BUS 4543 Quality Management Tools Learning Outcome Three Optimising and Controlling Processes Through Statistical Process Control

Statistical Process Control Defined Rationale for SPC Optimizing and Controlling Processes through Statistical Process Control (SPC) MAJOR TOPICS Statistical Process Control Defined Rationale for SPC Control Chart Development Management’s Role in SPC Role of the Total Quality Tools Authority over Processes and Production Implementation and Deployment of SPC Inhibitors of SPC

Optimizing and Controlling Processes through Statistical Process Control (SPC) Definition: SPC is a statistical method of separating special-cause variation from natural variation to eliminate the special causes and establish and maintain consistency in the process, enabling process improvement The application of statistical methods to the measurement and analysis of variation in a process. This technique applies to both in- process parameters and end-of-process (product) parameters. The origin of SPC was in the work of Dr. Walter Shewhart at Bell Laboratories 1931. Although SPC was ignored in the West after World War II, Japan adopted and subsequently developed it into total quality. The rationale (main objectives) for SPC includes the following: Enables the control of process variation. Makes possible continual improvement of the process. Results in predictability of processes. Results in elimination of waste. Makes less expensive inspection modes possible. SPC is essential today to elevate the quality of products and services while lowering costs in order to compete successfully in world markets.

Optimizing and Controlling Processes through Statistical Process Control (SPC) Control Chart Development: There are several types of control charts, the choice of which being determined by the kind of process under consideration. Further, some control charts are designed for variables data (something measured), others are concerned with attributes data (something that can be counted). Different procedures are used for developing these two types of control. Both require Upper Control Limits (UCL) and Lower Control Limits (LCL) and a Process Average. Upper and Lower Control Limits and Process Average calculations for constructing the control chart are made from the actual process data, which must be of sufficient quantity, and taken over a relatively short period. After drawing the blank control chart with UCL, LCL and process average, the data from which the calculations were made are plotted on the chart. No data points can penetrate UCL or LCL, and there must be no long runs of data on one side of the process average. That will only be true if the process is free of special causes of variation. If that is the case, the chart is ready for use.

ADVANTAGES OF DECREASING PROCESS VARIABILITY Reducing the variation in a process leads to some great benefits: Lower variability may result in improved product performance that is discernible by the customer. Lower variability of a component characteristic may be the only way to compensate for high variability in other components and thereby meet performance requirements in an assembly or system. For some characteristics such as weight, lower variability may provide the opportunity to change the process average. Reducing the standard deviation of fill content in a food package permits a reduction in the average fill, thereby resulting in cost reduction. Lower variability results in less need for inspection. In the extreme, if there were no variability, inspection of only one unit of product would tell the whole story. Lower variability may command a premium price for a product. Some electronic components have traditionally been priced as a function of the amount of variability. Lower variability may be a competitive factor in determining market share.

STATISTICAL CONTROL CHARTS—GENERAL Control chart, compares process performance data to computed “statistical control limits,” drawn as limit lines on the chart. The process performance data usually consist of groups of measurements (rational subgroups) from the regular sequence of production while preserving the order of the data. A prime objective of a control chart is detecting special (or assignable) causes of variation in a process—by analyzing data from both the past and the future. Process variations have two kinds of causes: (1) common (or random or chance), which are inherent in the process as currently designed, and (2) special (or assignable), which arise new and create abnormal variation. A process that is operating without special causes of variation is said to be “in a state of statistical control” because the variation is predictable and consistent with historical performance. The control chart distinguishes between common and special causes of variation through the choice of control limits. When the variation exceeds the statistical control limits, it is a signal that special causes have entered the process and the process should be investigated to identify these causes of excessive variation.

ADVANTAGES OF STATISTICAL CONTROL The process is stable, which makes it possible to predict its behavior, at least in the near term. A process in statistical control operates with less variability than a process having special causes. Lower variability has become an important tool of competition. A process is in statistical control is helpful to the workers running a process. A process is in statistical control provides direction to those who are trying to make a long-term reduction in process variability. A stable process that is statistical control also meets product specifications and therefore provides evidence that the process has conditions that, if maintained, will result in an acceptable product. A process having special causes is unstable, and the excessive variation may hide the effect of changes introduced to achieve improvement.

C Control Chart Formulas

P Control Chart Formulas

P Control Chart Formulas d2 = 2.326

Elimination of waste is another key element of SPC. Optimizing and Controlling Processes through Statistical Process Control (SPC)- the main elements of SPC Continual improvement of processes requires that special causes be eliminated first. Process improvement narrows the shape of the process’s bell curve, resulting in less variation. Continual improvement is a key element of SPC and total quality. SPC enhances the predictability of processes and whole plants. Elimination of waste is another key element of SPC. SPC can help improve product quality while reducing product cost.

SPC makes sampling inspection more reliable. Optimizing and Controlling Processes through Statistical Process Control (SPC) SPC makes sampling inspection more reliable. SPC supports process auditing as a substitute for more expensive inspection. SPC requires a capability in statistics, either in-house or through a consultant. Process operators should be key players in any SPC program. Understanding the process is a prerequisite to SPC implementation. All employees involved in SPC must be trained for their involvement. Measurement repeatability and reproducibility is essential for SPC.

SPC and the operator must have process-stop authority Optimizing and Controlling Processes through Statistical Process Control (SPC) Management’s role in SPC is similar to its role in total quality overall: commitment, providing training, and involvement. The seven tools, supported by flowcharting, five-s, FMEA and DOE are required for SPC. SPC and the operator must have process-stop authority SPC implementation must be carried out in an orderly, well thought- out sequence. SPC requires collaborative team activity.

Implementation and Deployment of SPC Requires commitment and time of management and other key personnel. Requires some expertise in statistics. Must be done in a well planned, orderly process. (Roadmap on page 329) Inhibitors of SPC Lacking statistics expertise. May have to bring in outside help. Assigning SPC responsibility to the wrong person/group. The process operator should “own” SPC on his process. Failing to understand how the process really works. Imperative that the process be accurately flowcharted first. Trying to implement SPC while the process still has special cause variation. Process needs to be cleaned up as much as possible before trying to make control charts. Inadequate training and lack of discipline in process operation. Users need training, and process procedures must be followed. Measurement repeatability and reproducibility lacking. Instrumentation and procedures must be made repeatable and reproducible. Otherwise data is not reliable.

Preparation, planning and execution are the three main phases of implementing statistical process control in an organization SPC implementation Phase RESPONSIBILITY ACTION   P R E A T I O N Top Management Consultant or in-house expert Commit to SPC Organise SPC committee Train SPC Committee L G SPC Committee assisted by consultant QA Management Set SPC objectives Identify target processes Train appropriate operators Ensure repeatability and reproduce ability of instruments and methods Delegate responsibility to operators in key role X C U SPC Committee, operator, suppliers and customers Operator W/ Expert assistance Consultant or inhouse expert operator SPC committee and management Operator with assistance as required Flow chart the process Eliminate special cause variances Develop control charts Collect and plot SPC data; monitor Determine process capability Respond to trends and out of limits data Track SPC data Eliminate root causes or any special issues in variation Continually improve the process

The steps the top management should take to implement statistical process control 1. Preparation Phase Step 1. Management must inform staff of its commitments to the entire activities of SPC. Communicate the need to maintain a high level of quality in output and reduce complaints.  Step 2. Form a SPC team comprising cross functional and process owners. Each must be trained in quality awareness and basics of quality management. This will put everyone in the same page when communicating about quality. Step 3. Train the members in basic quality awareness and also on the fundamental objectives of The SPC.SPC team must undergo knowledge of basic statistics in order to operate the charts and on the fundamental objectives of SPC. Educate the members on the benefits of SPC in order to have a better buy in from members. 2. Planning Phase Step 4: Set SPC objectives. This will involve the measurable objectives namely the elimination of gaps – existing problems and what are desired. The objectives could be reducing errors, rejects of entire products. Step 5 Identify target processes. Identify from the processes (name different processes in the case); how each process should be under control using SPC methodology. Step 6 Train process owners, operators and other members in the team. Train the team on basic statistics and the basic charts using the varied formulae relevant to the set objectives. Step 7: ensure all machines for acceptable performance. This will ensure that data plotted on control charts are reliable. Preventive maintenance of machines must be conducted periodically. Step 8: delegate responsibility for operators, process owners, and others to play a key role. Assign leadership roles among each member, resource allocation, data collection and maintenance of data.