How to use SPC to transform quality indicators

For Montgomery (2004), “quality is inversely proportional to variability”. The definition is one of several present in the article “Statistical Process Control (SPC) applied to automatic vending machine” (2019). The article aims to study the delivery of supplies by an automatic vending machine. In this context, it is possible to assume Statistical Process Control (SPC) as a quality tool widely used in the manufacturing industry, but also in other diverse sectors.

SPC originated in the 1920s when Walter A. Shewhart applied statistical tools to determine when corrective actions should be implemented in the process. At the time, he was working in the laboratories of Bell Telephone, the initial site of the application. Control charts on telephone lines were used to monitor wire resistance, connection quality, and electronic component performance.

His work was essential to the evolution of the SPC, which transitioned to industry with contributions from W. Edwards Deming. After World War II, Deming was asked to help rebuild Japanese industry. He introduced the concepts learned from Shewhart to improve the quality and efficiency of operations. By propagating the importance of controlled variability, it allowed Japanese industry to identify and correct production problems.

Currently, the SPC is one of the main methods of collecting and verifying the sample results of a process. The objective is to control the operation of the process and reduce the failures resulting from its execution. SPC achieved this through statistical and visualization methods, which allow it to identify non-compliant outputs, that is, to disapprove outputs that are outside the established parameters. This prevents subsequent variations and drives continuous improvement.

SPC Control Tools

Statistics is a science that, using probabilistic theories, seeks to explain the frequency of occurrence of events and model randomness and uncertainty. Its results are graphically analyzed, making it simpler to visualize the variation of the results and parameterize the data. From this, it is also possible to predict future phenomena. The graphically represented samples can be taken randomly, at specific times or in specific situations.

Thus, the SPC is a statistical method that use of quality tools, changing according to its objective. Although these tools are an important part, they only encompass technical aspects, and it is necessary to engage at all organizational levels in making improvements.

The histogram is a quality statistical tool that presents the frequency of a given value or class of values in a group of data. From graphical parameters, it provides an idea of the shape, dispersion, and central trend of the data. It is also widely used to draw quick and intuitive conclusions from aspects of the data that are not obvious when viewed in tables.

The scatter plot, on the other hand, graphically presents data collected in pairs on two variables (x,y), where the dispersion indicates the connection between them. If there is a connection, it is useful to identify deviations, as problems in one variable can lead to problems in others.

The Pareto chart is a tool used to organize data in order of importance, so that priorities for problem solving can be determined. It does not identify the most important defects, but the most frequent ones.

Among these various tools, the control chart is the most used in SPC processes, as seen below.

Control Charts

The control chart consists of a graphical tool to represent and record trends in the sequential or temporal performance of a process. The analysis of this graph indicates whether everything is occurring within the expected limits over time. By monitoring the indicators using this tool, it is possible to identify the causes of variation in the process.

This type of chart is composed of the following parts:

Centerline (LC): represents the setpoint of the process. It is the horizontal line that runs through the middle of the chart and serves as a reference to assess stability.
Upper Control Limit (LSC): the horizontal line above the centerline. It is typically defined as three standard deviations above the process average. It indicates the upper limit value of the acceptable variation.
Lower Control Limit (LIC): the horizontal line below the centerline. It is also defined as three standard deviations below the process average. It Indicates the lower limit of acceptable process variation.

Regarding interpretation, variations can be classified as either controlled or uncontrolled and can have common causes or special causes. Common causes of variation are controlled and are inherent to the process characteristics. They remain within the upper and lower limits already calculated in the process. On the other hand, special causes of variation are uncontrolled and cause deviations beyond the upper and lower limits, requiring corrective action. An example of common variation is the lack of standardization of operations. Now, when considering the special causes, it might include batches of raw material with issues, human error, or equipment deregulation.

Finally, it is important to mention that the connection between each of the variables is not taken into account here. It is assumed that they are independent.

Step by step of the application of SPC in the manufacturing industry

To start using SPC in manufacturing industry, you should follow these steps:

Determine the tool for data collection: the SPC involves data analysis, so it is necessary to choose the right tool for collecting process results. By using devices such as the Internet of Things (IoT), which collect data in real time, this process becomes more accurate and reliable.
Collect a dataset: define the sample that will be analyzed and collect the data. Here, it is necessary to consider critical characteristics of the process, such as weight, temperature, dimension, etc.
Define the limits of the process: analyze the historical data of the process to understand its behavior and calculate the standard deviation average of the obtained data to determine the UCL (Upper Control Limit) and LCL (Lower Control Limit). Alternatively verify the expected standards of the process to define its limits.
Build the chart: create control charts, based on the given thresholds and centerline to visualize the data and monitor the stability of the process. Depending on the purpose, other types of graphical control tools may also be used, as seen above.
Identify variations: check if there are data points that exceed the control limits and study them.
Determine the causes of variance and develop action plans: monitor control charts to recognize trends and variations. After that, implement corrective actions when the process presents a variation of special cause.
Continuous improvement: regularly review SPC results and make adjustments as needed to continuously improve the process.

Why use SPC?

Statistical Process Control (SPC) can be applied in several situations to ensure the quality and efficiency of processes:

When a new production process is implemented, SPC is indispensable for establishing a baseline for quality and quick identification of any initial problems.
Predict the expected range of results from a process.
Determine whether a process is stable (in statistical control).
Define whether the quality improvement project should aim to avoid specific problems or make fundamental changes to the process.
The SPC can be used to reduce costs associated with waste and rework.
Ensure that processes are in line with quality standards and regulations , as it helps to maintain consistency and predictability of results.

In addition, SPC can be combined with other methodologies, such as PDCA (Plan, Do, Check and Act). In the planning phase, SPC comes into action to identify the critical characteristics that need to be monitored, facilitating goal definition. Now, in the execution phase, the SPC is used to monitor the process in real-time, mainly through sensors and IoT devices, allowing for tracking changes and easily identifying deviations. In the verification phase, control charts play an essential role in attesting to the machine’s behavior in relation to the established standards. Finally, the action phase involves corrective actions to address variation, resulting in optimization.

At first glance, it is also possible to establish similarities between SPC and Centerlining, as they are complementary. However, both have different objectives, as the first focuses on monitoring and controlling the process, while the second on establishing optimal configurations for critical parameters.

Examples of SPC application in day-to-day production

Some examples of how SPC can be applied in the manufacturing industry:

Automotive industry: during vehicle assembly, SPC can be used in monitoring welding and painting processes to ensure the consistency and quality of parts and surfaces. It is also applicable in the manufacture of components, encompassing precision control in parts such as pistons and valves so that they meet technical specifications.
Pharmaceutical industry: throughout the production of medicines, monitoring mixing and filling processes ensures the correct dosage of inputs and the purity of products. Also, continuous verification of parameters such as temperature and humidity during production maintains the stability of the medicines.
Food industry: during food processing, controling pasteurization and packaging processes, for example, ensures food safety and product quality.
Beverage industry: monitoring fermentation and bottling processes ensures consistency in beverage flavor.
Chemical industry: monitoring parameters of the mixing process in chemical production ensures the homogeneity of the combinations and prevent unwanted and dangerous reactions;
Pulp and paper industry: during paper production, one possible application of SPC is the control of paper weight. By keeping the grammage within specifications, the industry reduces material waste and improves customer satisfaction.

SPC: the key for compliance with quality standards

According to the National Confederation of Industry (CNI), in the study “Industry in numbers” (2022), organizations that implement SPC reduce the defect rate by 50%. Also, the same material states that its implementation results in a compliance rate with quality standards of more than 95%. This helps to avoid fines and improve the market reputation. Compliance with these practices is often verified through audits conducted by regulatory agencies, such as the FDA in the United States and ANVISA in Brazil.

ISO 9001 is an example of these regulations. It is a quality management standard that requires industries to use data-driven methods to measure the effectiveness of processes. Additionally, it highlights the crucial role of leadership in implementing and maintaining an effective quality management system. SPC is an essential tool for meeting these requirements, as it helps to ensure that products and processes are under control.

ISO 22000 is an international standard that establishes the requirements for a food safety management system. It was developed by the International Organization for Standardization (ISO) and aims to ensure that all organizations involved in the food chain can consistently provide safe products. It also benefits from the SPC to monitor critical processes and ensure product safety.

Finally, HACCP is a preventive food safety management system. It focuses on identifying potential biological, chemical, and physical hazards that may occur at any stage of production. The SPC is used precisely to monitor these critical points, ensuring quality.

Improve SPC with Data Science

The integration of SPC with data science is a powerful approach to continuous process improvement. By combining traditional statistical methods with advanced data analysis techniques, industries can gain more assertive insights and make more informed decisions. This not only helps to comply with regulatory standards, but also to optimize production and reduce costs.

In summary, SPC, when combined with data science, provides a solid foundation for quality management and compliance with regulatory standards. At the same time, this union drives innovation and efficiency in the manufacturing industry. Learn more about us.

How to use SPC to transform quality indicators