Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Ten Questions

to Answer
Before Designing Mixed Model Value
Streams

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 1
Contents
The Impact of Product Mix on Flow .......................................................... 3

Value Stream Mapping Method for Mixed Model Flow .......................... 4

10 Questions for Mixed Model Flow ......................................................... 5

Four Questions In Depth for Value Stream Success .............................. 6

Mixed Model Value Streams for Growth .................................................. 10

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 2
The Impact of Product Mix on Flow
Flow is a key concept in lean, and also critical to achieving Operational
Excellence. While there is a lot of information available on creating flow
in cells, and even in the end-to-end value stream within a factory, there
is one large obstacle many companies face when trying to design and
achieve flow in their operation: product mix.

Product mix can have a staggering effect on flow. While products may
be grouped and shipped to the same customer on the same order, those
same products grouped together raise havoc in manufacturing since
they may have different cycle times, process steps, materials, and more.
To address these variations, companies apply the concepts of mixed
model production.

Pistons 45-80 units Collars 30-55 units Lifters 55-75 units


Brackets 50-90 units Diverters 35-60 units Levers 60-95 units
Arms 25-75 units Detectors 20-40 units Cranks 40-70 units

Stamp Paint Assemble Ship


FIFO
Max 50

Mixed model production means producing a variety or mix of products


or product variations through the same value stream at the pull of the
customer. In short, it means an operation can build to demand even when
the variety of products needed have different cycle times and processing
steps, providing a significant business advantage.

The key to establishing mixed model flow is to create value stream maps,
both current state and future state. Although value stream mapping is
commonly taught to identify waste or develop an ideal state, that approach
does not lead to a designed flow for mixed model, nor will it achieve
Operational Excellence.

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 3
Value Stream Mapping
Method for Mixed Model Flow
The value stream mapping method used to create mixed model flow is:

\\ Identify product families. Product families are groups of products


that have similar process flow and similar work content. Similar means
having about 80% the same process flow and less than 30% variation
in work content.

\\ Map the current state. The next step is to make a current state
map for each value stream. The current state map identifies “what
we do and how we do it” and is done by walking the shop floor
and recording what is observed. It captures data such as inventory
quantities, cycle times, customer demand, and more.

\\ Ask 10 questions for mixed model flow. Once an organization


creates the current state map, it applies a series of guidelines—not
brainstorming or anything else—to it to generate the future state map.

\\ Build the future state based on answers to the 10 questions.


The future state describes how the operation will perform once
the guidelines have been implemented on the shop floor so an
organization can determine if the implementation has been successful
by evaluating if the operation is performing as designed.

\\ Create an implementation plan for the future state. The final step is
to create an implementation plan by breaking the future state down
into implementation loops and establishing measurable goals.

Step three in this process—the 10 questions—is very important, and


also where most companies go wrong, so a more in-depth review of the
questions for mixed model production operations should ask during the
design phase is valuable.

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 4
10 Questions for Mixed Model Flow
In some operations, particularly those with steady customer demand
and low product variation, following basic guidelines are enough to
create robust value stream flow. However, in complex operations with
many custom product variations, shared resources, and uncertain
demand from day to day, operations need to design mixed model value
streams by following a specific set of guidelines. These guidelines,
from the book Creating Mixed Model Value Streams: Practical Lean
Techniques for Building to Demand, are:

1. Do you have the right product families? Group products together


based on similar processing steps and work content.
2. What is the takt time at the pacemaker? Determine the rate at
which the pacemaker must produce a part to keep pace with
customer demand for the product family.
3. Can the equipment support takt time? Determine if the existing
machine capacity is enough to support the product family within
the takt time.
4. What is the interval? Calculate how often the pacemaker will
produce all the parts in the product family.
5. What are the balance charts for the products? Create continuous
flow through the pacemaker by balancing work content, per operator,
to takt time.
6. How will you balance flow for the mix? Determine how to handle
variation within the product family at the pacemaker, either by
adjusting labor, scheduling, or work balancing.
7. How will you create standard work for the mix? Establish the one
best way to do a task, and then have everyone follow that method.
8. How will you create pitch at the pacemaker? Establish how often
you will release and take away work from the pacemaker. Pitch is a
visual management timeframe that lets you know if you are on time to
customer demand.
9. How will you schedule the mix at the pacemaker? Schedule
the pacemaker to handle variation within the product family, and
determine the mix the pacemaker can support.
10. How do you deal with changes in customer demand? Customer
demand will vary, and you need to have a “Plan B” pre-established
for when it does. “Plan B” can consist of pulling product from a
supermarket, rebalancing the pacemaker, and so on.

The first four questions are critical to a company’s efforts to successfully


creating mixed model value streams, so they require a closer look.

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 5
Four Questions In Depth for Value
Stream Success
Question 1: Do you have the right product families?

Establishing product families is not about looking for similarities in terms of


sales categories or commonality across customers. Instead, operations are
strictly looking at how the products are built and how long it takes to build
them. To begin the process of determining product families, an operation
creates a product family matrix. This is a grid where active part numbers are
listed down the left-hand side and the processing steps through which these
products go are listed across the top. Wherever a part number intersects with
a process through which it passes, a check mark is placed (Figure 1).

 PROCESS STEPS & EQUIPMENT 


 PRODUCTS 
Injection Mech. Electrical Final Configure
Stamp Welding
Mold Assembly Assembly Assembly & Test
Sensor Activated Arm       
Laser Activated Arm       
Manual Activated Arm       
Radon Detector   
XS2 Servo Motor   
XS4 Servo Motor    

Figure 1: A simple product family matrix.


Once this is done for all part numbers, the operation should sort them first
based on process commonality, ideally looking for 80% similarity. Once
the products have been initially sorted based on process commonality, the
operation replaces all the check marks with the processing time of each
part at each process. Within each family, it is important to ensure there is
no more than 30% variation in total work content; any more and creating
mixed model flow will be challenging due to the variation that could be
introduced to the line. If the variation within a family is greater than 30%,
an operation may have to remove parts to get the variation back within
the desired range. Sometimes, this won’t be possible, and the rest of the
design will have to be adjusted to compensate for this increased variation.

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 6
After an organization has completed these two steps, the final step is to
bring in some product experts who have extensive knowledge of the part
numbers manufactured to make sure the methodology has not yielded
any untenable manufacturing situations. For example, greasy gears should
not go down the same line as sensitive detection cells, even if the process
for determining product families indicated they could, since the risk of
contamination would be too high.

Each product family then becomes a value stream, and for each value
stream, each of the remaining nine questions for creating mixed model
flow need to be answered.

Question 2: What is the takt time at the pacemaker?

Takt time is a concept that has been around in lean for many years and
describes the rate at which a part must be produced in order to keep pace
with customer demand. To calculate take time, the following equation is used:

Effective Working Time


per Time Period

Customer Demand
per Time Period

However, it may be difficult to calculate takt time sometimes, such as when


demand is unstable or there are a high number of custom jobs. In cases like
these, an operation would use takt capability, which describes how much
can be produced in terms of volume and mix over a certain period of time.

Addressing the “mix” component of takt capability is critical. While an


operation may be able to produce 50 products in a day as long as those
products all have low cycle times, if it receives an order for 10 products
that are more complex and have high cycle times, there may only be
enough time left in the day to produce 20 products with low cycle times.
In this way, the mix is related to how much volume an operation can
produce, and vice versa.

It is usually necessary to create multiple takt capabilities, with each one


satisfying a different level of volume and mix. In that way, an operation
is able to switch between them quickly and effortlessly, like flipping a
light switch, and not disrupt the flow to the customer. However, this not
only means each takt capability has to be pre-established, but also any
associated standard work, cell rebalancing, product flow, and so on, must
be pre-established, too.

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 7
Question 3: Can the equipment support takt time?

Machine cycle times are typically fixed and difficult to change without
significant investment, so an important step in creating mixed model value
streams is to check whether the machines can support the anticipated
load. To determine this, an organization needs to know three things:

1. The effective working hours

2. The cycle time of each product at the machine

3. The demand for each product

To determine the equipment required, an operation divides the sum of the cycle
times multiplied by their respective demand by the effective working time:

Sum (C/T x Demand)


Equipment Required
Effective Working Time

Suppose there is one eight-hour shift (480 minutes) and the sum of the
the cycle times of the products is 420 minutes. The equipment required
would therefore be:

420 Minutes
Equipment Required .88 Machines
480 Minutes

With .88 machines required, an operation should be able to handle the


anticipated load. However, it may also need to consider other factors in
the load calculation, such as uptime. For example, suppose there have
been reliability issues at the machine and its uptime is only 90%. An
uptime of 90% means that 10% of the time, the machine is not available
for use, and this needs to be factored into the load calculations.
Before determining the equipment required, an operation would need
to recalculate the total time needed by dividing the sum of the cycle times
by the uptime:

420 Minutes
Total Time Needed 467 Minutes
.90 Uptime

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 8
Then take 467 minutes, not the original 420 minutes, and divide it by the
effective working time to determine the equipment required:

467 Minutes
Equipment Required .97 Machines
8 Steps to 480 Minutes

Operational With uptime factored in, the equipment required has increased from .88
Excellence machines to .97 machines. This means the machine can essentially never
miss a beat, or the operation will find itself behind. In reality, with such a
tight threshold, it is likely one machine would not be enough to meet the
1. Design lean value streams anticipated demand.
2. Make lean value However, rather than purchase another machine, an operation would
streams flow want to aggressively work to improve the uptime. Even a 5% improvement
to 95% would reduce the equipment required down to .92 machines.
3. Make flow visual This is still a bit high, but much more likely to succeed given there is now
some room for error. Increasing uptime in support of creating mixed model
4. Create standard work
value streams is where traditional continuous improvement techniques
for flow
like kaizens, Six Sigma, and more can be used.
5. Make abnormal flow visual
6. Create standard work for Question 4: What is the interval?
abnormal flow
Interval means how often an operation can cycle through all the parts in
7. Have employees in the flow the product family at the pacemaker. If it takes one week to produce all
improve the flow the parts in a given product family, then the interval would be one week.
If it takes one day to produce all the parts in the product family, then the
8. Perform offense activities interval would be one day. If it takes one shift, then the interval would be
one shift, and so on in smaller or larger intervals of time.

The length of the interval is tied to many other aspects of the operation:
lot size, lead time, and the time it takes to catch defects, to name just a
few. Consequently, lowering the interval as much as possible is critical
to the success of mixed model implementations by increasing flexibility
throughout the value stream, enabling the operation to become more
responsive to changing customer demand, and reducing the amount of
chaos in the system by limiting unpredictability.

For example, if the interval is one week, then it means that every time an
operation produces a part in the product family, it needs to produce one
week’s worth of parts. This is because it will be another week before the
operation gets back around to the part in question and it needs to make

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 9
sure there is enough of it on hand to support demand during that time
period. If the interval is only one day, then the lot size for each part
drops to one day.

Quality is directly impacted by interval, too. A one-week interval means


an organization will only be able to identify quality issues once per week,
since that is how frequently it produces each part. If a defect is found
Q&A using a one-week interval, then a week’s worth of parts would have to
be scrapped. However, if a defect is found using a one-day interval, then
only a day’s worth of parts would have to be scrapped.
Q. Does a lower interval
mean more changeovers? The lead time is also directly tied to the interval. With a one-week interval,
the lead time is also one week. However, if the interval could be reduced
A. Yes, but it doesn’t to one day, then the lead time would also only be one day, thereby
necessarily mean an presenting an instant competitive advantage.
operation will spend
the same total amount
of time on changeovers
as long as it reduces the
changeover time.
Mixed Model Value Streams
The key to lowering interval for Growth
is reducing the changeover
time. This can be done using By following the value stream mapping method outlined for mixed
well-known techniques, model flow, the end result is an operation capable of flowing a mix
including Single Minute of parts with varying cycle times and levels of customer demand
Exchange of Die (SMED), through the same value stream, all at the pull of the customer.
in which a rigorous process And that is the first step in the journey to Operational Excellence.
is followed with the goal of Once an operation designs lean value streams, it follows the
reducing changeover time. remaining steps in the process to achieve Operational Excellence and
By reducing the changeover create a self-healing, autonomous flow that frees managers from
time and thereby lowering having to fix flow problems and enables them to focus on business
the portion of total time taken growth. That means talking with product development, integrating
up by changeovers, more operations up front at the top of the innovation funnel, spending time
changeovers can be with sales and even visiting customers to discuss the development of
performed in support of a products operations can currently support to meet the customers’
needs. And that is exactly what operations need to do to improve their
lower interval.
market share and growth.

1130 Ten Rod Road Ste. A 202


North Kingstown, RI 02852

Tel: 401.667.0117 | Fax: 401.667.7298


info@instituteopex.org | www.instituteopex.org 10

You might also like