DMADV: A Complete Guide to Using the DMADV Method

STAGE 2: MEASURE

There are three steps involved in the Measure stage of the project. These include:

1. Convert the requirements identified in the Define stage to CTQs
2. Prioritize the CTQs
3. Determine measurement methods for the CTQs

Let’s break down each of these steps:

Step 1: Convert requirements to CTQs

In the measure stage of the project, you should start by converting all of the requirements that you identified from the VOC, VOE, and VOB in the Define stage to critical to quality parameters (CTQs). 

Some examples of common CTQs are: 

For example, if the VOC revealed that the customer wants a product that will not break easily, one of the CTQs will be durability.

Step 2: Prioritize the CTQs

Once you have the CTQs defined, you can then prioritize them.

The KANO model is a good tool to use to gain a deeper understanding of the customer’s requirements and to prioritize CTQs. The KANO model looks at each of the customer’s needs and analyzes how satisfied or unsatisfied a customer will be if the need is present or absent. 

The model has two dimensions represented by x and y-axes. 

• The x-axis is the supplier’s fulfillment of a particular need which ranges from “supplier did not fulfill at all” to “supplier fulfilled completely”. 
• The y-axis is the customer’s satisfaction which ranges from “completely dissatisfied” to “completely satisfied”.

This model can be used to prioritize customer needs by ranking each within the following three categories:

1. Must-haves
2. Satisfiers
3. Delighters

Must-haves are needs that MUST be fully satisfied for the customer to even consider purchasing the product.

Satisfiers are needs that the customer expects to see in the product. These needs are essentially must-haves if you want to be competitive in the market.

Delighters are properties or features that would “wow” a customer. These allow you to become a market leader.

Step 3: Determine measurement methods for the CTQs

Now that you have defined and prioritized the CTQs, you should determine how you will measure their performance.

At this point, you won’t have anything to measure because you have not designed the process yet, but it is important to understand how you will gauge if you are achieving your CTQs.

Your measurement system will differ depending on what you are trying to measure (for example, you may want to use a thickness gauge if thickness is a CTQ for your product), but it is critical that the instrument you choose is both accurate and precise.

This can be achieved using a Gage Repeatability and Reproducibility (Gage R&R) test. The goal of a Gage R&R test is to determine how much of the variation present in your data can be attributed to the measurement system.

If the Gage R&R determines that your measurement system is accurate and precise, you can move forward with using it.

Also, as a part of this step, you should be sure to fully define your measurement process. You can do so by asking yourself the following questions:

• What am I measuring?
• How am I measuring it (what measurement system and procedure am I using)?
• When during the process should I measure this?
• How often should I measure it?
• Who is responsible for measuring this and communicating if the measurement data does not match our expectations for performance?

Depending on the business and your goals, the questions you need to consider may differ from this list, but it is critical to go through this exercise to ensure you are not leaving any gaps in your process that will result in future failures.

STAGE 3: ANALYZE

The next stage is Analyze which consists of five steps. These steps include:

1. Convert the CTQs to process/product features
2. Prioritize the features 
3. Determine performance targets for each CTQ
4. Develop possible design concepts
5. Evaluate each design concept

Let’s dig deeper into each of these steps:

Step 1: Convert the CTQs to process/product features

The first step of the Analyze stage is very similar in concept to what you did in the Measure phase. In the Measure phase, you converted requirements from the VOC, VOE, and VOB to CTQs and now, in the Analyze phase, you will convert those CTQs to features.

The most important thing to remember here is that one CTQ can lead to multiple features.

For example, if you are designing a car, one of your CTQs will definitely be safety. This CTQ translates to many features, such as seatbelts, air bags, antilock brakes, and much more.

Step 2: Prioritize the features

The next step is also similar to the Measure phase. In the Measure phase, you prioritized the CTQs. Now in the Analyze phase, you must prioritize the features.

There will be some features that are non-negotiable. This should include any features related to safety even if they weren’t explicitly stated in the VOC, VOE, or VOB.

A tool that you can use to help with this prioritization is QFD (Quality Function Deployment or House of Quality). QFD shows the relationship between the customer’s wants and the product features and helps you to rate the importance level of each feature.

Step 3: Determine performance targets for each CTQ

In this next step, you will determine targets for each CTQ based on the customer requirements as well as competitive products. 

In general, your targets must at least meet the customer requirements and you should strive to exceed the targets set by your competitors in order to dominate the market.

Step 4: Develop possible design concepts

The fourth step in the Analyze phase is when you begin actually creating design concepts. One useful tool you can use to do this is called TRIZ which is a Russian acronym for the Theory of Inventive Problem Solving.

The idea behind TRIZ is that every problem has previously been solved by someone, somewhere. Therefore, you can save yourself a lot of time by generalizing your specific problem to something that has already been solved rather than reinventing the wheel.

There is a TRIZ effects databases that is a compilation of resources that answers a wide variety of different queries.

By generalizing your specific problem to something that has already been solved by someone else, you can spend less time trying to uncover the basics of solving your problem and can instead get a jumpstart on determining exactly what your specific solution might look like.

Step 5: Evaluate each design concept

Once you have some design solutions established, you will then need to evaluate each of these designs and land on a final decision. Some useful tools for this are the Pugh Matrix and the Six Thinking Hats.

The Pugh Matrix

The Pugh Matrix is a quantitative ranking system that allows you to compare alternative solutions to your baseline in order to determine which option best achieves your desired state. You can complete this using the following steps:

1. Determine what the most important characteristics are for your solution. You can use the VOC, VOE, VOB, and QFD to do this.
2. Weight each of these characteristics according to their relative importance.
3. Rate how well your current process/business achieves these characteristics. You can do this using a numeric scale such as 1-5 where 1 is the worst and 5 is the best. This will be your baseline.
4. Rate each of your alternative solutions based on how well they achieve each characteristic as compared to your baseline.
5. Calculate the sum of the ratings for each possible solution and choose the best.
6. Consider creating a hybrid design if it makes sense to do so.

The Six Thinking Hats

The Six Thinking Hats are a decision-making tool that forces you to consider all angles of a design. This will allow you to determine all the pros and cons that exist for the specific product. The six different aspects that you will need to consider are:

STAGE 4: DESIGN

After you select the most optimal design in the Analyze phase, you will then need to create a prototype of this design and test it. Therefore, the steps involved in the design phase include:

1. Create a prototype
2. Complete pilot runs
3. Revise the design as needed
4. Complete a risk assessment
5. Create an implementation plan

Overall, these steps are relatively straightforward to explain, but this will often be the stage that takes the longest to complete. Completing each of these steps thoroughly and accurately will ensure your success when you implement the final design.

Step 1: Create a prototype

The first step of the Design stage is to create a prototype for the design you selected during the Analyze phase. You will then be able to test this design through pilot runs or simulations.

Step 2: Complete pilot runs

Therefore, the next step is to complete these pilot runs. During these runs, you will want to complete the following:

1. Gain a deep understanding of how every input and process setting affects your outputs 
2. Test whether or not your design is able to achieve the target CTQs 

To do the first step, you should complete a design of experiments (DOE).

DOE will allow you to understand the relationships between different inputs, process settings, and outputs in your process. This is a critical piece if your process has many variables.

To complete the DOE, you should test different combinations of the inputs and process settings then analyze what effect they have on the outputs.

For the second step, you should use the measurement systems you chose and evaluated during the Measure stage to see if the design is achieving the target CTQs. If it is not, you will need to revise the design.

This is where the next step comes in.

Step 3: Revise the design as needed

This next step is where you will take your findings from the pilot runs and revise the design, if needed, to ensure you are achieving the target CTQ performance.

This will often be an iterative process. For example, once you revise your design you will need to redo steps 1-3 of this stage again and again until you have a final design that is fully functional and capable of achieving your target performance. 

Step 4: Complete a risk assessment

Once you have your final design defined, you need to complete a risk assessment for it. The most common form of risk assessment is a Failure Mode & Effects Analysis (FMEA).

The FMEA should be completed by a cross-functional team so that all perspectives are captured.

The goal of the FMEA is to identify anything that could potentially go wrong in the new process. These are referred to as your failure modes. Once you have identified these failure modes, you should go through them one-by-one and determine everything that can possibly happen if the failure were to occur. 

This can be a very tedious process, but it is critical to ensuring you have a risk-free process.

If you do identify any high risk failure modes, you should alter your design to get rid of them.

Essentially, this step forces you to “foolproof” your design.

Step 5: Create an implementation plan

The final step of the Design phase involves creating an implementation plan. This should clearly define who, what, when, and how you are going to complete the final go-live for the design. For example, this can include:

• Target implementation date
• Process owner establishment
• Training plan
• Change management process plan
• Plan to communicate the change to affected employees

STAGE 5: VERIFY

The final stage of the DMADV method is the Verify stage. During this stage, you will need to complete the following four steps:

1. Complete trials with the standard process setup
2. Make any necessary corrections to the design
3. Prepare for implementation
4. Go-live

Let’s take a deeper look at each of these steps:

Step 1: Complete trials with the standard process setup

During the Design phase, you completed pilot run trials which are typically completed under controlled conditions that may not mirror standard working conditions. If this is the case, you should be sure to complete trials with the standard process setup before officially implementing the new design.

This includes using standard raw materials, working conditions, and staffing.

As you do these trials, you should use check sheets to confirm that each CTQ is fully achieved. You should also note any issues that arise and complete a root cause analysis. This will allow you to resolve the source of any remaining issues.

Step 2: Make any necessary corrections to the design

Now that you have completed trials with the standard process setup and identified any issues, you should make any necessary corrections to the design to resolve those issues. 

You should add your findings from the root cause analysis to the FMEA and implement any necessary changes. This will give you the confidence that your design is fully operational and successful in achieving your goal.

Step 3: Prepare for implementation

The final step before implementation is to prepare. This is a critical step to ensure the go-live is smooth!

This includes actions such as:

• Writing standard operating procedures (SOP), work instructions, and/or standard work
• Updating any upstream or downstream processes to accommodate the new design
• Updating the process flow chart
• Training the employees
• Completing a change management process to communicate and track the change
• Creating a control plan

Step 4: Go-live

The final step is to officially go live with the new design!

At this point, the process will be handed over to the process owner and standard production will begin.

It is critical that you complete all of the previous steps thoroughly to ensure a smooth hand-off at this phase.

Once the process has gone live, you should continue to monitor it to confirm that the design is functioning as expected. You should always plan to look for potential areas for improvement to ensure you are maintaining a mentally of continuous improvement. This will allow you to stay on top of your competition.