The History of Six Sigma: 1980s–2000s

Motorola Story

The origin of Six Sigma is found at Motorola. In a study of the field life of a product, a Motorola engineer observed that products that were defect‐free at manufacturing rarely failed in early use by a customer. However, products which had failed during production and undergone rework were more susceptible to early failures in the hands of the customer. It was as if those products that were originally made well worked well, while those that originally had a defect might have other defects as well. Catching and fixing one defect did not assure a good product.

It was not enough to find and fix defects, they had to be prevented.

Preventing defects is possible only when the processes that produce the product are set up so that defect can’t or don’t happen. Motorola focused on “how” the work was done in each and every process. If defects were prevented, then there would be no need for rework and quality would be higher. Since there were lots of processes, Motorola developed a measure for Six Sigma work that could be applied to every process.

The measure Motorola developed was called the sigma level or sigma value for each critical‐to‐quality (CTQ) associated with a process. A CTQ analysis is a way of studying the flowchart of a process to find problems. The analysis studies inputs and outputs and identifies the steps that influence quality of the process and its outputs. If a process is in an ideal state, it’s stable and predictable and meets the CTQs. If it’s stable and predictable but doesn’t meet the CTQs, it’s in a threshold state. If it’s free from special cause(s), it should be relatively straightforward to move it into an ideal state, perhaps through a DMAIC project (see Section 8.12.2.3).

The sigma level is based on a statistical analysis linking the defects per million opportunities (DPMO) and the capability of the CTO with respect to customer requirements. A CTQ with 3.4 DPMO is associated with capability value of C_p = 2.0 and C_pk = 1.5 and is considered to have achieved the Six Sigma level. Motorola successfully applied the concepts, philosophy, and techniques of Six Sigma to the design, developments, and production of the Bandit pager. The quality level of this pager was unsurpassed. Motorola showed that the traditional approach of detecting and fixing defects resulted in CTQs at four sigma level of quality. Six Sigma quality, or 3.4 DPMO, led to eliminate of costly inspection and rework, which in turn led to decrease in manufacturing time and increase in customer satisfaction. Customers were happy and Motorola reaped staggering financial savings.

General Electric Story

In 1995, Larry Bossidy of AlliedSignal introduced General Electric (GE) top management to Six Sigma. Bossidy’s account of the benefits realized at AlliedSignal led Jack Welch, then CEO of GE, to launch a Six Sigma initiative which would bring all of GE’s products to a Six Sigma level by 2000. At the time, GE’s processes were operating between three and four sigma, which would mean an average of 35 000 DPMO. To go from three or four sigma to Six Sigma would require a monumental training and education effort. GE rolled out a rigorous training in the Six Sigma methodology with a heavy emphasis on statistical methods. From 1996 through 1998, GE’s training investment approached $1 billion and the returns on that investment were already keeping pace. By 1999 they expected $1.5 billion in savings. GE’s businesses range from a variety of products to various services including GE capital. Throughout GE, Six Sigma initiative returned benefits to the organization’s bottom line and its ability to serve its customers.

DMAIC Model

The five phases of the DMAIC model are define, measure, analyze, improve, and control (Figure 8.18). While this model suggests a linear progression with each phase leading to the next, there will always be some iterative work between the phases. Each phase involves a certain amount of work to accomplish which is integral to the successful completion of a Six Sigma project (Ward and Poling 2004).

The Phases of DMAIC

The define phase is for focusing the Six Sigma project and creating a roadmap with timeline to guide all project activities toward a successful conclusion. Under the champion leadership, the project team works on mapping the process associated with the project and setting up communication with all stakeholders of the process. Logistical decisions about team members and meeting times are decided. The charter is finalized and signed off by everyone. If questions arise about availability of resources such as time or people, the champion helps resolves these issues.

The measure phase concentrates on collecting data to assess current process performance. It involves deciding all of the measures that are pertinent to the project. These include the scorecard metrics that tie to the scoreboard of the organization as well as all CTQ characteristics for the process. All measure must be clearly identified and data collection plans for each one setup. Then data can be collected regularly for the analyze phase.

In the analyze phase, the various techniques are used to determine how the current process performance compares to the performance goals in the project charter. Data are analyzed with the express purpose of understanding why the current process works the way it does. Specifically, it is important to determine if the process has a predictable or unpredictable behavior. Predictable processes have routine sources of variation that are always present and impact the process day in and day out. Unpredictable processes have both routine sources of variation plus special, unpredictable sources of variation that can knock a process off track.

Activities in the improve phase are directed toward finding out how to improve the process. This phase involves investigations to see what changes to the process will make the process better to meet the project goals. The project team may conduct experiments to see what can be achieved and the associated costs. Once several solutions are identified, the project team can evaluate and pilot them for their potential benefits. Before a final improvement solution is selected and implemented, the project team needs to assess any potential problems with all solutions. Resolution of all potential problems will assure that the solution will not be undermined in the future.

Finally, the control phase in the DMAIC model must focus on making the changes permanent for sustained process improvement. This means that the project team will work on those activities that are critical for the project team to turn over an improved process to the process owners. To reap the benefits of any Six Sigma project, the improvements must be sustainable by the organization, not by the project team. Elements included in the control phase are training of employees, control plans, and process measures that will keep attention on the process. The transition from the control phase of the DMAIC model used by the project team back to the organization comes with the standardized and institutionalized phases.