understanding the magic of lean product development · pdf fileunderstanding the magic of lean...

Understanding the Magic of Lean Product Development

JAOO 2010Århus, DanmarkOctober 4, 2010

Donald G. ReinertsenReinertsen & Associates

600 Via Monte D’OroRedondo Beach, CA 90277 U.S.A.

(310)-373-5332Internet: Don@ReinertsenAssociates.com

www.ReinertsenAssociates.com

No part of this presentation may be reproduced

without the written permission of the author.

2

Any sufficiently advanced technologyis indistinguishable from magic.

– Arthur C. Clarke

3

Lean Manufacturing

• Lean Manufacturing is a best practice.

• Best practices lead to superiorperformance.

• Why not adopt these best practices inproduct development?

4

The TPS Emergency Room

• We desire to rigorouslyimitate the practices ofToyota.

• All arriving patientswill be processed on aFIFO basis.

• We will stop admittingwork when we reachour preset WIP limit.

5

Thus, since the Toyota ProductionSystem has been created from actualpractices in the factories of Toyota, ithas a strong feature of emphasizingpractical effects, and actual practice andimplementation over theoreticalanalysis.

– Taiichi Ohno

From Foreword to 1983 FirstEdition of Toyota ProductionSystem by Yasuhiro Monden,

6

Turning Magic into Technology

LPD1 LPD2

• Repetitive Tasks• Low Variability• Homogenous Flows

Queueing Theory

Traffic Flow Theory

Computer OS Design

Maneuver Warfare

The Internet

Lean Manufacturing

Toyota

• Non-Repetitive Tasks• High Variability• Non-Homogenous Flows

DOMAIN

Use Some Ideas ofLean Manufacturing

Add Concepts and Sciencefrom other Domains+

7

Queueing Theory

8

Hvem er jeg?

9

Alle har mødt Erlang meningen ved rigtig hvem har er!Det skyldes, at de berømtefircifrede logaritmetabeller,som fleste har brugt iskolerne, baerer Erlangsnavn.

Det, han virkelig blevverdensberømt for, varteletrafik-teorier.

– Bjarne Kousholt

10

Traffic at rush hourillustrates the classiccharacteristics of aqueueing system.

Ph

oto

Co

pyri

gh

t 2

00

0 C

om

sto

ck, In

c.

11

The Effect of Capacity Utilization

Note: Assumes M/M/1/Infinite Queue

Queue Size vs. Capacity Utilization

0

5

10

15

20

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Capacity Utilization

Qu

eue

Siz

e

12

Total Cost

Cost of Delay

Cost of E

xcess Capacity

Economics of Queues

Excess Product Development Resource

Dollars

13

Batch Size

14

Setting Batch SizeEconomic Batch Size

0

5

10

15

20

25

1 2 3 4 5 6 7 8 9 10

Items per Batch

Cost

Transaction Cost Holding Cost Total Cost

15

CheaperCorrection

CheaperDebug

Lower CostChanges

BetterEconomics

Fewer StatusReports

Faster CycleTime

EarlyFeedback

Faster Learning

Less RequirementsChange

Less DebugComplexity

More EfficientDebug

CheaperTesting

Less Non-Value-Added

BetterCode

FewerOpenBugs

More Uptime

Higher Validity

SmallerChanges

Benefits of Small Batch Testing

16

WIP Constraints

17

Little’s Famous Formula

λ

λ

=

=

=

=

Rate Departure Average

Queue in Customers of Number Average

Queue in TimeWait Average

q

q

q

q

L

W

LW

18

Visual WIP Boards

ReadyQueue Coding

Readyto Test Testing

TestComplete

A D

E

C

B

WIP Constraint = 10 units

19

Arrivals

Departures

Time in Queue

Quantityin Queue

QueueTime

CumulativeQuantity

Cumulative Flow Diagram

20

1 6 11 16 21 26 31 36 41

Arrivals Departures

Time 21: 400Passengers Arrive,

Queue up 5xby Time 22 Time 41:

Cycle Timeup 2x

CumulativeQuantity

• Queues give instant indication of a problem.• This is very important when problems age poorly and when fastresponse times matter.

Control Queues Not Cycle Time

21

SynchronizedCadence

22

Fixed

Variable

Interval

Train Length

Variable

Fixed

23

Cadenced Purchasing Availability

BEFORE

• One buyer willsupport you with 10percent of his time.

AFTER

• Buyer will be at deskin team area from 8:00AM to 9:00 AM daily.

• During this period hishighest priority issupporting yourproject.

24

Asynchronous Processing

1 2 3 4 5 6 7

THE THEORY

• With the new IT system we can tell exactly who has eachECR at any point in time.

• Work could be done instantly instead of waiting for ameeting.

• 40 day Processing Time• Low First Pass Yield• High Processing Cost

THE REALITY

25

Variability

26

Taking Rational Risks

Stakes Payoff Probability EMV Bet?

$15,000 $100,000 50% $35,000 ?

$15,000 $20,000 90% $3,000 ?

$15,000 $16,000 100% $1,000 ?

We cannot maximize economic value byeliminating all bets with uncertain outcomes.

Choice

A

B

C

EMV=Expected Monetary Value

27

Payoff vs. Price

Price

Pay

off

Expected Price

Price

Pro

bab

ilty

Asymmetric Payoffs and Option Pricing

Strike Price

Expected Payoff

Price

Exp

ecte

d

Pay

off Strike

Price

x

=

28

Sequencing

29

Queueing Disciplines

• FIFO

• Highest Profit (or ROI/IRR/EVA) First(HPF)

• SJF (FCFS)

• High Cost of Delay First (HDCF)

• Minimum Slack Time First (MSTF)

• Weighted Shortest Job First (WSJF)

30

Fast Feedback

31

Front-Loaded Two Digits at Same Time

Buy Second Digit After Receiving Feedback

Pay $2

Make Nothing

Make $200

99%

1%

ExpectedPayoff: Zero

Pay $1Make Nothing

Pay $1

90%

10%

ExpectedPayoff: $0.90

Make Nothing

Make $200

9%

1%

The Value of Feedback

32

The Importance of Math

• There are underlying mechanisms ofaction behind lean methods.

• These mechanisms can be used in LPD.

• These methods affect more than onemeasure of performance, so tradeoffs arenecessary.

• This requires that you use a common unitof measure for your decisions.

33

Economics

34

Making Economic Decisions

Waste

Cycle Time

Variability

Efficiency

Unit Cost

Value-Added

Revenue

Life CycleProfits

Economic SpaceProxy Variable Space

Transformations

35

The Modeling Process

Create Baseline Model

Determine Total Profit Impact of Missing a MOP

Calculate Sensitivity Factors

ModelExpenseOverrun

ModelSchedule

Delay

ModelValue

Shortfall

ModelCost

Overrun

ModelRisk

Change

36

The Model Output

Life-Cycle Profit Impact

-$80,000

-$500,000

-$100,000-$150,000

-$40,000

1 PercentExpenseOverrun

1 PercentProduct Cost

Overrun1 Percent Value

Shortfall 1 Month Delay1 Percent

Increase in Risk

37

Range of Cost of Delay Estimates

Poor Intuition

Average Intuition

Best Case Intuition

Average Analysis

Quality Analysis

Any Analysis Beats Intuition

200:1

50:1

10:1

2:1

1.2:1

Source: Reinertsen & Associates Clients

38

Example: The Goal of Conformance

Performance

Target

Actual 101%

90%

FeatureA

FeatureB

Correct thePerformance Gap

Actual

39

Marginal Economics

Value Value

CostCost

PerformancePerformance

$ $

Feature A Feature B

Target Actual Target Actual

?

Should our goal be to optimize conformance,or to make good economic choices?

40

Decentralizing Control

Engineer Supervisor ProgramManager

Boeing 777 Weight Reduction Decision Authority

$300

$2,500

$600

Dollars per Pound

1991 / 1997 1997 2009

Going Further

42

Seminars in Scandinavia

Copenhagen: November 29-30, 2010

Stockholm:December 2-3, 2010