lean process development: key questions to connect with ... · • process re-selection: in-line...

10/18/2016

1

© Copyright 2016, Lean Enterprise Institute, Inc., Cambridge, MA, lean.org. Lean Enterprise Institute, the

leaper image, and stick figure image are registered trademarks of Lean Enterprise Institute, Inc. All rights

reserved.

Lean Process Development:

Key Questions to Connect with

New Products

By Matt Zayko leanpd.org

[email protected]

© Copyright 2016, Lean Enterprise Institute, Inc.

What is the purpose of

Development in your organization?

“The purpose of Development is to create useful

knowledge and profitable operational value streams.”

--the late Allen Ward

Lean Product & Process Development Book,

LEI 2007

10/18/2016

2


What is the Operational Value Stream?

Raw Supplier Value Stream

Subassembly Supplier VS

Customers Final Assembly Value Stream

Fabrication Supplier VS

FIFO FIFO

Forming

1 Area

Feeder Line

2 lines PCT= 8

Fin ASM

3 Mixed PCT=12

FIFO

12 sec 12 sec 12 sec 12 sec 12 sec 12 sec

Product Flow

The Operational

Value Stream

is…..

….a series of

connected

processes


Development Casts a Large Shadow

5 5 2 0

7 0

%

Influence

on

Total Cost

100

90

80

70

60

50

40

30

20

10

0

5% 5% 20%

70%

10/18/2016

3


A Tale of Two Processes at Device Tech

“A bad process will defeat a good person

every time.”

--W. Edwards Deming


Device Tech’s Traditional Development Approach

Operations Process Selection Product Design

10/18/2016

4


The High Voltage Switch (HVS) at Device Tech

0.60

0.70

0.80

0.90

1.00

1.10

1.20

Jan-March April-June July-Sept Oct-DecWe

igh

ted

Vo

lum

e

Timeframe

Typical Seasonality for Switch Product

Normalized


Are you doing Kaizen…or Touzen?

• Operational Excellence in place at the Device Tech plant-

level after five years of focused effort, but…

• …every new product that launched in that same

timeframe required significant improvements to

incorporate the learning and operational standards

• Much of the improvement or kaizen activity the prior five

years had been valuable and necessary, but

• …a good portion of it was actually engineering re-work

or touzen due to poorly selected process flows

10/18/2016

5


Difference of Opinions…6 Months Prior to Launch

Metric Product Team’s Proposal

Operational Team’s Counterproposal

Cost per unit $2.26

Labor Efficiency 51%

Capital Spend $3.6 Million



Cost per unit $2.26 $1.88

Labor Efficiency 51% 86%

Capital Spend $3.6 Million $2.5 Million


Process Development Thinking at Device Tech:

6 Key Questions

• What is the current work for process development

with your new products?

• Where/What are the main failure modes?

• Where/What are the risks and opportunities?

• What is the ideal work for process development with

your new products?

• What is the improved future state of work for new

process development?

• What is the action plan to test out your future state

hypothesis?

10/18/2016

6


Phases of Development P

rocess D

evelo

pm

ent

Levels

Q1: What is the Current Work for Process Development?


Q2: What/Where are the Failure Modes?

• Too Much, Too Early (TMTE)

• Critical decisions are made too far in advance

• Knowledge incomplete

• Prematurely locks-in a process design

• Penny-Wise, Pound-Foolish (PWPF)

• Short-sighted decisions are made for local

optimization throughout the design cycle

• Results in overall performance reduction

• Too Little, Too Late (TLTL)

• Insufficient upfront design activity before launch

• Creates a flurry of rework after launch

10/18/2016

7


Q3: What/Where are the Risks & Opportunities?


Q4: What is the Ideal Process for Creating New Processes?

Context

• Scope

• Value Stream

Design

• Capacity Plan

Outputs •Short Lead Time

•Built-in Quality

•Flexible & Low

Cost Capacity &

Resources

•Simplicity &

Manageability

System Architect &

Expert Creation Team

4-2-1

Development Cycle Idea Delivery

Learning Executing

10/18/2016

8


Q5: What is the Future State for Doing Process

Development Work?

Phases of Development

Pro

cess D

evelo

pm

ent

Levels


Device Tech’s Lean Process Creation Team Experiment

• Pick System Architect(s): Technical Design Leader,

Operational Design Leader

• Select the Expert Creation Team Member Areas:

Operations, Production Control, Manufacturing

Engineering, Product Design, Quality, Equipment

Design, Purchasing, Test Engineering

• Understand the Timing: Six months until launch for the

HVS Process

10/18/2016

9


Lean Scorecard: Total Lifecycle Focus


Observations from Device Tech

PROCESS: Device Tech’s High Voltage Switch

Failure Modes What was Observed? What was the impact?

Too Much, Too Early HVS Project Team designs a semi-automated conveyor line to meet their goal of minimizing direct labor

10 Year Lifecycle Cost Difference = $5.6 Million between Concept 1 & 8

Penny-Wise, Pound-Foolish

HVS Project Team decides to combine Overweld and Leak Test to save space on their conveyor concept

Bottleneck of 28 seconds vs. 15 seconds would have required investing in a second line

Too Little, Too Late HVS Project Team hands-off the process design to the Chicago plant without any input from the Operational Team

Concept 1 is 20% more costly than Concept 2, and 138% more costly than Concept 8

















10/18/2016

10



(Concept 1)


(Concept 2)

LPPD Team’s Final Proposal (Concept 8)

Cost per unit $2.26 $1.88 $0.95

Labor Efficiency

51% 86% 84%

Capital Spend

$3.6 Million $2.5 Million $0.5 Million

Total Lifecycle Cost


Results

Device Subs

Low Vol: Make-to-Order

High/Med Vol: Supermarket

Device Subassembly Cell

Straight Line for Plate & Final Assembly Cell

100

10b

50a 60a 90 70

20a

20b

10a

80

40 30

FIF

O

120 110b 110c 110a


(Concept 1)


(Concept 2)


Cost per unit $2.26 $1.88 $0.95

Labor Efficiency

51% 86% 84%

Capital Spend





(later, Concept 1)

Operational Team’s Counterproposal (later, Concept 2)


Cost per unit $2.26 $1.88 $0.95

Labor Efficiency

51% 86% 84%

Capital Spend





Key Learning at Device Tech

• Lifecycle Focus: Over 8x Payback for Upfront Engineering Effort

• Product Design Influence: Re-designed part eliminated Crimp step, reducing capital & manual time

• Quality Improvement: Crimp re-design improved quality & reduced key warranty issue

• Process Re-Selection: In-line bolt feeders reduced manual station time by 41% and process variability by 80%

• Resource Efficiency: People, Capital, Environmental

• Six different flexible process rates for Concept 8; One for Concept 1

• Learning Organization: People were given a chance to use their creative talents to add value for customers and Device Tech

• Knowledge Supermarket: Identified future opportunities for next generation product

• Simplified Process Management: Manual work close together, load & go stations close together

20

10/18/2016

11


How to embed LPPD

into the LPPD

System?


“Monday Morning”

• Pick a new product / process for your “learning laboratory”

• Nominate the system architect

• Take responsibility

• Link to product development for one overall Development Team

• Identify your responsible experts

• Be prepared to collaborate & co-locate

• Take a value stream view of the work

• Consider these phases

• 6Cs (Context, Concepts, Converge, etc.)

• 4:2:1 or some variation of set-based learning early-on

• Focus on the work & the key questions

• It will define the process

22

10/18/2016

12


Failure Modes Homework Template

lean process development: key questions to connect with ... · • process re-selection: in-line...

Documents