24 Quality Measurement Statistical Process Control

Sudhanshu Joshi

1. Statistical Process Control (SPC)

Statistical Process Control (SPC) as a set of quality improvement methods developed in industrial settings and more recently generalized for medical processes. Attention is directed to the control chart, SPC’s primary tool used to evaluate process variation. The control chart’s logical and statistical conditions are reviewed.

SPC encompasses a set of improvement methods used to “understand, monitor, and improve process performance over time” (Woodall, 2000). It incorporates statistical, analytic, and managerial processes that: (1) organize team support for the improvement process, (2) model the specific production process, (3) specify a quality measurement strategy, (4) monitor process variation using control charts, (5) identify special causes associated with non-random time series, and (6) control production through the elimination of special causes. An extensive body of literature documents the successful application of SPC to improve different types of industrial processes in different countries, spanning seven decades (Walton, 1986). SPC tools are “user- friendly” and can be managed on the production line. One expert writes, “This is not a theory; this has been proven time after time” (Wheeler, 2001).

However, questions remain about the robust nature, operation, and appropriate analysis of control charts (CC), perhaps the most prominent of SPC’s statistical tools. Control charts enable visual and statistical analysis of time series data. As a descriptive tool, the CC enables the practitioner to visually monitor patterns in the production process. As an inferential tool, the control chart establishes baseline, and, as additional data are added during subsequent observations, probability tests are applied to distinguish special from common cause. When the data series has been shown to be stable or controlled, prediction intervals can be calculated. The control chart uses historical observations to infer to future production outcomes.

2. Evolvement of SPC

The two people recognized for having developed and promoted SPC worldwide are Walter Shewhart and W. Edwards Deming (Shewhart, 1931 and 1939, Deming 1950 and 1982). Shewhart’s seminal work on SPC was published in 1931, although his work at Bell Telephone Laboratories introduced the control chart concept in 1924 (Montgomery, 2001). Shewhart developed the fundamental logic of SPC, advanced measurement practices, and developed control chart methods. Deming contributed in the areas of sampling theory and research design; however, he is also recognized for incorporating management theory and promoting worldwide dissemination of SPC through his work to re-build industry in post-war Japan. Deming is frequently cited for his “fourteen points for management which include the elimination of mass inspection and work standards that prescribe quotas and numerical goals (Wheeler, 1995). Deming considered improvement a fundamental personal and corporate value that requires “constancy of purpose” at all times, and in all aspects of production. He assigned responsibility for product quality to corporate management, estimating that 80 – 85% of all product-related quality problems involve production systems (DeVor, 1992). Deming is also recognized for placing emphasis on consumer perspective in the determination of quality. He considered the consumer an essential part of the production line, emphasizing that products should respond to consumer needs, present and future (Deming, 1982).

SPC has conceptual roots in probability theory. Shewhart and Deming, however, distinguished SPC from traditional statistics in their discussions of analytic and enumerative research (Deming, 1950). Analytic studies use practical methods that respond to environmental realities apparent in many industrial processes; they infer to a future production process for the purpose of assessing process capability (Ramberg, 2001). The Analytic Study has no fixed, static inferential population; rather, it infers to a future process. In contrast, Enumerative Studies have a well-defined population that provides an inferential target using traditional sampling methods. Enumerative Studies represent pure scientific research carried out under controlled conditions. The Analytic/Enumerative dichotomy distinguishes SPC from other quantitative methods and provides a framework for understanding its application (i.e. using smaller samples, inferring to future process populations, and preferences for managerial pragmatism over statistical purism)

3. The Quality Construct

Quality has been defined in the engineering sense as “fitness for use,” referring to the product’s ability to achieve its intended purpose (Montgomery, 2001). The language derivation is from the Latin word qualitas, means what kind or how constituted. Quality has also been defined as (1) Any of the features that make something what it is, or (2) The degree of excellence that a thing possesses (New World Dictionary, 2″ Edition). The first That quahty can be evaluated from both objective and subjective perspectives is important. Each perspective requires a different measurement strategy. Objective quality can be reliably measured independent of time and, in the abstract, has a constant value (Shewhart, 1939). Subjective quality is more difficult to measure given differences in perceived value based on individual proclivity. Objective measurement is more concrete and directly observable, while subjective measurement relies on indirect methods of observation. An example of measuring objective quality would include counting the number of scratches on a paint surface or reporting consumer queue time. An example of indirect measurement of subjective quality includes consumer/patient satisfaction. Objective and subjective quality measurement require different approaches, although the same product or service may be evaluated using both. Subjective quality represents a construct, or abstract variable. Constructs vary in complexity and may have “fuzzy edges,” making them difficult to measure. Constructs also vary by the number of observable variables related to their domain and how well defined the indicator variables are (Nunnally, 1978). Subjective constructs tend to have larger domains and be associated with variables having different degrees of association with the underlying domain. This means that subjective quality will likely be measured less reliably and with greater error, challenging construct validity.

4. Animation: Understanding Statistical Process Control (U-tube)

https://www.youtube.com/watch?v=i-lPys7pVlY

5. Various Functions of Statistical Process Control

5.1 Objective analysis of variation

SPC must be practiced in 2 phases: The first phase is the initial establishment of the process, and the second phase is the regular production use of the process. In the second phase, a decision of the period to be examined must be made, depending upon the change in 4 – M conditions (Man, Machine, Material, Method) and wear rate of parts used in the manufacturing process (machine parts, jigs, and fixture)

5.2 Emphasis on early detection

An advantage of SPC over other methods of quality control, such as “inspection”, is that it emphasizes early detection and prevention of problems, rather than the correction of problems after they have occurred.

5.3. Increasing rate of production

In addition to reducing waste, SPC can lead to a reduction in the time required to produce the product. SPC makes it less likely the finished product will need to be reworked.

6. Limitations

SPC is applied to reduce or eliminate process waste. This, in turn, eliminates the need for the process step of post-manufacture inspection. The success of SPC relies not only on the skill with which it is applied, but also on how suitable or amenable the process is to SPC. In some cases, it may be difficult to judge when the application of SPC is appropriate.

7. Applications

In manufacturing, quality is defined as conformance to specification. However, no two products or characteristics are ever exactly the same, because any process contains many sources of variability. In mass-manufacturing, traditionally, the quality of a finished article is ensured by post-manufacturing inspection of the product. Each article (or a sample of articles from a production lot) may be accepted or rejected according to how well it meets its design specifications. In contrast, SPC uses statistical tools to observe the performance of the production process in order to detect significant variations before they result in the production of a sub-standard article. Any source of variation at any point of time in a process will fall into one of two classes.

1) “Common Causes” – sometimes referred to as non-assignable, normal sources of variation. It refers to many sources of variation that consistently acts on process. These types of causes produce a stable and repeatable distribution over time.

2) “Special Causes” – sometimes referred to as assignable sources of variation. It refers to any factor causing variation that affects only some of the process output. They are often intermittent and unpredictable.

Most processes have many sources of variation; most of them are minor and may be ignored. If the dominant sources of variation are identified, however, resources for change can be focused on them. If the dominant assignable sources of variation are detected, potentially they can be identified and removed. Once removed, the process is said to be “stable”. When a process is stable, its variation should remain within a known set of limits. That is, at least, until another assignable source of variation occurs.

7.1. Applications to Non-Manufacturing Processes

In 1988, the Software Engineering Institute suggested that SPC could be applied to non-manufacturing processes, such as software engineering processes, in the Capability Maturity Model (CMM). The Level 4 and Level 5 practices of the Capability Maturity Model Integration (CMMI) use this concept.

The notion that SPC is a useful tool when applied to non-repetitive, knowledge- intensive processes such as research and development or systems engineering has encountered skepticism and remains controversial.

In 2014 a method for data validation of measurement data, based on SPC, was tried out. The method enabled the user to validate data containing static wave components (process noise), a requirement when working on hydro power plants where slowly damping surges are abundant during normal operation

The application of SPC involves three main phases of activity:

Understanding the process and the specification limits.
Eliminating assignable (special) sources of variation, so that the process is stable.
Monitoring the ongoing production process, assisted by the use of control charts, to detect significant changes of mean or variation.

Suggested Readings:

(a) Gaudreau (1994): Total Quality Management for Custodial Operations: A Guide to understanding and Applying the key elements of Total Quality Management. CRC Press.

(b) Elearn (2016): Quality and Operations Management. Routledge.

(d) Liker and Ross (2016): The Toyota Way to Service Excellence: Lean Transformation in Service. McGraw-Hill Education.

e) Noori (1994): Production and Operations Management: Total Quality and Responsiveness. McGraw-Hill.

(f) Besterfield (2011): Total Quality Management. Pearson.