To monitor, control, and improve manufacturing processes, a systematic approach such as Statistical Process Control (SPC) is indispensable in evaluating the stability of a particular process over time. Control charts are a commonly used graphical tool, and the determination and application of the appropriate method is key to obtaining accurate and useful actionable data.
What is a control chart?
In the automotive manufacturing industry especially, consistency is critical. A control chart helps ensure that:
- Parts meet dimensional and performance expectations
- Processes remain predictable and capable
- Quality issues are caught early before becoming defects or recalls
All control charts share a similar structure of components:
- A vertical axis, which represents the characteristic being monitored
- A horizontal axis, which represents the sample number or time
- A centerline, or CL, which represents the average or target level of the performance of the process
- The Upper Control Limit, or UCL, and the Lower Control Limit, or LCL, which represent the acceptable range of performance of the process
There are two broad classifications of control charts:
- Variable Data: the numerical value that can be measured on a continuous scale, such as length, weight, temperature, and time.
- Attribute Data: data that can be classified into discrete categories, or attributes, such as pass or fail results and yes or no responses.
In this blog, we will focus on control charts for variable data, explain the three main types, and clarify when each should be used. Control charts for attribute data will be discussed in a separate blog.
What is variable data?
Variable data is summarized by measuring both process centering and process spread, Because both of these important characteristics are tracked, a control chart for variable data is always represented as a pair of charts. Process centering is the adjustment of the process output to bring it closer to the target value. Process spread is the measure of variation from the average. This dual view ensures that a process is both accurate and consistent.
There are three main types of control charts used for variable data:
- X̅ and R chart (spoken as X bar and R chart)
- X̅ and S chart (spoken as X bar and S chart)
- I-MR chart (spoken as I M R chart)
The selection depends primarily on subgroup size, or total number of data values, and the situation in which the data is collected. Choosing the correct chart ensures the variability of the process is addressed accurately.
X̅ and R Chart
An X̅ and R chart is a pair of control charts used to examine the performance of the process over time by monitoring the process average and range within each subgroup.
- X̅ (X bar): Represents the average of data values in each subgroup
- R (Range): Represents the variability within the subgroup, calculated by subtracting the smallest value from the largest value
When to use it:
The X̅ and R chart is preferred when :
- Multiple subgroups of data are collected at specified time periods of the process
- The subgroup size is greater than one but less than ten (The typical subgroup size is five.)
How to use it:
First, sample data are collected where each time period represents a subgroup. Then, initial data calculation is performed, followed by control limits calculation. Finally, the X̅ and R chart is plotted with the X̿, spoken as X double bar, which represents the grand average across all subgroups, the R̅, spoken as R bar, which represents the average range across all subgroups, the UCL, and the LCL.
Example Scenario: Using an X̅ and R Chart
A machining operation produces five steering knuckles every 30 minutes. Dimensional measurements, such as bore diameter, are taken for each subgroup of five. The X̅ and R chart helps identify:
- Drift in the cutting tool affecting the average bore size
- Increased variability due to tool wear or improper fixturing
X̅ and S Chart
An X̅ and S chart is another pair of control charts used for examining the performance of the process over time by monitoring the process average, but instead of using range, it uses standard deviation to measure variability.
- X̅ (X bar): Average of data in each subgroup
- S (Standard Deviation): Measures variation of all data values in the subgroup
When to use it:
Because standard deviation uses all readings, it offers a more complete picture of variation than the range, especially for larger samples. It is preferred when:
- Subgroup size is greater than ten
- A more accurate measure of variability is required
How to use it:
First, sample data are collected where each time period represents a subgroup. Then, initial data calculation is performed, followed by control limits calculation. Finally, the X̅ and S chart is plotted with the X̿, which represents the grand average across all subgroups, the S̅, spoken as S bar, which represents the average standard deviation across all subgroups, the UCL, and the LCL.
Example Scenario: Using an X̅ and S Chart
A supplier provides rolled steel sheets used in stamping. Quality engineers collect 12 thickness readings from each sheet. Because the subgroup size is large, the X̅ and S chart is appropriate to monitor:
- Uniformity of the sheet thickness
- Process changes at the supplier (e.g., rolling mill variation)
I-MR Chart
An I-MR chart, also known as an Individual-Moving Range chart, is a pair of control charts used for examining the performance of the process over time by monitoring the individual data values, as a measure of process setting, and moving range, as a measure of variability.
- I Chart: Plots individual measured values
- MR (Moving Range): Represents the absolute difference between consecutive values
When to use it:
The I-MR chart is used when:
- Subgroup size is one
- Data collection is destructive, costly, or occurs one unit at a time
- The process naturally produces one measurable output per cycle
How to use it:
Before the I-MR chart can be plotted, the sample data must be collected first. Then, the moving range calculation is performed, followed by control limits calculation. Finally, the I-MR chart is plotted with the X̅, which represents the average of the individual data values, the , spoken as MR bar, which represents the average moving range, the UCL, and the LCL.
Example Scenario: Using an I-MR chart
A torque tool applies a critical bolt on an assembly line. The torque value is recorded for each event—one at a time. The I-MR chart helps engineers detect:
- Gradual drift due to tool calibration issues
- Sudden variability caused by lubrication or fastener thread problems
Quick Comparison: How to Choose the Right Chart
Understanding when to use each type of variable data control chart is essential for maintaining stable, predictable processes in automotive manufacturing.
| Situation | Subgroup Size | Best Chart to Use | Why? |
|---|---|---|---|
| Machining parts with several measurements per hour | 2–9 | X̅ and R | Simple, effective for smaller subgroups |
| High-volume data collection with large samples | 10+ | X̅ and S | More accurate standard variation measurement with larger subgroups |
| One measurement per cycle (robot torque, destructive tests) | 1 | I-MR | Tracks individual values and cycle-to-cycle variation |
Selecting the appropriate chart ensures that both process centering and spread are monitored effectively, enabling early detection of process issues and promoting continuous improvement.
If you’d like to find out more about using variable control charts and other types in SPC, as well as see examples of other applications and calculations, enroll in the THORS course, Statistical Process Control (SPC) Fundamentals today.
