Concept to Launch

Lean Concepts for

Product/Process Design

A Collection of Learnings from

Project Reviews with Mr. Yamada

Compiled and Summarized

by Carl Nehmer
Release 1.4
Revised: December 15,2000
Delphi Energy & Chassis

Delphi Confidential
 Lean Concepts for Product/Process Design

Table of Contents
Forward....................................................................................................................................................................... iii
Think! The 80/20 Rule............................................................................................................................................. iii
The Four Rules of the Toyota Production System ................................................................................................... iii
The Good, The Bad and The Ugly........................................................................................................................... iv
Five Steps to Kaizen – O,S,K,K,K............................................................................................................................ v
Pilot Time Concept Issues ....................................................................................................................................... vi
Steps in Balancing a Cell Design:............................................................................................................................ vi
Lean Enterprise Concepts............................................................................................................................................ 1
Natural Cycle Time................................................................................................................................................... 2
Cell Design Strategy - Rabbits.................................................................................................................................. 3
Fab-Assembly Concept............................................................................................................................................. 4
Couple / Decouple .................................................................................................................................................... 5
Purchasing Strategy .................................................................................................................................................. 6
Mixed Model Error Proofing .................................................................................................................................... 7
Back, Back, Back...................................................................................................................................................... 8
Information Flow ...................................................................................................................................................... 8
Design for Manufacturing and Assembly ................................................................................................................... 9
PPAP-at-Beta.......................................................................................................................................................... 10
10 x 10 x 10 = Customer Differentiation ................................................................................................................ 11
Linear Assembly Flow............................................................................................................................................ 12
Repeated Process Operations.................................................................................................................................. 13
Transferable Work Elements .................................................................................................................................. 14
Cell Design Concepts ................................................................................................................................................. 15
Cell Design Strategy - Scaleable Investment .......................................................................................................... 16
Staff Out of Sequence ............................................................................................................................................. 17
Mixed Supercells .................................................................................................................................................... 18
Balanced Operations ............................................................................................................................................... 19
Small Lots – Create Time for Changeover ............................................................................................................. 20
A-B Control ............................................................................................................................................................ 21
Appendix A: Agenda for Yamada Reviews ............................................................................................................... 22
Appendix B: Concept to Launch News Clippings..................................................................................................... 23
Appendix C: Revision Record and Distribution........................................................................................................ 24

Yamada_Concepts14a.doc / can ii Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Forward
Mr. Yamada began meeting in May 1999 with Delphi-E Project Teams (See Appendix A for the suggested agenda
and content for the Project Reviews). During the discussions, Mr. Yamada makes observations on different Project
Team issues. These observations prompt Mr. Yamada to communicate “Concepts” learned during his years of
experience with Toyota. This report is a collection of those concepts that are important to the planning and execution
of projects.

The underlying themes of all concepts are:

• Lead-time reduction

• Utilization of Investment

• Utilization of People

• Quality
All of these concepts are relevant to Project Team responsibilities for Concept to Launch. Many of these concepts
may also be applied to Phase 3 continuous improvement activity on existing manufacturing systems.

Think! The 80/20 Rule

Mr. Yamada says, THINK! Creating a Product or Process concept is the standard approach 80% of the time. The
other 20% of the time the Engineer must study the problem and THINK before recommending a solution.
Engineering management must work the complex problems and make them simple to the floor.

The Four Rules of the Toyota Production System1

Mr. Yamada recommends this paper for learning about the Toyota Production System. It comes the closest to
understanding the Toyota Production System. The following is an excerpt from the paper:

“The tacit knowledge that underlies the Toyota Production System can be captured in four basic rules. These rules
guide the design, operation and improvement of every activity, connection and pathway for every product and
service. The rules are as follows:

1. All work shall be highly specified as to content, sequence, timing, and outcome.

2. Every customer-supplier connection must be direct and there must be an unambiguous yes-or-no way to send
requests and receive responses.

3. The pathway for every product and service must be simple and direct.

4. Any improvement must be made in accordance with the scientific method, under the guidance of a teacher, at
the lowest possible level in the organization.

All the rules require that activities, connections, and flow paths have built-in tests to signal problems automatically.
It is the continual response to problems that makes this seemingly rigid system so flexible and adaptable to changing
circumstances.”

1
“Decoding the DNA of the Toyota Production System” by Steven Spear and H. Kent Bowen, Harvard Business
Review, September-October 1999, pg. 98.
Yamada_Concepts14a.doc / can iii Printed: 01/08/2001
 Lean Concepts for Product/Process Design

The Good, The Bad and The Ugly

Transition
The
Road
to
Lean
Bad Good

The road to a lean enterprise poses an interesting paradox. Mr. Yamada observed that a “Bad/Weak
Company looks the same as an “Excellent/Strong” Company.2 Only the company in transition to a lean
enterprise looks different. Delphi is a company in transition. He cited the following observations (the
observations for the transition company were supplied by C. Nehmer):

Observation Bad Company Good Company Transition Company

Kanban None None Everywhere
Slogans: 1 piece flow, … None None Everywhere
Process Islands Everywhere Everywhere None
Molding Machine No Wait, always No Wait, always Waiting
running parts running needed parts
Assembly Machines Decoupled Decoupled Coupled 1 piece flow
Batch Size Large Different Sizes One Piece Flow
Muda (Waste) Muda is Muda Muda is Muda 7 types of Muda
No understanding With understanding Trying to understand

2
Project Review 11-Nov-1999 – Spin Form Converter
Yamada_Concepts14a.doc / can iv Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Five Steps to Kaizen – O,S,K,K,K

Plan, Do, Check, Action
1. Observe – Every operation performed the same
2. Standardization – all shifts; Motion, Material; Safety, Quality, Management
3. Kaizen – Flow & Process - Material & information first
4. Kaizen – Equipment – Simplify machine layout
5. Kaizen – Layout – always last

1. Observation
What is happening on the work area? Focus on machine / material stoppages. Most important- Don’t jump to the
next step. Make detailed observations and notes.
• Watch, review, and record each step of the process
• Break down the process and motion into discrete steps
• Identify sources of variation in material and information flow
• Identify all opportunities for standardization
- Maintenance System
- Quality System
- Material System
- Management System
- Always start and stop the line with material in the same status for breaks, shutdowns, lunch, etc. The
operator has a standardized motion to restart flow in the line
- Cell Process
- Information Systems
• After an initial observation
- Make a process vision for each department / area
- Track progress using a map based on these 5 Kaizen steps

2. Standardization
Why can we not standardize?
• Each activity, motion, method or process is performed in the same manner every time
• It provides predictability and the baseline for measuring progress
• Starting point for Kaizen
• Management responsibility to enforce and teach
• Interruptions must be standardized and then eliminated. Management issue not an operator issue
- Must be categorized as normal and extraordinary
- Normal are cyclical and must have countermeasures developed.
• Management must learn how to take advantage of interruptions based on the interruption’s cycle time, e.g.
changeover or maintenance during breaks

3. Flow & Process Kaizen

• Improve the material and information flow on or to the work area
• Rebalance and combine various parts of the baseline process
• Improve all routes (scrap, audit, maintenance)

4. Equipment/Tooling/Machine Kaizen
• All improvements should support the operator
• Improve future placements
• Improve machine cycle times
• Simplify machines
• Group but do not combine process steps
• Standardize

Yamada_Concepts14a.doc / can v Printed: 01/08/2001

 Lean Concepts for Product/Process Design

5. Cell Layout Kaizen

• Improve the Man-Machine-Material ratios through changes to the overall layout of the cell
• Scope can be simple rearrangement or complete cell shape/flow redesign
• Taichii Ohno, author of “Toyota Production System”, always asks for O-S-K-K before Cell Layout
Kaizen. Never signed an approval without O-S-K-K.
• Work on old machines to find problem; always reduce investment
• Expect new machines to provide a 10 times benefit
• Kaizen existing machine then we can make better expectations to the machine maker for a new machine.

Pilot Time Concept Issues

Mr. Yamada presented this concept as a teaching at the Concept to Launch Workshop 17-Nov-1999 – Fuel
System Products.
1. Study Machine Cycle time – Program change, gear change, motor change to achieve cycle time.
2. Machine shape / design
3. Parts Presentation
4. Safety concept – Buttons, Light curtain, shield, In/Out slide
5. Preventive Maintenance of machine –
a) Replace after quantity? Plan all steps of PM
b) Accidental – Not so often
c) Adjust for position due to machine vibration; visual controls
d) Dirty – How to clean machine
6. Study of Up Time – Top Ten and Top Three, drive requirements back to machine builder

Steps in Balancing a Cell Design:

1. Machine Balance first
2. Motion Balance – Limit the number of motions for an operator; 80 motions are not repeatable, 12 motions are
repeatable.
3. Part Presentation
4. Machine Uptime and Quality planning.
A cell concept with only two operators is very bad for volume fluctuations. One person out changes the capacity by
50%. A good cell design staffing level is 5 to 15 people. It is difficult to balance and manage for greater than 15
people. It is very difficult to make a kaizen of 20% to remove one person from a line with 5 people.

Yamada_Concepts14a.doc / can vi Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Lean Enterprise Concepts

Mr. Yamada has repeatedly emphasized the first five concepts in this section. The concepts are listed in the
sequence he numbered them at the Converter workshop on November 11, 1999. These first five concepts represent
the need to think about the cell design. Mr. Yamada emphasizes that there is not a single answer for lean
manufacturing. Many consultants have fostered digital thinking for lean manufacturing: “Single piece flow”, “Zero
Inventory”, “Coupled processes”, “No Process Islands”, “No Buffers”. There are logical exceptions to all of these
lean-manufacturing rules. Mr. Yamada keeps emphasizing that there is no substitute for thinking. Do what is smart,
not what a consultant or book tells you.

Yamada_Concepts14a.doc / can 1 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Natural Cycle Time

Takt time = 200 sec.

80
Cycle time target
= 15 sec. J
15
Assembly Operation 12
Cycle Times 10 8

Keywords: Manufacturing System Strategy

Concept: Design processes to a Cycle Time NOT to a Takt Time. Every product family should have a Cycle
Time target. All of the operations for the product family should be designed to operate at the
target. The Natural Process Cycle Time will be based on part size and technology. The Chief
Engineer for the product family is responsible for setting the cycle time target.

Observation: This concept was observed during the Escalade Engine Air Control Valve project workshop.3 The
Takt time for the project was about 200 seconds. The cycle time varied greatly between
operations. The MSD workshop concluded with an operator “chase” to balance work.

Discussion: Operations less than the target cycle time are opportunities to reduce machine complexity, reduce
investment and/or add content to the operation. There is no value in having any operation less than
the target cycle time.
Operations above the target cycle time require effort to break apart or reduce content to bring the
operation down to the target. Duplicating the equipment to balance cycle time only works for high
volumes. Balancing through duplication is a waste of capital for low volumes.
With common Cycle Times, Takt time becomes an integer multiple of the cycle time. The line can
be operated at many different Takt times.
At Toyota, Takt time is a variable. Toyota management sets Takt time monthly. Nobody knows
the future. Customer requirements change all the time. Customer estimates are usually too
optimistic.
Run-at-Rate should be a demonstration of cycle time, not Takt time. Toyota looks at the company
management. They trust the company to ramp up capacity.

Contributed by: Rob Garrick, Carl Nehmer

3
Project Review 8-Oct-1999 – Air Control Valve
Yamada_Concepts14a.doc / can 2 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Cell Design Strategy - Rabbits

Rabbit Minimum Cell Design Strategy Minimum

Strategy Cycle Time Cycle Time
1st increment of
Capacity 30 20 10 30 sec. 30 20 10 30 sec.

2nd increment of
Capacity 30 20 10 30 sec. 30 20 sec.

3rd increment of
Capacity 20 15 sec.

4th increment of
Capacity 30 20 10 30 sec. 30 10 sec.

The numbers in the boxes represent the cycle time of different processes in a cell.

Keywords: Cell Design Strategy

Concept: The investment plan should comprehend many small increments of investment for customer
volume changes. Keep expanding a cell until all equipment is fully utilized. Do not make the
expansion investments until the customer volumes materialize.

Observation: The Mr. Yamada presented this concept as a teaching following a discussion of the Spin Form
Converter project4. He has observed in the Milwaukee Converter plant and elsewhere that we
design a cell and then multiply it like “rabbits”. We don’t think about fully utilizing investment
before building the next cell.

Discussion: Mr. Yamada used the very simple example pictured above to demonstrate this concept. The simple
example fully utilizes all operations after the 4th increment of capacity. Under these circumstances
it would be acceptable to begin duplicating this perfectly balanced cell. The real world seldom
achieves a perfect cycle time balance across all operations. The “Cell Design Strategy - Scaleable
Investment” on page 16 discusses expansion of capacity beyond one basic cell.

The Rabbit strategy defines a product cell and then duplicates the cell for additional capacity
requirements. This appears to be very foolish when compared to the Cell Design Strategy.
However, let’s assume that the 20 sec. and the 10 sec. processes represent only 5% each of the
total product cell investment. Many engineers and plant managers will argue for the small
additional investment for product cells that all look and operate the same way. Engineering
management must work the complex problems and make them simple to the floor.

Each Rabbit Product Cell will typically be dedicated to a single customer and/or part number. This
mentality limits the ability to adjust capacity for shifts between customers and part numbers. This
strategy leads to over capacity as each new customer application gets its own product cell.
Manufacturing does not make changeovers a routine.

This concept is also important for scaling down investment at the end of a product’s life.

4
Project Review 11-Nov-2000 – Spin Form Converter
Yamada_Concepts14a.doc / can 3 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Fab-Assembly Concept

3 Shifts Decouple 2 Shifts

Buffer
Fab Assembly

High Capital Processes Buffer High Labor Processes

Running, Running, Running Empty Every Day Day by Day Catchup

Keywords: Manufacturing System Strategy

Concept: High Investment and Low Labor content characterizes Fabrication processes. They should be run
on THREE shifts. Low Investment and High Labor content characterizes assembly processes.
They should be run on TWO shifts. Decouple high investment processes from assembly
operations. Buffers are good. Centralized high investment processes are good for maintenance.

Observation: Mr. Yamada has presented this concept as a teaching in several workshops.

Discussion: Buffer cost versus Investment Cost – Labor has an investment cost of hiring, shift premiums and
absenteeism to operate on third shift. Operating with no inventory is more expensive than
operating with a properly managed inventory.

Mr. Yamada tells the story of a Toyota plant manager that was criticized by his boss for having
too much inventory. The plant manager eliminated most of the inventory. His boss visited the
plant a second time and criticized the plant manager for having too little inventory. Too much
inventory and too little inventory are both signs of a poorly managed operation.
Toyota in Japan has 250 suppliers. 60% of the suppliers operate on one shift; 40% operate on two
shifts. 70% of Toyota’s material needs are produced on two shift operations.
Mr. Yamada discussed two-shift operations separated by a shutdown period of about two hours.
This reduces parking lot size and congestion at the arrival/departure times. Most importantly it
allows for making up volume on the same day and same shift. Toyota workers stay on the job until
the day’s schedule is completed. Management should not accumulate shortages during the week
and then schedule Saturday overtime to complete the schedules. This leads to bad worker and
management habits during the week.

Yamada_Concepts14a.doc / can 4 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Couple / Decouple

Countermeasure Buffers

100% 85% 3 2
shifts shifts

Uptime, Yield, Complexity or

Many-to-One / One-to-Many Cycle Time Differences - Different Number
Process Cell Relationships Decouple with Buffers of Shifts

Keywords: Manufacturing System Strategy

Concept: There are many good business reasons for decoupling processes. Decisions to couple or decouple
are a project by project discussion. Buffers are a countermeasure for uptime, yield, complexity or
cycle time issues. The size of the buffer is a visible measure of the magnitude of the uptime and
yield issues.

Observation: The MFI Project5 required a group of three machines in parallel to feed 20 leak test stands. The
processes following leak test were another group of three machines in parallel. Mr. Yamada
decoupled the three groups of processes.
The Multec 3 Project team6 has a process with an uptime issue in the middle of the assembly line.
A buffer capability was added on both sides of the process to protect the line from flow
interruptions.

Discussion: The Countermeasure buffer on the Multec 3 line is a visual control for process capability. The
buffer inventory is always in the buffer following the process. Buffer inventory in front of the
process means the process experienced a breakdown. The operator must continue working after
the line stops until the inventory is in the buffer following the process. The inventory is sized for
the standardized repair procedure. The same type of buffer can also be used for yield problems and
changeover requirements.
The MFI Project team is challenged with providing a process solution for a 15-to-1 difference in
capacity requirements. The high volume requirement is for two years or less. The low volume will
be for many years. There is also customer uncertainty on an intermediate capacity. Mr. Yamada
worked with the team to develop a plan to transition between the different capacity levels. The
high volume plan involved decoupling into three groups of processes.

5
Project Review 19-Nov-1999 – MFI Project
6
Project Review 19-Nov-1999 – Multec 3 Project
Yamada_Concepts14a.doc / can 5 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Purchasing Strategy

We want YOU to supply these very important parts.

Keywords: Purchasing

Concept: Engineering must make the supplier recommendation on the most important/critical parts.

Observation: Mr. Yamada has presented this concept as a teaching in several workshops

Discussion: Purchasing for important / unique / critical parts is a teamwork effort. The Engineer makes the
supplier recommendation. Kaizen with a good supplier to reduce cost. This is much easier for the
enterprise then developing the capabilities of a low cost supplier.

Toyota has 250 good suppliers. Delphi has 4500 suppliers. Fixed cost is in many places. Contact
with the suppliers is limited. Delphi should spend its time with good suppliers and work with them
to lower cost.

It is very difficult to manage volume ramp ups and build-outs when you have seven suppliers for
the same material.

Yamada_Concepts14a.doc / can 6 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Mixed Model Error Proofing

Pack Quantity

A B C D E

Keywords: Error proof based on pack quantity

Concept: Prevent quality problems based in our material flow

Observation: This concept was observed during a Wabash/Delta project review at MTC Mexico. The cell was
designed to build product for two different customers. The challenge is how to keep from mixing
parts.

Discussion: When you have several models running in a cell, a system to control that you are shipping the
correct model and not mixing materials is to supply the line with the same pack quantity/multiples
that you have for your final customer. You use the Kanban system when you supply the material
to the line. The same Kanban quantity is used for shipping orders. This way you can feedback to
the system that you used all the material. Also if we have rejects during the line or loses for any
cause, we can take action for this abnormal situation (using a safety stock), so we can keep the
abnormal separated from the normal.

Contributed by: Monica Feregrino

Yamada_Concepts14a.doc / can 7 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Back, Back, Back

Keywords: Error Proofing, Quality, Manufacturing, Supplier Quality

Concept: No part quality inspections at Final assembly. Move all part quality inspections “Back, Back,
Back” to the supplier of the part.7

Discussion: Move all quality inspections back, back, back to the supplier of the part. The cost of a part defect
found at final assembly now includes the final assembly labor and the part cost of other good parts
assembled to the defective part.

Part feature error proofing equipment at final assembly is redundant to quality processes at the part
level. This increases investment and labor cost for final assembly without adding value. The
information acquired from the error-proofing equipment is not effective in controlling the part
fabrication process.

Information Flow

Keywords: Error Proofing, Manufacturing, PC&L

Concept: The Kanban pack information identifies the customer information for changeover of the line.
Avoid making in-process labels specific to customers.

Observation: A serialized label is required in the Airmeter process to associate sensor measurements made at an
early step in the process to the calibration late in the process.7 The serialized label also contained a
customer code to select the calibration points in the process. The customer code on the label
requires a changeover of the label for each different customer part number.

Discussion: Let the information flow make the association to customer specific requirements. The leadoff
operation in the cell has the customer build information. Electronically associate that information
with the serialized process flow label to identify customer specific requirements.

7
Project Review 20-Sept-1999 – Bidirectional Airmeter
Yamada_Concepts14a.doc / can 8 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Design for Manufacturing and Assembly

The following concepts relate primarily to Product Design issues for lean manufacturing systems. The concepts
suggest ways that we need to think differently about designing the product. They encompass project management
and the design of the product for the lean manufacturing system.

Yamada_Concepts14a.doc / can 9 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

PPAP-at-Beta

I’ll bet a production tool on your first design!

Keywords: Lead-time Reduction, Knowledge Base

Concept: Decouple Project Timing from Customer Timing. Engineer the product and process completely
and correctly the first time and deploy the engineers to the next project.8 Supply Beta parts from
Lean Production Equipment and Tools.

Observation: Mr. Yamada observed that the durations on Delphi’s Project Schedule expanded to fill the time
available when the customer delayed introduction of their vehicle. He also observed that we
expend effort re-engineering parts of the product/process that are functionally carryover from the
prior design.

Discussion: Mr. Yamada’s observations on program timing, reuse of knowledge from prior designs and
learnings from the Method’s Lab workshop9 formed the basis for the PPAP-at-Beta concept. Lean
Tooling concepts are simple to build and require minimal investment. This enables building the
production assembly tools concurrently with the first customer samples. Customer samples can
then be assembled using regular manufacturing workers. Customer sample pricing and lead-time
to production parts now become a competitive advantage.

Mr. Yamada has stated that the leading vehicle manufacturers are focused on lead-time reduction.
Concept to Launch for vehicle manufacturers (Toyota) is expected to drop to 9 months in the near
future.10 Volkswagen’s Piëch stated in an interview that “Our time capability from design freeze to
Job One is a secret, but sometimes it can be less than a year.”11 That leaves suppliers about 3
months to complete their work. There is no room for engineering the product multiple times. It
must be right the first time. It must be engineered based on a strong Product Knowledge Base.
Imagine the competitive advantage of a 3-month Concept to Launch capability for Application
Projects. We must practice being fast to market so that we’re ready when a customer needs the
capability.

Customer changes after PPAP-at-Beta are a non-issue for our Fast-to-Market customers. Changes
by customers that operate at three to four levels of prototypes are a commercial issue.

Contributed by: Jose Avila, Carl Nehmer

8
Project Review 6-June-1999 – Delta Cruise Control
9
Method’s Lab Workshop 20-May-1999
10
Project Review 20-Aug-1999 – Multec 3
11
Automotive Engineering International / June 2000, pg. 69
Yamada_Concepts14a.doc / can 10 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

10 x 10 x 10 = Customer Differentiation

10 10 10
Fabricated Sub- Final
Parts X Assembly X Assembly
Configurations Processes

= 1000 Product Variations!

Keywords: Design for Assembly, Product Planning

Concept: Minimize Final Assembly product variations dependent on fabricated parts.

Maximize product variations dependent on the Final assembly process. Final assembly is the
easiest place to make changeovers to comply with customer orders.
10 designs of fabricated parts (reuse of existing tooled parts) TIMES
10 sub-assembly configurations of fabricated and commodity parts TIMES
10 Final assembly configurations of sub-assemblies and fabricated/commodity parts EQUALS
1000 Unique Product Variations for Customer Differentiation.

Observation: Mr. Yamada shared this concept during a Modular Reservoir Assembly Lean workshop with the
Product Chief Engineers in the Flint IE Methods Lab.12 The MRA consists of many fabricated
parts and sub-assemblies that are combined in various configurations at Final Assembly. The ease
of moving different final assembly workstations in and out of the process was apparent during the
exercise.

Discussion: The Best Design is the one that accommodates all customer variations anywhere in the last
continuous process flow, typically final assembly.
This is a very important concept when it comes to supply chain management for just-in-time and
sequenced-in-line delivery to our customers. Delphi received the “Excellence in Logistics” award
for the supply system that “allows for Mercedes to install a finished cockpit assembly into a
vehicle in just 120 minutes from the time we receive the order.”13

12
Method’s Lab Workshop 20-May-1999
13
Delphi Dateline 23-Sept-1999
Yamada_Concepts14a.doc / can 11 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Linear Assembly Flow

Complex Sub- B C
Assembly
A D E

Linear Flow A B C E
D

Keywords: Design for Assembly

Concept: Linear product build sequences enable balancing cycle times and smooth material flow.

Observation: This concept was observed during a Modular Reservoir Assembly (MRA) Lean workshop with the
Product Chief Engineers in the Flint IE Methods Lab.14 An MRA sub-assembly operation was
required in the middle of the final assembly process. This disrupted material flow and operator
balance.

Discussion: A sub-assembly operation that is included in the final assembly process poses challenges for the
MSD workshop. The high work content at the sub-assembly operation drives complexity into the
material flow of the work cell. The partial assembly completed upstream from the sub-assembly
must flow around the sub-assembly in the work cell. The sub-assembly could be performed
offline; this limits the staffing flexibility and operator balance for the product cell.

The L6 Ignition Coil Project Team15 applied this concept (based on MSD workshops) to the
secondary coil winding operation. The conventional approach was to build the secondary coil
subassembly, wind the coil and then add it to the in-process assembly with the primary coil. The
project team made innovative product and process changes to achieve Linear assembly flow. The
secondary coil bobbin was first assembled to the in-process primary coil assembly. The entire
assembly was then placed on the winding equipment to wind the secondary coil.

14
Method’s Lab Workshop 20-May-1999
15
Concept to Launch Workshop 30-Sept-1999 – Ignition Products
Yamada_Concepts14a.doc / can 12 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Repeated Process Operations

Assembly Stations
Process Sequence A W1 C W2 E

Concept Example A C W1,2 E

Keywords: Design for Assembly, Lean Investment

Concept: Revise the product/process design to combine high investment processes into the same process
step.

Observation: The original Bidirectional airmeter design16 concept had wirebond operations at three different
locations in the manufacturing process. Changes to the product and process allowed all of the
required wirebonds to be performed at the same station in the manufacturing process. This
eliminated two wirebonders and resulted in higher utilization of the remaining wirebonder. A
substantial investment expense was eliminated.

Discussion: Many products require the repeated use of capital intensive manufacturing processes to complete
the assembly of a product. In the case of the Bidirectional airmeter, the repeated process is a
wirebonder. In the case of a Fuel Injector, the repeated process is Laser welding (3-6 welds per
assembly).

During a Design for Manufacturing workshop, the DI-G Injector Team examined the weld
parameters for each weld operation. Next they reviewed design changes required to achieve
convergence of the weld parameters. They also reviewed product and process changes to move
multiple welds into the same process step.17 Some progress was made in the workshop; a lot of
homework was assigned to address the product and process design issues.

16
Project Review 20-Sept-1999 – Bidirectional Airmeter
17
Design for Manufacturing Workshop 28-Sept-1999 – DIG Injector
Yamada_Concepts14a.doc / can 13 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Transferable Work Elements

Assembly Stations
Build Sequence A B C D E

Required Assy time: 10 5 10 5 10

Balanced Assy time: 10 10 10 10 10 = 50

Concept Example A B D C E

Required Assy time: 10 5 + 5 10 10

Balanced Assy time: 10 10 10 10 = 40

Keywords: Design for Assembly

Concept: A transferable product build work element enables balancing cycle times by moving an assembly
operation to a different position in the assembly process.

Observation: This concept was observed during the Modular Reservoir Assembly (MRA) Lean workshop with
the Product Chief Engineers in the Flint IE Methods Lab.18 The MRA permitted some assembly
operations to be performed in different sequences. This facilitated balancing of cycle times.

Discussion: The product build sequence requirements can significantly impact the success of MSD workshops.
Fixed build sequences limit the ability to shift work between operator stations to achieve Operator
Balance.

A transferable product work element also supports operating the assembly line at different staffing
levels. The above example is balanced with four operators at 10 time units each. It will also
balance with two operators at 20 time units each. The original assembly sequence will not balance
with two operators. A small imbalance occurs for three operators with 15,15,10 time units.

18
Method’s Lab Workshop 20-May-1999
Yamada_Concepts14a.doc / can 14 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Cell Design Concepts

This section contains additional concepts applicable to Cell design.

Refer to the Lean Enterprise Concepts section for cell design strategies as follows:
• Natural Cycle Time, page 2
• Cell Design Strategy - Rabbits, page 3
• Couple / Decouple, page 5

Yamada_Concepts14a.doc / can 15 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Cell Design Strategy - Scaleable Investment

Cell #2

Investment $’s
Cell #1

Cell #1 with
4-1-4 Material Flow and
additional equipment

Tooled Capacity

Keywords: Cell Design Strategy

Concept: The investment plan should comprehend many small increments of investment for customer
volume changes. Keep expanding a cell until all equipment is fully utilized.

Observation: The Bidirectional Airmeter investment plan comprehended only three different capacity levels for
the product cell. The capacity plan duplicated the entire cell even though several stations
representing 5 % of the total investment were operating at half capacity.

Discussion: The following is a paraphrase of Mr. Yamada’s discussion19:

“Delphi is not a rich company. We need to use ours heads to figure out how to
fully utilize every piece of equipment in a product cell. A product cell should
not be replicated until all equipment is fully utilized. Don’t waste capacity even
on low investment stations. It is a management problem to use their heads to
figure out how to flow material in a more complex product cell configuration.
Capital $’s are a very precious resource; thinking is free.”

The expansion of the cell should be constrained within the range of manageable staffing
levels for a cell. A good cell design staffing level is 5 to 15 people. It is difficult to
balance and manage for greater than 15 people. It is very difficult to make a kaizen of
20% to remove one person from a line with 5 people.

19
Project Review 20-Sept-1999 – Bidirectional Airmeter
Yamada_Concepts14a.doc / can 16 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Staff Out of Sequence

Assembly Stations
Build Sequence A B

B&D same
operator C
E D

Required Assy time: 10 5 10

10 5
Balanced Assy time: 10 + 10 + 10 + 10 = 40

Keywords: MSD Workshop

Concept: Staffing the cell operations out of sequence enables balancing operator workload. Arrange the cell
to accommodate the operator.

Observation: An operation at the beginning of the line and one at the end of the line both had high operator wait
times.20 The adjacent operations on the line were balanced. Operator workload was balanced by
physically arranging the line so that one operator could perform both the first and the last
assembly operations.

Discussion: Work cells are sometimes arranged to accommodate one operator working at two operations that
are non-adjacent in the build sequence. The preferred approach to balancing is “Linear Assembly
Flow” on page 12.

20
Project Review 20-Sept-1999 – Bidirectional Airmeter
Yamada_Concepts14a.doc / can 17 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Mixed Supercells

Rabbit Strategy Mixed EACV SuperCell

Existing EACV Cell

“Escalade” EACV Cell

Unique tooling and equipment for the Escalade EACV was integrated into an
existing EACV cell.

Keywords: Manufacturing System Design

Concept: Integrate new product programs into existing product cells. Add a few pieces of equipment and
achieve production through changeovers and staffing levels.

Observation: This concept was observed during the Escalade Engine Air Control Valve project workshop.21 A
new product cell was planned to accommodate the new EACV design for the Escalade application.
The cycle times for the cell ranged between 8 and 80 seconds. The Takt time for the application
was 200 seconds. This created a situation of severely underutilized equipment.

Discussion: The project team returned home after the workshop and developed a plan to integrate the
production of the Escalade EACV into the existing Engine Air Control Valve product cell. This
created a super cell that could manufacture both the existing product and the Escalade Engine Air
Control Valve.

Contributed by: Rob Garrick

21
Project Review 12-Oct-1999 – Air Control Valve
Yamada_Concepts14a.doc / can 18 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Balanced Operations

30 60 seconds per
piece equals
60 pieces per
Assembly Operations hour with one
10 10 10
operator
Staffing Options:
1 operator = 60/hr
2 operators = 120/hr Takt time = 30 sec.

Balanced Operations 10 10 10 10 10 10
Staffing Options:
1 operator = 60/hr
2 operators = 120/hr
3 operators = 180/hr
6 operators = 360/hr

Keywords: Manufacturing System Design

Concept: Break complex machines into less complex work operations. This permits the cell to be operated
at different staffing levels to achieve a wide range of production rates with no change in
investment.

Observation: This concept was observed during the Modular Reservoir Assembly (MRA) Lean workshop with
the Product Chief Engineers in the Flint IE Methods Lab.22 Several operations for the MRA were
performed in the same workstation. The workstation operated at the current Takt time for the cell,
but limited the ability of the cell to be operated at other Takt times.

Discussion: The high content workstation was broken into several simple workstations. Production was tripled
from 120 parts/hour to 360 parts/hour without the need to purchase new equipment. This simple
example was created to demonstrate a point and get you thinking about the advantages of a “lean
process” scenario. The advantages as volume increase should be obvious. Also, do not forget the
aftermarket or service part of a product life cycle. This same concept can be used to lower your
production as the volumes go down. This one line can run anywhere from 60 pieces per hour to
360 pieces per hour based upon the needs of the customer.

Contributed by: Gene Dura

22
Method’s Lab Workshop 20-May-1999
Yamada_Concepts14a.doc / can 19 Printed: 01/08/2001
 Lean Concepts for Product/Process Design

Small Lots – Create Time for Changeover

Buffer Screen Buffer

Printer

Auto Auto
Load Unload
External
Changeover
Tasks

Keywords: Manufacturing System Design, Cell Design Strategy

Concept: Small lot automatic load and unload for a machine can be used to collect small pieces of operator
time into larger blocks of operator time. The large block of time can then be utilized to
prepare/complete the external changeover tasks for the machine. The operator motions for the
changeover are standardized.

Observation: The screen print operation for the Flat Plate Oxygen Sensor Element required a significant
changeover time for each of the seven printing operations on each assembly. The high-investment
screen printer equipment had adequate capacity for the initial volumes. Future capacity would be
constrained by the changeover time. The Project Team was seeking assistance to manage material
flow and machine changeover schedules.

Discussion: The changeover process for the screen printer was divided into internal work and external work.
Internal work is the activity that can only be performed while the equipment is down and not
producing parts. External work is the activity that can be performed while the machine is operating
and producing parts. This had an immediate impact on capacity but created a challenge of how to
perform the external changeover work while keeping the material flowing through the screen
printer.

Mr. Yamada proposed a buffer before and after the screen print operation. This allowed the
operator to perform the load/unload task for an entire batch of parts. While the machine cycled
through the buffer, the operator had sufficient blocks of time to complete the external work
changeover operations. Capacity was significantly increased.

The size of the buffer was also discussed. Downstream from the screen printer are several batch
operations. Mr. Yamada proposed the criteria for buffer size is an integer multiple/divisor of the
downstream batches. He also stressed the importance of standardizing the external changeover
work. The duration of the standardized external changeover work elements would help define the
size of the buffer.

Yamada_Concepts14a.doc / can 20 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

A-B Control

Mach #1 Mach #2
5sec. 5sec.
B A B A

<5 sec.
A in the machine
Conveyor travel time is less than the
B ready to go
machine cycle time, no disruption of flow.

Mach #1 Mach #2
10sec. 5sec.
A B A

>10 sec.
Conveyor travel time is greater than the A out of the machine
machine cycle time, disrupted flow. B is NOT ready to go
1. Conveyor Stopping
2. Free Conveyor

Keywords: Manufacturing System Design

Concept: Conveyor systems must be designed such that with Part A in the machine, Part B (the next part)
must be ready to enter the machine. No waiting by the machine.

Observation: Almost every conveyor system suffers from this condition. Conveyor controls are designed
without understanding of the processes on the line. More complexity is designed into the conveyor
system than is required for lean operation of the line.

Discussion: The first graphic depicts the optimum situation: all machines are balanced and the travel time
between machines is less than the cycle time of the machines. Conveyor controls are very simple.
Both machines release their part at the same time and the controls stop the part in front of the next
machine. The line requires two pallets for each machine; one pallet in the machine and one pallet
traveling to the machine. The line always starts and stops with one part in the machine and one
part waiting.

The second graphic depicts several complications may occur in conveyor systems. The first issue
is unbalanced cycle times. The concept of AB Control says that we should normally hold the part
in the machine with the short cycle and release the part simultaneous with the longest cycle time
machine. The line can only produce parts at the cycle time of the slowest machine. There is no
value to complicate controls and/or conveyor design by releasing the part early.

The other issue with the second graphic is the travel time between machines. The travel time is
greater than the cycle time of the line. This requires that a second part/pallet is required between
Machine #1 and Machine #2 so that a part is always waiting for Machine #2. It may be possible to
utilize the short cycle time of Machine #2 to compensate for the long travel time.

Yamada_Concepts14a.doc / can 21 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Appendix A: Agenda for Yamada Reviews

Jose Avila (8/15/99) and Mr. Yamada have agreed on a suggested review agenda for Strategic/Forward projects.
These are working meetings, not customer PowerPoint presentations. Mr. Yamada is demanding on people to BE
PREPARED. We must make good use of his time.
1. Product description (Typically 15 minutes)
Presented by a very knowledgeable product engineer. The purpose is to give Mr. Yamada some product
background; he will ask questions to further his understanding. The discussion should include the following:
• Project Scope – Include target price, cost, investment.
• Discuss the design by SBCE (Set Based Concurrent Engineering) Sub-systems. Is it something we know
how to do? Is it new to Delphi and familiar in the industry? Is it high risk to Delphi?
• Features and parts that change with customer requirement changes.
2. Competitive Analysis / Spider charts (Typically 10 minutes)
• What is really good about the product and these ideas? What is your competitive position?
• What is the marketplace strategy: Low Cost Producer? Technology Leader? Intimate with Customer?
• Project planning life cycle volume curve and manufacturing location strategy; include alternate rollout
scenarios if applicable. Include end-of-life service volume estimates.
3. Design for Manufacture features. What have we done in the design to help manufacturability? Design for lower
cost, manufacturability or both?
4. Program timing, especially emphasis on PPAP-at-Beta. What are we doing to accelerate timing? Mr. Yamada’s
concept is that Capital is made of two things: 1) Equipment / Tools and 2) Utilization of skilled people. We
want to see fast-to-market, PPAP-at-Beta. What is the utilization of both people, machines and tools in an
accelerated way to achieve create good results?
5. Process Flow in divisional standard charts: Cycle time, Machine/Operator balance, scalability of line with
volume (volume at different operator staffing levels) and Process flow. Have an overall layout of the process
and number each step/station in the process. Complete a matrix that shows the manual operator time, machine
time, changeover time, uptime and scrap rate by process step/station. When possible, number parts and
subassemblies to correspond to the process step/station.
6. Lean Capital. What approach are we taking to lean equipment design by the process tool engineers?
7. Mock-up and build demonstration review (Typically 2 hours)
The project team must be prepared to demonstrate the operator motions for each step/station in the process.
The manual operator time, machine time, changeover time, uptime and scrap rate should be available at each
step/station. The mock-up should include the part presentation plan. Talk over the mock-up with Mr. Yamada. He is
very experienced in working through the operator motions and balancing cycle times.
It is very important to begin a mock-up of the process concurrently with the development of the product. The
process can be developed with surrogate parts that represent the general size and shape of the proposed part.
Parts for the mock-up should be creatively gathered, not fabricated.

It is suggested that 4 hours be allowed for a Project review. Two hours of that time should be planned at the Mock-
up and build demonstration review.

Yamada_Concepts14a.doc / can 22 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Appendix B: Concept to Launch News Clippings

Lead-time Reduction Enablers
Ford’s Excursion: “Platform sharing and computer-aided engineering (CAE) permitted them to go directly into
production tooling, shaving approximately six months off the program.” (“Making The Monster”,
Automotive Industries, August 1999, pg. 45)
Nissan Cuts Development Time: “In cutting development time (from 29 months five years ago to 15 months on
the all new Tino small car), Nissan went to more computer-based validation processes, application of high-
speed prototype machining and more off-line programming of robots. The company has been able to cut its
launch costs by 40%, prototype outlay by 60%, engineering man-hours by 30% and capital investment by
40%, Mr. Takagi says (Senior Vice President).” (“Nissan Cuts Cycle From 29 to 15 Months”, Ward’s Auto
World, September 1999, pg. 75)
Quick quality is goal at Ford: “In the industry-wide race to get new models out faster, Ford Motor Co. has opened
the first phase of a test facility that may cut in half the time it takes the company to bring a new car or truck
to market. The automaker hopes the $100-million wind tunnel and dynamometer test facility in Allen Park
will help it shave 12 or 13 months off the 24 to 28 months now required to move vehicles from final
design to first production. Giving customers more for less -- and faster -- has become a sort of
subcompetition in the auto industry. Japanese automakers are the fastest. Toyota, Nissan and Honda take
15 to 18 months for product development, according to Daron Gifford of Deloitte Consulting in Detroit.
U.S. automakers, in general, are still trying to get down to 18 months. (Apollo 3/31/2000, by Charlotte W.
Craig Detroit Free Press Automotive writer)
VW’s Piëch: Quality Improves as Design Time Decreases: Says the enigmatic Piëch, “In the morning we
develop a product, in the afternoon another; sometimes there are slight conflicts, so in the evening we fix
them. We find a solution if things are in conflict. We do not believe the design time, cost, and quality
curves are going in different directions. We find that quality improves as design time comes down, not the
reverse. Our time capability from design freeze to Job One is a secret, but sometimes it can be less than a
year. Can we go further? Why not?
“We have a car coming at the start of 2001 and made major changes to the body-in-white in January this
year. I cannot tell you about the outcome yet, but in January we had the first drawings of the dramatic
change in the car and four weeks later we could order the tooling. This kind of parallel work does not cost
more. It only costs more if, for example, a new engine block is needed and we have to re-tool for that.
Technologies are not the answer. It is experience that really matters—and good people. Use technology but
don’t rely on it too much; that is my philosophy.” (Automotive Engineering International, June 2000, pg.
69, by Stuart Birch, European Editor)

Yamada_Concepts14a.doc / can 23 Printed: 01/08/2001

 Lean Concepts for Product/Process Design

Appendix C: Revision Record and Distribution

This is a work in progress. Please forward all suggestions and comments to:
Carl Nehmer
Delphi-E&C Mailcode 483-3DB-210
Brighton Technical Center
12501 E. Grand River
Brighton, MI 48116-8326
810.494.5800 Fax: 810.494.4689

Release 1.0 02Nov1999 Initial Release

Release 1.2 14Jan2000 Completely revised
Revise Appendix A: Agenda for Yamada Reviews
Release 1.3 29Mar2000 Add “Mixed Model Error Proofing”
29Jun2000 Add new clipping: VW’s Piech
Release 1.4 15Dec2000 Add “A-B Control” and OSKKK

Distribution:
The following are on distribution for each release:
Engr. Board: S. Bailey, J. Avila, W. Gillespie, L. Ostrander, J. Overly, J. Botti, D. Kessler, M. Gustanski
EMS/P Staff: R.Bremer, J. Hutson, S. Kiefer, J.Luckman, T.Reddington, M. Sheline, M.Shost, H. Sullivan,
B.Timmers, S. Aikman
CSC Staff: G. Cameron, M. Depoyster, Nick Jones, J. Zizelman, G. Sims
Lean Team: D. Logozzo, F.C. Brown, E. Ethington
ME Staff: C. Buttry, D. Draeger, J. Galonska, M. Golla, D. Hahn, D. Haney, G. Hill, R. Kimes, A. Lee,
D. Mann, R. Perlet, F. Pirrello, C. Ramsey, M. Rasmussen, J. Rolwing, J. Robeson, J. Schmidt,
A. Webster, W. Matuschek, Jean-Jacques LAMBERT
IE Staff: M. Dietz, J. Rolecki, M. Feregrino
Project Leaders: J. Sofianek, P. Wendt, P. Jankowski, D. Ehle, M. Rasmussen, M. Murphy, C. Buttry, W. Keller,
S. Caster, J. Niemeier, S. Medwid, R. Garrick
Others: D. Kraush, G. Mansfield, E. Northern, T. Branciforte, K. Roessler, S. Reid

Yamada_Concepts14a.doc / can 24 Printed: 01/08/2001