Blog · Board and management reporting · Manufacturing
Use exported measurements and supplied control limits to review process signals by line and batch. This guide gives spreadsheet flags, a worked example, and the checks that keep process limits separate from product specifications.
A manufacturing SPC dashboard compares time-ordered measurements with supplied statistical control limits, then identifies the line and batch behind each signal. Preserve the original order and the applicable limit version. Keep the product's specification test separate so a conforming part does not hide a process change.
Start with the process measurement export your quality team already uses. Keep sequence number, timestamp, line, batch, measured characteristic, measurement unit, and value. Include the gauge identifier and applicable control-plan reference when available.
This method reviews individual observations against existing individual-measurement limits. If the approved chart uses subgroup averages or ranges, obtain those subgroup records and matching limits instead. Comparing individual values with limits designed for subgroup averages answers a different statistical question.
NIST describes control charts as measurements or statistics plotted against sample number or time, with a center line and control limits. A summary sorted from smallest to largest removes the sequence that makes this review useful.
Use the manufacturing dashboard template for the surrounding plant workbook. Keep the measurement review at the line and characteristic level.
An illustrative Ohio plant measures shaft diameter in millimeters. Its quality team supplies the following approved limits for Line L1. These are example inputs, not limits estimated from the observations below.
| Cell | Supplied parameter | Value (mm) |
|---|---|---|
| K2 | Lower specification limit | 9.95 |
| K3 | Lower control limit | 9.97 |
| K4 | Control center line | 10.00 |
| K5 | Upper control limit | 10.03 |
| K6 | Upper specification limit | 10.05 |
The control limits describe the process's expected statistical behavior under the approved method. The specification limits describe the product requirement. NIST's process capability guidance compares the behavior of a stable process with specification limits; the two concepts have different jobs.
Record the limit version and effective date beside these cells. Confirm that the limits apply to this line, shaft design, measurement method, and period. This short review does not establish a new baseline or calculate capability indices.
Place sequence in A, line in B, batch in C, and diameter in D. Rows below are in production order. Keep timestamps in the underlying register.
| Sequence | Line | Batch | Diameter (mm) | Outside control limits | Outside specification |
|---|---|---|---|---|---|
| 1 | L1 | A | 10.00 | 0 | 0 |
| 2 | L1 | A | 10.01 | 0 | 0 |
| 3 | L1 | A | 9.99 | 0 | 0 |
| 4 | L1 | B | 10.02 | 0 | 0 |
| 5 | L1 | B | 10.04 | 1 | 0 |
| 6 | L1 | B | 10.01 | 0 | 0 |
In E2, flag a value outside the supplied control limits:
=IF(D2="","",IF(OR(D2<$K$3,D2>$K$5),1,0))
In F2, test the specification limits separately:
=IF(D2="","",IF(OR(D2<$K$2,D2>$K$6),1,0))
Fill both formulas through row 7. They treat a value exactly on a limit as inside; use the boundary convention in the approved procedure. Check that measurements and limits are numeric before relying on the flags.
The fifth observation is 10.04 mm. It exceeds the 10.03 mm upper control limit by 10.04 − 10.03 = 0.01 mm, while remaining below the 10.05 mm upper specification limit. It is a process signal even though this dimension meets the supplied product requirement.
The review calculations are:
| Measure | Spreadsheet formula | Result |
|---|---|---|
| Numeric measurements | =COUNT(D2:D7) |
6 |
| Measurement sum | =SUM(D2:D7) |
60.07 mm |
| Mean | =AVERAGE(D2:D7) |
10.0117 mm |
| Minimum | =MIN(D2:D7) |
9.99 mm |
| Maximum | =MAX(D2:D7) |
10.04 mm |
| Outside-control count | =SUM(E2:E7) |
1 |
| Outside-specification count | =SUM(F2:F7) |
0 |
The mean is 60.07 / 6 = 10.0117 mm, rounded to four decimal places. It sits near the center line, but it cannot cancel the individual signal. Retain sequence 5, Batch B, and its source record in the exception review.
A separate inspection export can summarize broader batch quality. Suppose Batch A has 3 nonconforming units among 100 inspected, and Batch B has 8 among 200. These are a different population from the diameter observations and may cover several product requirements.
| Batch | Inspected units | Nonconforming units | Nonconforming share |
|---|---|---|---|
| A | 100 | 3 | 3.00% |
| B | 200 | 8 | 4.00% |
| Total | 300 | 11 | 3.67% |
The total is (3 + 8) / (100 + 200) = 3.67%, rounded to two decimals. NIST defines the nonconforming proportion using nonconforming items divided by the relevant item population. These descriptive rates have no control-limit test here. The quality assurance dashboard example covers rework and disposition checks.
The quality engineer should examine the flagged observation against the approved response procedure. Check the original measurement, gauge status, material lot, setup changes, and adjacent observations. Record the evidence and disposition rather than assigning a cause from the flag alone.
Also inspect the ordered chart or record for the pattern rules specified in the control plan. NIST cautions that values inside the limits can still show systematic behavior. Passing the simple boundary test therefore does not establish process stability.
Keep plant output and downtime in the broader manufacturing dashboard; use this view to examine the measured characteristic.
Bring your ordered measurement export, supplied limits, and matching inspection counts to Covirage. Covirage maps the columns; its tools compute and check the figures, and the external AI model explains the results with every figure traced to its rows. Create a free Covirage account, set up your business, and bring a sample of your own data.
See Covirage for industrial manufacturers for the manufacturing context.
Control limits describe expected process behavior under an approved statistical monitoring method. Specification limits define acceptable product characteristics. A measurement can be inside the product specification and still signal a process change. Keep the limits in separate fields and label both tests clearly.
A simple outside-limit count cannot establish that conclusion. Ordered observations can show nonrandom patterns even when all values remain inside the limits. The quality team must apply the approved chart type, pattern rules, measurement checks, and investigation procedure to the appropriate process data.
Use the approved limits applicable to that process and monitoring period. Recalculating them from each batch can make a changing process appear normal. Any limit revision needs the quality team's baseline review, documented rationale, effective date, and consistent treatment of earlier observations.