BACK TO BASICS
To DMAIC or Not to DMAIC?
Identify when you need a structured method for problem solving
by Carl F. Berardinelli
Define, measure, analyze, improve and control (DMAIC) is a structured problem-solving method. Each phase builds on the previous one, with the goal of implementing long-term solutions to problems. Sometimes, project leaders or sponsors don’t feel a formal approach is necessary, but most problem-solving efforts benefit from a disciplined method.
The tools used in the define phase lay the foundation for the project. The team accurately and succinctly defines the problem, identifies customers and their requirements, and determines skills and areas that need representation on the project team. Individuals who must be part of the core team or be ad-hoc members are identified, and project measures, financials and a communication plan are established.
The measure phase is when the true process is identified and documented. Process steps, and corresponding inputs and outputs are identified. Measurement systems are identified or developed, and validated and improved as required. Baseline performance is established with trustworthy data.
In the analyze phase, the critical inputs are identified. Inputs that have a strong relationship with the outputs and root causes are determined. These critical inputs are the drivers of performance.
In the improve phase, potential solutions are identified and evaluated, and the process is optimized. The critical inputs that must be controlled to maintain performance that reliably satisfies the customer are determined. Process capability and project financials are estimated.
The control phase establishes mistake-proof, long-term measurement and reaction plans. The team develops standard operating procedures and establishes process capability. Project financials are updated, verified and reported. Control plans are established with reaction plans, ownership and control is transitioned to the process owner, and lessons are documented. The team documents opportunities to spread the outcomes to other areas in the organization.
When to use DMAIC
When improving a current process, if the problem is complex or the risks are high, DMAIC should be the go-to method. Its discipline discourages a team from skipping crucial steps and increases the chances of a successful project, making DMAIC a process most projects should follow.
If the risks are low and there is an obvious solution, some of the DMAIC steps could be skipped, but only if:
- Trustworthy data show this is the best solution for your problem.
- Possible unintended outcomes have been identified and mitigation plans have been developed.
- There is buy-in from the process owner.
If the obvious solution can’t be proven with trustworthy data, a DMAIC project should be launched.
There are two approaches to implementing DMAIC. The first is the team approach in which individuals who are skilled in the tools and method, such as quality or process improvement experts, lead a team. The team members work on the project part-time while caring for their everyday responsibilities. The quality or process improvement expert might be assigned to several projects. These are long-duration projects taking months to complete.
The second tactic involves the kaizen event method, an intense progression through the DMAIC process typically done in about a week. Prep work is completed by the quality or process improvement expert, and is centered on the define and measure phases. The rest of the phases are done by a team of individuals who have been pulled from their regular duties for the duration of the kaizen event.
In most cases, the changes are piloted during the event, and full-scale implementation is completed after the event. It is crucial the impact of these changes—whether they are wanted or not—are monitored. The advantage of this approach is the ability to make rapid change.
DMAIC is not an implementation method for best practices; it is a method to discover best practices. When developing a new process from scratch, the design for Six Sigma (DFSS) approach should be followed. DFSS builds on traditional DMAIC tools.
DMAIC is a data-driven, customer-focused, structured problem-solving framework. It builds on learning from previous phases to arrive at permanent solutions for difficult problems. Define will tell the team what to measure. Measure will tell the team what to analyze. Analyze will tell the team what to improve. And improve will tell the team what to control.
- Joseph M. Juran and Frank M. Gryna, Quality Planning and Analysis, McGraw Hill, 1993.
- Michael L. George, David Rowlands, Mark Price and John Maxey, Lean Six Sigma Pocket Toolbook, McGraw Hill, 2003.
Carl F. Berardinelli is a senior principal health systems engineering analyst at the Mayo Clinic in Rochester, MN. He has a bachelor’s degree in electronics technology from Indiana State University in Terre Haute. He is a member of ASQ.