equipment is taken offline for planned maintenance.
Soon, however, these teams begin to greatly outperform
teams following a cost-minimization strategy.

Source: Business Dynamics: Systems Thinking and Modeling for a Complex World,
Sterman, John D., 2000.
[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 7

Part IV: Disconnects

Part V: Conclusion

When Does PM Make Sense?

PM makes sense when the cost of doing PM is less than the
cost of NOT doing PM.
PM makes sense if CDoingPM < CNotDoingPM
CDoingPM =

f(hours of not running equipment, loss in employee

morale from doing PM instead of real work, materials
and man-hours consumed in PM, potential for making
things worse, etc.)
CNotDoingPM = f(cost of losing/reworking a failed batch (unless PM
makes no difference in preventing the failure), materials
and man-hours spent repairing equipment, loss of
equipment lifetime, loss in employee morale from NOT
doing PM, reduced employee familiarity with equipment,
etc.)

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 8

Part IV: Disconnects

Part V: Conclusion

Optimizing a PM Schedule
Question:
If a certain piece of production equipment requires ~10 hours of
preventive maintenance per week, how should those 10 hours
be scheduled?

Answer:
In a 24x7 manufacturing operation, it is typically better to
perform the ~10 hours of activities in several smaller periods of
time, for instance 5 PM activities that take ~2 hours each
Duration and variability in preventive maintenance are key
factors in whether equipment will be able to maintain a steady
flow of output

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 9

Part IV: Disconnects

Part V: Conclusion

PM Durations: Simulation 1
Simulation of equipment with a 10 hour average PM duration, std dev 20
hours (85% availability)
Drop Page Fields Here
Trend of Queues for the Toolset

Avg Queuelength
450

Average QueueLength (# of Lots)

400
350
300
250

Scenario
1 - 10h mean, 2 CV

200
150
100
50
0
83 88 99 109114119125133139147152157162167172177182187192198203211216221226231236241246267
Days into Sim ulation

Day

Data from a simulation run at Intel Fab 11X.

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 10

Part IV: Disconnects

Part V: Conclusion

PM Durations: Simulation 2
Simulation of equipment with a 5 hour average PM duration, std dev 10
hours (85% availability)
Drop Page Fields Here
Trend of Queues for a Toolset

Avg Queuelength
450

Average QueueLength (# of Lots)

400
350
300
250

Scenario
2 - 5h mean, 2 CV

200
150
100
50
0
83

88

93

98 103 108 113 118 125 137 155 164 194 201 213 228 233 244 259 264 277 285
Days into Sim ulation
Day

Data from a simulation run at Intel Fab 11X.

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 11

Part IV: Disconnects

Part V: Conclusion

PM Durations: Simulation 3
Simulation of equipment with a 2 hour average PM duration, std dev 4
hours (85% availability)
Drop Page Fields Here
Trend of Queues for a Toolset

Avg Queuelength
450

Average QueueLength (# of Lots)

400
350
300
250

Scenario
3 - 2h mean, 2 CV

200
150
100
50
0
84

91

102 116

125

138

148 160

173

196

209 225

230

238

261

273 279

287

Days into Sim ulation

Day

Data from a simulation run at Intel Fab 11X.

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 12

Part IV: Disconnects

Part V: Conclusion

PM Durations: Simulation 4
Simulation of equipment with a 2 hour average PM duration, std dev 6
hours (85% availability)
Drop Page Fields Here
Trend of Queues for a Toolset

450

Avg Queuelength

Average QueueLength (# of Lots)

400
350
300
250

Scenario
4 - 2h mean, 3 CV

200
150
100
50
0
84 98 112 117 122 127 133 138 143 148 153 158 163 168 173 178 183 188 193 204 222 242 258 264 275 286 291
Days into Simulation
Day

Data from a simulation run at Intel Fab 11X.

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 13

Part IV: Disconnects

Part V: Conclusion

PM Durations: Why they matter

For the exact same availability on a piece of equipment:
Simulation

Avg QueueLength

StdDev QueueLength

1) =10h, =20h

56

63

2) =5h, =10h

38

52

3) =2h, =4h

4) =2h, =6h

56

57

Shorter PM durations mean a difference in days of lot

cycle time!
Each day of factory cycle time = millions of $$

Regardless of the mean, performing PMs inconsistently

is functionally equivalent to consistently having much
longer downtime durations

Data from simulations run at Intel Fab 11X.

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 14

Part IV: Disconnects

Part V: Conclusion

The Waddington Effect

First observed by C.H. Waddington during WWII for
British aircraft maintenance
Background theory: unscheduled downtime should
be a random phenomenon
If all unscheduled downtime events are plotted with
respect to the last PM, there should not be any
pattern evident
A pattern of increased unscheduled downtime
immediately following PMs is a Waddington
Effect

Information from a presentation given at Intel Fab 11X.

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 15

Part IV: Disconnects

Part V: Conclusion

The Waddington Effect

"Waddington Effect" Report - Trend of Unscheduled Downtime
Frequency Following PMs for the last 12 work weeks
70

# of Unsched Downs

# of Unscheduled Down Events

60

50

Event_Type

40

REPAIR
QUAL
30

OUT OF CNTRL

20

10

0
0.5

1.5

2.5

3.5

4.5

Days since End of Last PM

Days Since PM

Impact of Waddington Effect = 1.4% Availability

Data from a set of production tools at Intel Fab 11X.
[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 16

Part IV: Disconnects

Part V: Conclusion

Challenges to Implementing
Preventive Maintenance
Social Factors

Technical Factors

Organizations are
frequently structured in
ways that promote local
optimums (cost, shiftly
output goals, etc.)

Breakdown maintenance
is typically cheaper than
preventive maintenance
in the short-term

The benefits of preventive

maintenance are not
always well understood

Under-trained technicians
can cause more damage
than they prevent

The focus on minimizing

maintenance costs has to
shift to maximizing overall
organizational
performance

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 17

Part IV: Disconnects

Part V: Conclusion

Concluding Comments
Performing preventive maintenance is almost always the best
long-term strategy to maintain equipment
PM scheduling and strategy are keys to maximizing output
while reducing work-in-process inventory
Due to short-term cost increases and local optimums, there are
barriers to implementing a preventive maintenance strategy at
some plants

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 18

Part IV: Disconnects

Part V: Conclusion

Appendix: Instructors Comments and Class

Discussion for 7.2

PM is an important tool for establishing stability

necessary for other lean elements:
Andon, 5s, etc.

How do you escape the crisis management

whirlpool?
Social disconnects:
Change the mindset of management, maintenance
group from reaction to prevention
May need to be done in steps
[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

6/9/04 -- 19

Appendix: Instructors Guide

Slide

Time

Topic

Additional Talking Points

1-3

2-3 min

Introduction, overview and

learning objectives

Acknowledge contributors to the presentation

Provide an overview of the learning objectives

4-5

3-5 min

Key Concepts

Describe the 3 types of maintenance

Explain the benefits of preventive maintenance

6-15

7-15 min

Exercises/Activities

Go through the key learnings of the Manufacturing

Game
Go through a PM optimization example
Discuss Waddington Effect

16

2-3 min

Disconnects

Discuss the difficulties of implementing preventive

maintenance in a reactive maintenance culture

17

1-2 min

Concluding comments

Summarize key points of preventive maintenance

[LFM Students] ESD.60 Lean/Six Sigma Systems, LFM, MIT

Part I: Introduction

Part II: Concepts

Part III: Application

6/9/04 -- 20

Part IV: Disconnects

Part V: Conclusion