Industry is no stranger to statistical process control (SPC). The control chart and the idea that a process may be statistically controlled were invented in 1924 by a worker at Bell Laboratories. William A. Shewart was his full name. He went on to write a book called “Statistical Method from the Perspective of Quality Control,” which he finally published (1939). The military used the SPC process extensively in ammunition and weapons installations throughout World War II. They needed a better and more effective approach to check product quality without jeopardizing safety because of the increased demand for the product. This was met by SPC. Following World War II, the usage of SPC tactics in the United States waned. It was later adopted by Japanese manufacturers, who continue to use it today.
Because American industry was under pressure from high-quality items imported from Japan in the 1970s, SPC began to regain popularity. SPC is now a commonly utilized quality technique in a variety of sectors.
Statistical Process Control – An Overview
SPC stands for statistical process control, which is a way of evaluating and regulating quality through monitoring the production process. Product or process measurements, as well as readings from various equipment or instrumentation, are used to collect quality data. Data is gathered and analyzed in order to evaluate, monitor, and regulate a process. SPC is a powerful tool for achieving continuous improvement. We can ensure that a process is operating at its full capacity by monitoring and managing it. The document released by the Automotive Industry Action Group is one of the most extensive and significant sources of knowledge about SPC (AIAG). You can also look for implementing SAFe course to have a grasp of SPC.
SPC Implementation – Step-by-step Process
While SPC may assist manage processes and assure consistent output, it is rarely used in China’s industries. It is best to follow the logical phases in the implementation of this tool in a factory to understand how it works:
1. Determine which product features are crucial to quality (CTQ)
A screwdriver, for example, might not be able to do its intended purpose if the metal hardness is lower than predicted. A CTQ trait is hardness.
2. Determine which procedures are crucial
What defines a “critical” process? If the procedure fails, it will very certainly have a significant influence on at least one CTQ feature.
On the control plan, critical processes are frequently marked with an asterisk (*).
3. Determine whether machines are capable of calculating SPC on their own
Many sophisticated pieces of technology gather and analyze data before sending out a warning if things get out of hand. However, as this is unusual in China, let’s suppose that isn’t the case; this means you’ll have to go through stages 4 and 5.
4. Gather data and process knowledge of what influences the process’s outcome
The variables that may influence the result of a gluing process, for example, are given below:
- Glue as a process input
- Process parameters include cycle time and fixture.
- Climate: outdoor temperature, humidity, and so on.
As a result, you may deduce that the glue’s viscosity and the ambient humidity are two factors that need to be managed.
But where do you look for that information? It frequently comes from engineers, chemists, physicians, and other professionals, but it may also come from operators who have worked on the process for a long time and have recognized certain cause-and-effect linkages.
5. Maintain control over independent variables with a significant influence on the process outcome
There are two strategies to ensure that your output does not go too far from the norm:
On a regular basis, collect data on the variables you defined before, for example, 5 random samples every 4 hours.
Teach the essential calculations to the production operators and leaders, as well as how to plot them on a chart, such as the “Xbar – R” chart, which is the most often used tool. We usually establish a target for the process capability index (Cpk) and let operators plot the index’s change over time for simplicity.
6. Look for strategies to cut down on variance
After you’ve made variances evident, the following step is to figure out how to minimize them. It should look something like this if you use statistical process control charts:
Engineers and production leaders are encouraged to experiment with alternative ways as long as the Cpk index remains within the regulated limits. In most cases, achieving a Cpk objective of 1.0 takes only a few weeks. 1.33 is more difficult. 1.66 is a lot more difficult!
You can use another statistical approach known as Design of Experiments (DoE) to assist you come close to the optimal settings for variables that affect your process output if necessary. However, be aware that there are various options here, and it may become rather complicated.
7. Continue to do so in the long run
Let’s pretend you’ve succeeded in lowering variation. Set a new goal and maintain control over the critical factors. Your statistical process control system will notify you whenever the characteristics of your process change in any manner. Otherwise, don’t mess with the process; making little tweaks all the time will automatically make the process “unstable” in statistical terms.
Despite its relative ease of implementation, SPC is seldom implemented in China’s factories across most industries. Would implementing SPC assist your factory in most cases? If yes, you can work on Agile management courses online.
Process control can be implemented in a variety of ways. The following are some of the most important monitoring and investigation tools:
- Histograms
- Check Sheets
- Cause and Effect Diagrams
- Defect Concentration Diagrams
- Pareto Charts
- Scatter Diagrams
- Control Charts
Why Use SPC?
Manufacturing companies are facing rising competition nowadays. At the same time, raw material prices continue to grow. These are factoring those enterprises, for the most part, have little control over. As a result, firms must concentrate on the aspects of their operations that they can control. Companies must strive for continuous improvements in quality, efficiency, and cost-cutting. To find quality issues, many firms still rely on post-production inspection. The SPC approach is used to move quality control from detection to prevention in a business. By monitoring the process’s performance in real time, the operator can see trends or changes in the process before they result in non-conforming output and scrap.
1. Instantaneous SPC aids in lowering the margin of error
Control charts assist operators to spot and solve faults before they develop deeper problems in processes and products since they show what’s going on in a production line in real time. This eliminates the requirement for product rework or additional product costs to correct an offering.
“Control charts in your real-time monitoring software act as an early warning detection system, notifying you when it’s time to make a change,” explains Steve Wise, Vice President of Statistical Methods at InfinityQS.
“That way, you won’t finish this only to learn three hours later that you should have made improvements and now have to pay for the accompanying expenditures.”
2. Having access to high-quality data avoids manipulation
Knowing when your process is working well is just as crucial as knowing when something is wrong. Operators commonly over-tamper with a process that was operating correctly while trying to discover whether a problem exists, which can lead to greater variations.
Operators must decide whether to “do something” (change a process behavior) or “do nothing” after studying a control chart (let the process run as is). Operators are often deterred from over-tampering with their processes after understanding that they can do nothing.
3. Control charts give essential stakeholders with operational knowledge
Control charts provide a range of information for all important stakeholders engaged in the construction of a produced product, from operators and engineers to managers and executives.
Control charts may bring together data in ways that give practical insights into whether a process needs to be altered when they pull SPC-based data from a single, unified data source. Engineers can use aggregated data to assist them enhance a process. Managers can also use more advanced box and whisker and Pareto charts to get a holistic picture of the whole plant floor—or even numerous plants.
Control charts provide a key need for a range of organizations responsible for industrial quality control, allowing them to make choices based on facts rather than guesswork.
4. Access to and visibility of data level the playing field
Operators, engineers, and managers may improve their abilities and procedures based on views and “instinct” if they don’t have access to data. Even though many experienced industrial employees have solid senses, control charts are likely to confirm what they already know to be true.
But as Wise points out, not everyone is “an expert at reading the tea leaves of their processes.” Control charts, he claims, “provide beginners and new persons with rapid access to the same information and abilities.”
Control charts not only level the playing field, but they also confirm what individuals already know and eliminate process fallacies. Decision-making is improved across the business when everyone has access to the same data and is on the same page. Production knowledge gleaned from process control software will set you apart from the competition and make manufacturing difficulties easier to overcome.
Too much knowledge is never a negative thing when it comes to creating intellect. After all, having real-time insights into your manufacturing quality control processes is critical since it may save you money by preventing unexpected expenses caused by process failures and product quality concerns.
Conclusion
While statistical process control has a high initial resource cost, the return on investment generated by the information and knowledge the tool generates has shown to be a successful activity time and time again. This tool necessitates a tremendous level of cooperation, but if implemented correctly, it may significantly increase a process’ capacity to be managed and studied during process improvement efforts. One can try knowledgehut implementing SAFe course to get their hands on Statistical Process Control.
Frequently Asked Questions (FAQs)
1. How many steps are there to implement SPC?
Step 1: Select an Appropriate Method of Measurement…
Step 2: Choose a time frame for data collection and plotting.
Establish Control Units in Step 3…
Step 4: Plot Data Points and Identify Data Points That Are Out of Control
Correct Out-of-Control Data Points in Step 5….
Calculate Cp and Cpk in Step 6.
2. What do you mean by SPC?
Statistical Process Control is an acronym for statistical process control. A bad product—defined as one that does not satisfy specifications—is frequently the outcome of a faulty process in the manufacturing business. SPC is a statistical approach of quality control in which data from product and process measurements is collected and analyzed.
3. What are SPC Methods?
SPC (Statistical Process Control) is a quality regulator approach that uses statistical methodologies to monitor and control a process. This ensures that the process runs smoothly, resulting in more specification-conforming items and less waste (rework or scrap).
4. How to use SPC tool?
- Identify the processes: Identify the major process that influences the product’s output or the process that is extremely important to the consumer. In a manufacturing firm, for example, plate thickness affects product performance; then analyze the plate production process.
- Determine the process’s quantifiable characteristics: Determine the qualities that must be measured throughout manufacturing. Consider the plate thickness as a quantifiable feature in the above example.
- Select a measuring technique and execute Gage R&R: Make measurement technique work instructions or a procedure that includes the measuring device. Consider using a thickness gage to measure the thickness and developing an acceptable measuring process. Gage Repeatability and Reproducibility (Gage R & R) tests are used to determine the level of variance in the measurement.
- Create a subgroup strategy and sampling approach: Determine the subgroup size depending on the criticality of the product, as well as the sample size and frequency. For example, in a time sequence with a subgroup size of 4, gather 20 sets of plate thicknesses.
- Gather data and create an SPC chart: Collect data according to sample size and then choose a suitable SPC chart depending on data type (continuous or discrete) and subgroup size.
Discussion about this post