Chapter 6: Statistical Process Control

Introduction

The origin of statistical process control (SPC) dates back to 1931 and Dr. Walter Shewhart’s book The Economic Control of Quality of Manufactured Product. It can be applied not only in manufacturing but also in any process carried out in the workplace. SPC is a statistical method of separating variations resulting from special causes, e. g. environmental and the Five M’s, from variations resulting from natural causes in order to eliminate the special causes. The Five M’s that affect a process are: manpower, machine, material, method, and measurements.

The requirements described in ISO 9001 are based on the quality management principles that are customer focus, leadership, engagement of people, process approach, improvement, evidence-based decision-making, and relationship management.

Process Approach

Process is defined by Cambridge Dictionary as “a series of actions that you take in order to achieve a result”.

A process has five main features:

1. Definability: Determination of the main elements/activities.
2. Measurability: It is the feature of the process to be monitored with performance criteria.
3. Repeatability: It is the ability of the output that is formed as a result of processing the inputs that activate the process to meet the customer needs and requirements continuously.
4. Controllability: It is the ability of the process managers who are always aware of the performance of the process and can take corrective actions when necessary.
5. Value added capability: It is an ability to have a positive effect on the quality of the output of the process and the satisfaction of the customer.

The processes are classified in three main groups:

• Main processes: Processes that start directly on demand from the external customers of the organization and provide a product or service to the external customer
• Management processes: Planning the activities of all processes in line with the organization’s objectives and regularly reviewing and reporting the performance indicators related to them,
• Support processes: Process that consists of different areas of expertise gathered under one roof in order to ensure the most appropriate use of resources throughout the company.

The benefits of management with processes:

• The processes, sub-processes and activities are defined as process map, procedures, and
• instructions. Therefore, every employee can work at the same format in an organization.
• Process performance indicators are defined, measured and reviewed periodically. The problems or barriers can be seen clearly. By eliminating these problems or barriers, the process can be improved according to its performance criteria.
• The relationships and interactions of the process are defined. By this way, the activities can be planned more accurately.

The SPC points are presented as “test” symbol. The bottleneck and non-value added processes are defined and eliminated according to lean philosophy. Ways to Eliminate Non-Value-Added Activities:

1. Rearrange the sequence of work steps and the physical location
2. Change work methods, forms, documents
3. Change the type of equipment used in the process
4. Improve operator training
5. Try to eliminate unnecessary steps or merge steps

Quality engineers measure all process activities and determine value added activity time and process cycle time. Also, process cycle efficiency is calculated by using the following equation:

Seven Basic Quality Tools for Process Improvement

For customer satisfaction, product or service should meet some specifications. In this way, the manufacturing process is capable of operating with little variability around the target or nominal dimensions of the product’s quality characteristics. Statistical process control (SPC) is a powerful collection of problem-solving tools useful in achieving process stability and improving capability through the reduction of variability.

The seven major quality tools for detecting and reducing the variability of process are:

1. Histogram
2. Check sheet
3. Pareto chart
4. Cause-and-effect diagram
5. Box plot
6. Scatter diagram
7. Control chart

Process Optimization

Quality of product improves by reducing the variation in process. The root causes of variation can be found by using some statistical techniques; confidence intervals, hypotheses testing. Design of experiments (DoE) is also a methodology to determine the best level of effective controllable factors on quality characteristics. The experiments are planned with a systematic procedure and carried out in a random order under controlled conditions. For the analyses of experiments, Variance analysis (ANOVA) method is used to test a hypothesis for discovering effective factors and factors’ interactions.

1. Experiments allow us to set up a direct comparison between the treatments of interest.
2. We can design experiments to minimize any bias in the comparison.
3. We can design experiments so that the error in the comparison is small.
4. Most important, we are in control of experiments, and having that control allows us to make stronger inferences about the nature of differences that we see in the experiment. Specifically, we may make inferences about causation.

Statistical Process Control

SPC is a methodology that collects data from process, analyze with several tools, trace and control of process for improving quality of product and service. Contributing to the widespread deployment of the SPC intervention is an abundance of anecdotal evidence attributing quality, productivity, and costs benefits to this particular quality improvement intervention. By systematic deploying of SPC with various statistical techniques and problem solving tools; organizations develop their perspective conceptualized the implementation and practice of SPC as a multi-facet change phenomenon involving the organization’s task, technology, structure, and people.

Variation is unavoidable due to controllable or uncontrollable (environmental effects) factors in process. It is important to reduce variability on process to product quality and “do it right in the first time”. The practical benefits of implementing SPC are:

• Catching the variability in a process and define the causes
• Explaining the variation with statistical tools by using samples from the process
• Using and understanding some statistical tools such as histogram, Pareto chart, cause and effect diagram, and Shewhart charts
• Catching and analyzing the patterns on process with statistical tools and interfering with them.

Elements of a Successful SPC Program:

1. Management leadership
2. A team approach, focusing on project-oriented applications
3. Education of employees at all levels
4. Emphasis on reducing variability
5. Measuring success in quantitative (economic) terms
6. A mechanism for communicating successful results throughout the organization

Measuring all quality characteristics of each product is costly and time consuming. So, sampling from process is an applicable approach to get data periodically for the monitoring process. Quality engineers take samples from the process and make descriptive analysis of the data. Descriptive data include mean, variation, standard deviation, and maximum and minimum values.

It is costly and invalid to control finished product at the end of the manufacturing process. Quality must be created in process by controlling meaningful steps of process. Therefore, the product can be built right the first time. Also, inspecting all products is impossible especially in mass product due to time, cost, and applied destructive tests. SPC is an inevitable tool for monitoring product by controlling key quality characteristics (QCs) of both product and process. Control charts were first described by Shewhart.

Variable control charts for measurement data and attribute control charts for counted data are introduced in detail in the following sections. Variability is unavoidable in both production and service process. The cause of variability comes from raw materials / semi products, machines, operators, methodology, and environment. This variability always exists naturally which is the sum of many small variabilities. If the process operates under natural variability named chance causes of variability, it is said to be in statistical control. But the process is affected by certain causes such as wearing insert, changing machine settings, operating errors, unacceptable raw material characteristics, etc. The variability comes from assignable cause and is relatively large in comparison with the chance causes of variability, so the process is out-of statistical control. Unless this assignable cause is eliminated, the process continues to produce the faulty output.

Three main sources of assignable causes:

• Lack of raw material / semi product’s QCs
• Improper machine settings or process parameters
• Operator errors

Failure Mode and Effect Analysis (FMEA)

FMEA is a specific methodology to evaluate a system, design, process, or service for possible ways in which failures (problems, errors, risks, concerns) can occur. FMEA is the method that is used to determine the potential failures of product, process, service, and/or system. The main purpose of FMEA is to predict and prevent the known or potential problems based on the historical data of a similar product or process in the design stage.

By definition, the FMEA becomes a systematic technique using engineering knowledge, reliability, and organizational development techniques; in other words, teams to optimize the system, design, process, product, and/or service.

The benefits of FMEA are given as follows:

• Identifies and ranks the known or potential failures. Therefore, risks are determined and it is possible to eliminate the important risks of product, process, services, and systems;
• Improves the quality, reliability, and safety of the products or service;
• Improves the company’s image and competitiveness;
• Helps increase customer satisfaction;
• Reduces product development time and costs;
• Helps select the optimal system design.

There are three elements of FMEA:

• Occurrence is the frequency of known or potential failures. It can be obtained from the service rates of the similar product or process or customer feedbacks.
• Severity is the effects of failures on product/ process and customers
• Detection is the ability of detecting the failures with test instructions of quality characteristics

The methodology of FMEA: Firstly, known or potential failures and modes are identified. Secondly, the occurrence, severity, and detection scales are determined for each of the failures by using defined scales. These scales range from 1 to 10 for occurrence, severity. But scale for detection ranges from 10 to 1 because of good to detect the failures via several testing methods. Risk priority number (RPN) is a product of the occurrence, severity, and detection:

RPN = Occurrence x Severity × Detection

It may be changed to reflect different situations:

• Under minor risk, no action is taken;
• Under moderate risk, some action may take place;
• Under high risk, definite action will take place;
• Under critical risk, definite actions will take place and extensive changes are required in the system, design, product, process, and/ or service.