If you work in manufacturing, you’ll understand how critical it is to identify defects as early as possible. Statistical process control (SPC) is a powerful toolset developed in the 1920s to help achieve this. W. Edwards Deming was instrumental in standardising SPC during the Second World War, later introducing it to Japan, where it became a core part of Six Sigma and, by extension, lean manufacturing.
What is "Statistical Process Control (SPC)"?
Statistical Process Control (SPC) is a methodology used to monitor, control, and improve manufacturing processes through statistical analysis. It involves collecting data from various stages of production and using that data to detect variations in output—allowing manufacturers to intervene before these variations result in defects or process failure.
SPC operates on the principle that all processes exhibit some level of variability. However, when that variability exceeds a certain threshold, it may indicate a problem requiring immediate correction. By charting process behaviour over time, SPC helps manufacturers determine whether a process is stable and predictable or if it's drifting out of control due to unforeseen issues.
Core elements of SPC include:
-
Control charts that visualise data trends
-
Process capability analysis to assess whether a process can meet specification limits
-
Ongoing data collection for continuous monitoring and improvement
Used properly, SPC ensures consistency, predictability, and higher product quality, making it a foundational tool in quality assurance, Six Sigma, and lean manufacturing frameworks.
Why is SPC so useful?
SPC evaluates process output by tracking small, statistically significant changes—allowing you to correct issues before defects occur. Initially used in manufacturing to minimise waste and scrap, SPC is now also applied in healthcare and service industries.
SPC involves:
-
Statistical analysis
-
Design of experiments
-
Monitoring via control charts
-
Two-phase implementation: establishing process capability, then ongoing monitoring
The effectiveness of SPC depends on establishing the correct monitoring frequency, based on various influencing factors.
Key concepts in statistical process control
A core idea within SPC is process variation. This can arise from two sources, as first defined by Shewhart:
Assignable (special) causes vs. Chance (common) causes
-
Common causes: Random, unavoidable fluctuations inherent to the process
-
Special causes: Specific, identifiable disruptions (e.g. a change in operator, material, or temperature)
Understanding the source of variation allows manufacturers to predict performance and take appropriate action. Modern terminology often refers to assignable causes as special causes and chance causes as common causes.
SPC and measurement system analysis (MSA)
SPC concepts align closely with measurement system analysis:
-
Common causes ↔ precision issues in MSA
-
Special causes ↔ bias or trueness problems in MSA
The objective is to remove special cause variation and achieve a stable process influenced only by known, random effects.
Control charts and variation monitoring
Control charts are central to SPC. They help visualise whether a process is stable or if a special cause has emerged.
Key benefits include:
-
Predicting defect probability
-
Identifying tool wear, bias, or other emerging issues
-
Maintaining consistent quality throughout production
Only after special causes are addressed can a process be considered stable and limits reliably defined.
Basic statistical concepts for SPC
You don’t need to be a statistician to benefit from SPC. A basic understanding of a few key ideas is sufficient:
Standard deviation
A measure of dispersion within a data set. If you monitor 20 manufactured parts, standard deviation tells you how much they vary from the mean value—providing a more reliable measure than simply using the range.
Dispersion and probability
Dispersion refers to how far individual values deviate from the mean—regardless of direction. The triangular distribution arises when two random effects are combined. When multiple effects with varied distributions combine, they form a normal (Gaussian) distribution, as described by the central limit theorem.
Probability distribution in SPC
Once standard deviation is known, you can determine the probability of a defect occurring based on where a measured part falls within the distribution.
If a value lies outside the expected range, it may signal:
-
A process that’s out of control
-
A new, unidentified special cause
Early detection via SPC enables timely intervention, preventing small problems from escalating into full production line failures.
Understanding and applying statistical process control is vital for maintaining manufacturing quality. SPC enables early detection of anomalies, helping to stabilise processes and reduce waste. With tools like control charts and basic statistical methods, manufacturers can predict performance, pinpoint variation, and take corrective action—before costly defects occur.
Implementing SPC is not just about compliance or process documentation—it’s about creating a predictable, reliable, and efficient production environment where quality is built into every step.
