Sterling MortensenPresident, Applying Lean Inc.sterling@applyinglean.com

1Copyright Applying Lean Inc.

Number of products growing quickly

New more complex products

Number of features growing exponentially

Moving to common software platform

Demand far exceeded ability to supply

2Copyright Applying Lean Inc.

Speed (Fast Processes) Less Work

Because (Less of People’s Time is being Wasted on Non-Value Added Work)

3Copyright Applying Lean Inc.

4Copyright Applying Lean Inc.

Corporate Value $70 million program

cost avoidance in 20 months on Edgeline

$27 million in annual savings (FW lab budget reduction) going forward

Program TTM reduced by 1 year

7.4x productivity improvements sustained during staff & budget reductions

5Copyright Applying Lean Inc.

Theory of Constraints - Eliyahu Goldratt

Global Optimization

Queuing Theory – Don Reinertsen

Continuous Improvement

Value Stream Maps + Information Flow

Discovered the Laws that make Lean work

6Copyright Applying Lean Inc.

Natural Law’s&

Human Psychology

7Copyright Applying Lean Inc.

21

3

5

4

8Copyright Applying Lean Inc.

Productivity is inversely related to complexity

• Productivity is Restored by breaking large complex projects into smaller –Iterative Development projects

Productivity = aComplexity

1

9Copyright Applying Lean Inc.

Technology delivery

Technology delivery

Note the consistent “heartbeat”

Small Batch + Fixed Cadence

Technology delivery

Technology delivery

Technology delivery

Technology delivery

Technology delivery

Technology delivery

Product A delivery Product B delivery

10Copyright Applying Lean Inc.

Going to small batch iterative development reduced defects, every time the process went faster, the quality got even better!

Quality Goes Up as Time to Feedback Goes Down!

People are self motivated to improve quality if they can see quickly that what they are doing has problems. But if there is a long delay their

motivation drops significantly.

11Copyright Applying Lean Inc.

53% Less

80% Less 87%

Less

12Copyright Applying Lean Inc.

Too much WIP (Work In Process) is the greatest source of waste

•Use a Kanban system to keep the WIP constant•A constant WIP delivers a consistent r te of delivery, which makes planning and status more easy and accurate

13Copyright Applying Lean Inc.

Cumulative Flow of Features

0

200

400

600

800

1000

1200

1400

3/2

2/2

004

4/5

/2004

4/1

9/2

004

5/3

/2004

5/1

7/2

004

5/3

1/2

004

6/1

4/2

004

6/2

8/2

004

7/1

2/2

004

7/2

6/2

004

8/9

/2004

8/2

3/2

004

9/6

/2004

9/2

0/2

004

10/4

/2004

10/1

8/2

004

11/1

/2004

11/1

5/2

004

11/2

9/2

004

12/1

3/2

004

12/2

7/2

004

1/1

0/2

005

1/2

4/2

005

2/7

/2005

2/2

1/2

005

3/7

/2005

3/2

1/2

005

Time

Cu

mu

lati

ve F

eatu

res

Cum Proposed

Cum Accepted

Cum Evaluated

Cum Designed

Cum Committed

Cum Completed

Cum Verified

WIP

WIP >

500

WIP >

900

Work In Process Example

14Copyright Applying Lean Inc.

800

85 99 100 100

950

92

0

100

200

300

400

500

600

700

800

900

1000

Mar-05 Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05 Dec-05 Jan-06 Feb-06

Time

Fe

atu

res

in

WIP

48 months of inventory

20 features per month

2 1/2 months of inventory

39 features per month

15Copyright Applying Lean Inc.

Thro

ughput

Load

Optimal

Throughput

Kanban Control

16Copyright Applying Lean Inc.

Work on small chunks of work

Take the work to completion

Use Kanban system to control total WIP

17Copyright Applying Lean Inc.

Small batch = low WIP

high productivity

Predictability◦ Linear vs. non-

linear closure

Manager satisfaction

Employee satisfaction

Get done sooner◦ Continuous quality

reduces rework

◦ Better economics

Req’d scope

18Copyright Applying Lean Inc.

1. Product Development is Learning

2. Product Development is high variability

3. Learning doesn’t happen in nice, phased steps

4. Risk reduction requires accelerating learning in high risk areas

5. Rate of learning = f(1/time to feedback)

19Copyright Applying Lean Inc.

Done

WIP

Backlog

Average Closure Rate

Avg WIP Arrival Rate

Avg Backlog arrival rate

Eventual convergence indicates that backlog can be worked off.

Amount of WIP = queue size. Queue size has a direct correlation to quality and cycle time

Avg rates can be computed using run charts and, later, SPC charts. Variance can be accounted for to yield predictability

20Copyright Applying Lean Inc.

Jedi SysRDAll Analysis

0

100

200

300

400

500

600

700

3.1 3.2 3.3 3.4 3.5 3.6 4.1 4.2 4.3 4.4 4.5 4.6 5.1 5.2 5.3 5.4 5.5 5.6 6.1 6.2 6.3 6.4 6.5 6.6 7.1 7.2

Core Release Cycle

Co

nd

or

Req

uir

emen

ts

0

200

400

600

800

1,000

1,200

1,400

Tim

e to

Mai

n (d

ays)

Total Submitted Total Started Total Committed

Total FW Implemented Total FW Verified Total Sol'n Verified

Time to Main [avg] Avg Time To Main Pre/Post Adoption of Lean (4/25/2005) UCL

21Copyright Applying Lean Inc.

• Uncertainty = Risk = Areas the learning is needed• Focusing early iterations on high risk areas enables learning, reduces uncertainty

& risk22Copyright Applying Lean Inc.

A requirement or feature chunk is not doneuntil it has passed ALL tests◦ This means that tests have been run, defects fixed

◦ Demos if used are done and changes made based on feedback

◦ Then the tests are rerun

◦ Learning is complete once ALL tests pass and demo approved!

◦ No Credit is given until it is completely done!

23Copyright Applying Lean Inc.

“Continuous Flow is the Ultimate Objective of lean production, and creating continuous flow has been the goal of countless kaizen projects. We work at a wide range of manufacturing facilities and see many commendable efforts to create flow. Unfortunately, we see very little of it actually achieved.” Mike Rother & Rick Harris from the book “Creating Continuous flow”

24Copyright Applying Lean Inc.

0

20

40

60

80

100

120

140

160

180

Cumulative Demand

Total Tasks Completed

Cumulative Actual

Potential savings of about 15%

in time with reduced headcount

Projected timeline with less resource DO

Added features

Doing 15% more in the same timeline

+20

VR-Sep09

Tasks/F

eatu

res

Highlights:

-Bocce (1) got defined in Nov07 with aggressive schedule, and added more features as OEM integrates

-Delivering more with reduced headcount, attrition management

-Improved efficiency due to

- Closer/direct interaction with OEM/AE for issue resolutions and responsiveness

- More knowledge gained as the team makes the progress along PLC

- Increased rigor in the overall review processes

25Copyright Applying Lean Inc.

LonePine Program EE deliverables for Lone Pine - End to end prediction with

accomplishment as on 7/25/2008

0

50

100

150

200

250

Sep-

07

Oct-

07

Nov-

07

Dec-

07

Jan-

08

Feb-

08

Mar-

08

Apr-

08

May-

08

Jun-

08

Jul-08 Aug-

08

Sep-

08

Oct-

08

Nov-

08

Dec-

08

Jan-

09

Feb-

09

Mar-

09

Time (months)

De

livera

ble

s

Cumulative Completed Completed Each Month Cumulative Started Cumulative Demand

Toda

y

If something is in WIP, it is

probably safe to assume there

has been progress. This graph

only shows completed

deliverables though.

The blue line reaching the red line

indicates that everything due for this

milestone should be started.

The gap between cumulative

completed and started indicates

WIP

As of on 7/25, EE HW is

pending with one deliverable

(ESD issue)

26Copyright Applying Lean Inc.

Each improvement opens the way to find another improvement,

leading to continuous improvement

27Copyright Applying Lean Inc.

Measure only Items Done

Requirement chunks that are in process at the end of an iteration continue on into the next iteration, until they are done

Variability is natural.

28Copyright Applying Lean Inc.

29Copyright Applying Lean Inc.