Statistical process control (SPC) software
With JMP, go beyond static control charts to detect variation, uncover root causes, and drive continuous improvement.
This guide covers how various SPC software differs, what to look for in a solution, and why manufacturers choose JMP.
The challenge with process variation
Process variation sounds simple in theory. In practice, teams are managing hundreds of variables, multiple production lines, supplier inputs, and measurement systems – each with its own sources of error.
Most manufacturers can identify a problem after defects appear. The real challenge is detecting meaningful change early enough to act while avoiding false alarms that create unnecessary adjustments. Overreacting to normal process noise can make a stable process worse.
SPC succeeds when teams can separate common-cause variation from special-cause variation and quickly investigate process shifts. It becomes difficult when engineers are forced to piece together information from spreadsheets, dashboards, disconnected databases, and static reports.
What SPC software actually does
Statistical process control software monitors process behavior over time using statistical methods to distinguish expected variation from signals that indicate a real process change.
At its core, SPC software helps manufacturers answer four questions:
- Is the process stable?
- Is the process capable of meeting customer specifications?
- If an issue is detected, what changed?
- What should we do next?
Most SPC platforms support control charts, capability analysis, alerts, measurement systems analysis, and reporting. Where they differ significantly is in what happens after an alarm – whether the software helps engineers investigate and resolve the issue, or if it remains hands-off.
SPC software also differs from quality management system (QMS), manufacturing execution system (MES), or business intelligence dashboards. Those systems collect, route, or display data. SPC is the analytical layer that helps engineers interpret that data and act on it.
What to look for when evaluating SPC software
Basic control charts are no longer enough for many manufacturing environments. Software that only supports basic Shewhart charts may create bottlenecks when processes become more complex.
Flexible control charts and SPC rules
Manufacturing environments vary widely.
Teams should evaluate whether software supports:
- Standard and advanced control chart types.
- Rational subgrouping.
- Custom SPC rules.
- Western Electric and Nelson rules.
- False alarm reduction strategies.
Rigid systems that lock teams into fixed monitoring rules can reduce trust in SPC workflows over time.
Interactive investigation workflows
A static dashboard can indicate that something has changed. Engineers still need to determine why it changed.
Interactive SPC environments allow users to:
- Drill into out-of-control points.
- Compare shifts, lots, tools, suppliers, and operators.
- Explore linked visualizations.
- Investigate process relationships without rebuilding reports.
With SPC, engineers spend less time rebuilding analyses and more time finding answers.
Support for non-normal process data
Real manufacturing data is often non-normal. Capability analysis that assumes normality without verification can produce misleading conclusions.
Strong SPC software should support:
- Distribution fitting.
- Goodness-of-fit testing.
- Non-normal capability analysis.
- Data transformation workflows when appropriate.
This functionality is especially important in regulated manufacturing environments where capability metrics may support compliance documentation.
Measurement systems analysis
If the measurement system is unreliable, SPC conclusions also become unreliable.
A strong SPC platform should support:
- Gauge repeatability and reproducibility (R&R) studies, including evaluating the measurement process (EMP).
- Bias analysis.
- Linearity studies.
- Stability analysis.
MSA should be integrated into the workflow rather than treated as a separate afterthought.
Integration with engineering workflows
SPC should not exist in isolation.
The best platforms connect SPC analysis to broader quality and engineering workflows, including:
- Root cause analysis.
- DOE.
- Reliability analysis.
- Predictive modeling.
- Process optimization.
Auditability and compliance support
In pharmaceutical, aerospace, semiconductor, and medical device manufacturing, reproducibility matters.
When evaluating software, be sure to determine if it supports:
- Saved workflows.
- Scripted analyses.
- Version control.
- Traceable reporting.
- 21 CFR Part 11 expectations.
Why engineers choose JMP for SPC
The distinction that matters most for engineering teams is whether the software can carry an investigation from detection through to resolution without forcing engineers out of the environment. Here's how JMP is built for that workflow.
Choose the right SPC method for your data
Apply the SPC methods that best fit the data and process being monitored, from standard control charts to advanced analyses for multivariate, non-normal, and complex process data.
- XBar-R and XBar-S charts.
- I-MR charts.
- Attribute charts.
- EWMA and CUSUM charts.
- Rare event and short run charts.
- Multivariate SPC.
- Process screening.
- Process capability analysis for normal and non-normal data.
Investigate anomalies interactively
Explore process behavior from multiple perspectives in Control Chart Builder. Drag variables to build charts on the fly, switch chart types or add phases without starting over, and click straight through from a flagged point to the underlying data or a full capability report.
- Build charts by dragging variables directly into the workspace.
- Add grouping or phase variables to compute separate control limits automatically.
- Hover over a flagged point to see which test it failed, or click it to highlight the source rows in your data table.
- Filter live to isolate specific operators, shifts, or lots for focused investigation.
- Add spec limits with one click to surface a capability report in the same window.
Validate measurement systems
Ensure that measurement data is accurate, consistent, and reliable before using it to monitor and improve process performance for both continuous measurements and pass/fail inspection.
- Gauge R&R: Separate repeatability and reproducibility using the EMP or AIAG method.
- Bias: Compare measurements to a known standard to catch systematic offset.
- Linearity: Check whether bias changes across the entire measurement range.
- Stability: Monitor a gauge's bias over time to catch drift early.
- Attribute agreement: Evaluate pass/fail or go/no-go inspection systems with Kappa statistics.
Screen processes at scale
Analyze many processes simultaneously to quickly identify instability, shifts, and capability issues, then drill into exactly the ones that need attention.
- Screen large numbers of processes from a single data table in one pass.
- Spot unstable or incapable processes instantly with color-coded stability and capability indices.
- Catch significant shifts and gradual drift with dedicated shift and drift graphs.
- Rank processes by business importance and preview any of them with a hover-over graphlet before diving in.
- Drill into any flagged process with a full control chart or capability analysis in one click.
Connect SPC and DOE
Many SPC tools identify that a process has changed. JMP helps engineers determine why it has changed so they can then find the optimal settings going forward. Move directly from SPC into designed experiments and process optimization without exporting data into separate analytical systems.
- Pinpoint exactly what kind of shift occurred using built-in tests for special causes.
- Add a suspected variable directly into the control chart layout to see if it visually explains the shift before designing a full experiment.
- Screen dozens of candidate variables at once to see which ones are statistically linked to the shift, narrowing down what's worth testing.
- Build a cause-and-effect diagram to map out potential contributors to the shift, then flag the ones that belong in your next experiment.
- Design an experiment that builds on your historical process data instead of starting from scratch, then interactively explore the results to identify your optimal settings.
Collaboratively monitor and resolve issues
After publishing SPC dashboards and reports to JMP Live, your whole team can work from the same live view with no separate exports or static reports to pass around.
- Review live control charts and dashboards from any browser, with no JMP installation required.
- Get notified automatically when an issue needs attention, then assign, resolve, and document it in one place.
- Screen thousands of processes in a shared dashboard to keep everyone focused on what matters most.
- Pull in data from external suppliers and multiple production sites to catch issues early and hold every step of production to the same standard.
A real-world example
Seagate Technology
Global manufacturing operations
Seagate runs continuous 24-hour SPC monitoring across global manufacturing operations using JMP, screening large numbers of variables and moving from process signals to decisions faster than disconnected tools allow.
The role that JMP plays is to really help us explore that data in the beginning… to determine which parameters and interactions are the most important. And then explore which types of SPC monitoring tools might be really effective.
Ted Ellefson
Managing Principal Engineer, Mechanical R&D
Seagate