Design for Six Sigma (DFSS, DMADV)

When seeking to improve an EXISTING process, we apply the DMAIC methodology. But what do you do when the process doesn’t exist–like you’re starting from scratch? That is where Design for Six Sigma (DFSS) takes place.

Design for Six Sigma (DFSS) helps organizations create NEW products, services, and processes in a way that ensures customer satisfaction by using a structured framework (DMADV, IDDOV, or DMADOV).

What is Design for Six Sigma (DFSS)

Use DFSS when the business is seeking to build a new product, service, or process. Design for Six Sigma is effective in that it takes all the guesswork out of designing new goods for customers. Designing new goods can be daunting since no feedback exists, as commonly found in DMAIC. Oftentimes this leads to companies spending LOTS of money designing something in trial and error to build the best characteristics for this new good. But with DFSS, customer needs are built into the product’s design up front.

When Should You Use DFSS Over DMAIC?

Use Design for Six Sigma when:

• The business outcome is showing ‘stair-shaped graphs.’
• Building new products, services, and processes.
• Exponential change to an existing system.
• Introduction of new technology to the business.

DFSS vs. DMAIC Example

My first solo Six Sigma project required the use of DFSS over DMAIC, but I didn’t realize that at the time. You see, my project involved introducing automatized vending machines to distribute personal protection equipment (PPE) to manufacturing employees. In this case, I was introducing new technology to the business. Our existing process had a centralized store area to distribute PPE to each department twice daily, so I immediately assumed to use DMAIC to modify our current process. However, the introduction of vending machines completely changed the existing process.

Complications in my project occurred during the ANALYZE phase of DMAIC because the analysis of moving from a centralized PPE location to a decentralized system didn’t require any descriptive or inferential statistics to prove the new program’s benefits. I collected internal customer feedback months before the vending machines arrived to capture what employees needed from our new PPE program. These surveys proved to be very beneficial in understanding the needs of our employees. Overall, it would have been much more effective for me to have taken internal customer feedback and designed features into that vending program (as found in DESIGN of DMADV). It would have also cleared up vending software expectations from the suppliers participating in our bid process.

How to Implement DFSS

Similar to DMAIC, the Six Sigma team must be cross-functional to fully grasp the scope of the project. It is easy to assume that most DFSSs are composed of engineers and QA personnel since the focus of DFSS is on design. However, like playing with dominos, introducing one new product, process, or service can affect the sum of all parts.

The steps taken in DFSS vary depending on the methodology used. The most common method is DMADV (DEFINE, MEASURE, ANALYZE, DESIGN, VERIFY). Because DMADV is the most common and the other methods follow a similar pattern to DMADV, most of this article’s content will focus on DMADV.

DMADV (Define, Measure, Analyze, Design, Verify)

DMADV stands for DEFINE, MEASURE, ANALYZE, DESIGN, and VERIFY. It may require a large scope and budget in order to incorporate a large amount of change. This kind of event could be triggered by new research or technology breakthroughs, a mandate to leapfrog the competition, radical new customer requirements, new regulations, or legislation.

Resources are typically allocated earlier than regular product designs. Finally, success depends on the involvement of senior management, the team retaining a customer-centric focus, fostering creative design, and executing sound project management.

Similarities and Differences Between DMAIC & DMADV

DEFINE

Practically the same as DMAIC, with slight changes in tools emphasized. You want to understand the goals and customer requirements. Thus, the project goals remain customer-focused. Like DMAIC, DMADV includes a project charter and project plan.

MEASURE

Like DMAIC, here you want to determine the needs and specifications for your new process. However, in DMAIC, MEASURE involves gathering information from an existing process. Since DFSS involves creating a new process or product, there is no process to pull existing data; things differ. You can still benchmark other areas, like your competition, especially where processes are unknown.

Therefore, the MEASURE phase is best defined in terms of customer needs and wants, then translating them into measurable design requirements.

A Kano Model is a great tool to measure and rank customer needs (must-haves) and wants (exciter) features.

Both DMAIC and DMADV emphasize the importance of surveying and interviewing customers and plugging survey results into a CTQ matrix.

Also, like DMAIC, DFSS focuses on Critical to Quality (CTQ) metrics, which include tools like House of Quality.

• benchmark opportunities

ANALYZE

Here you want to examine your options and determine creative solutions. DMADV utilizes tools like Quality Function Deployment (QFD) during the ANALYZE Phase, whereas QFD is often employed in the DEFINE phase of DMAIC. In the ANALYZE phase, the team draws up various concepts that are tailored to customer requirements. Comparison tools such as benchmarking measures show how well the concept meets customer requirements. The final design is chosen by the end of the phase.

DESIGN

Here, you’ll likely use the previous phases’ outputs to develop a product and process. You may even design experiments to verify design meets customer needs.

This is where 3D modeling or preliminary drawings occur. Once complete, the design is evaluated against tools such as FMEA and Design of Experiment (DOE) to ensure all CTQ metrics are met. Other design tools include:

• QFD Matrix
• FMEA Failure Mode Effects Analysis
• TRIZ
• Brainstorming
• Pugh
• Taguchi Robust Design
• DFX – Design for X

VERIFY

This is where validation testing of the design occurs. Pilot runs are employed to validate design success– can the process meet or exceed customer need? After the product or process is validated, a plan for full implementation, allowing metrics to continue to monitor gains. Detailed process maps, work instructions, flow charts, and checklists are created to ensure a fool-proof transition to the operations team. Finally, you deploy the new process and ensure controls are in place to ensure improvements.

Finally, the DFSS Team is closed, and team members take on new projects.

Other Design for Six Sigma Methodologies

DMADOV (Define, Measure, Analyze, Design, Optimize and Validate)

The only difference between DMADOV and DMADV is the O–for optimize. Optimize involves several product evaluations across multiple performance levels and redesigning if necessary. Further experimentation in DOE occurs here.

Optimize the Design

Optimize the design that produces the maximum output. Use the statistical and process capability methods to build parameters and tolerances. Furthermore, develop detailed designs and functional parameters that meet the customer’s needs and expectations.

Key Tools

• Process Capability: Process Capability Analysis tells us how well a process meets a set of specification limits based on a sample of data taken from a process. Also, it can be used to establish a baseline for the process and measure the future state performance of the process for comparison.
• DOE: The objective of the Design of Experiments (DOE) is to establish optimal process performance by finding the right settings for key process input variables. Design of Experiments is a way to intelligently form frameworks to decide which course of action you might take. This is helpful when you are trying to sort out what factors impact a process.
• Taguchi Robust Design: Robust design has been the key development in the design process in recent years. The robust approach in design is a key aspect as it produces reliable design during manufacturing and product use. The basic concept of robust design is that parameter control that makes the design strong enough that it does not cause failure due to random “noise.”

Other tools, such as Response Surface Methodology (RSM), a more efficient take on DOE utilizing partial factorial designs, and Evolutionary Operations (EVOP) may also be used.

IDDOV (Identity, Define, Develop/Design, Optimize, Verify/Validate)

IDDOV takes the existing DEFINE phase and divides it into two modules, with IDENTIFY focusing on translating customer needs into the design of the product while DEFINE ensures the product specifications meet the consumer’s needs. Essentially, IDDOV includes more upfront work to establish customer requirements and then translate them into the product design. IDDOV also includes the same elements of DESIGN and OPTIMIZE as found in DMADOV.

See more about IDDOV here.

Design for X (DFX)

Also known as Design for Excellence. DFX captures the knowledge of experts as a guide for designers to create the actual product design. Design for X successfully captures the finite details of the design process in order to avoid manufacturing design and production problems down the road.

See more about DFX here.

Why Use Design for Six Sigma (DFSS)?

Business studies by Hockman (2001), Suh (1990), and Paul (1996) found that 70-80% of quality problems are centered on design. With 70-80% of quality problems coming from the design of the product or service, it becomes very difficult to take an existing process to perform at a 6 Sigma Level.

Suppose a company wants to achieve 3.4 defects per million opportunities. In that case, they need to use DFSS because no matter the gains made in DMAIC, a 6 Sigma performance capability requires near perfection. Remember that not all businesses need to perform at a 6 Sigma Level due to certain diminishing returns. For example, we can all agree that we want commercial airplanes (that cost millions of dollars to build) built to the highest quality possible. However, a company producing low pencils, a low-cost good, can be successful at a 4 Sigma Level with a DPMO of 6,210.

Design for Six Sigma (DFSS) Videos