SPC (Statistical Process Control)

Detect process anomalies in real time and respond immediately

With WebSocket-based real-time streaming, measurement data is reflected on control charts the instant it is entered. Nelson Rules (all 8) and Western Electric Rules are applied simultaneously, automatically detecting not only simple control limit violations but also patterns such as trends, shifts, and mixtures.

• Zone classification (A/B/C) for each measurement point enables visual position verification. Even within control limits, pattern anomalies (e.g., 7 consecutive rising points, 9 consecutive points on one side of the center line) are detected proactively.
• Upon deviation detection, depending on severity, responsible personnel are notified, a CAPA is automatically generated, or a production hold event is issued. Warning-level and immediate-action-level responses are distinguished to set response priorities.
• 3-level escalation: If the primary responsible person does not respond within a set time, the issue is automatically escalated to a higher-level manager. An automatic rejection feature is also included, preventing unresponsive states from persisting.
• Shop floor POP screens: Measurement values can be entered via touch on the manufacturing floor, or data automatically collected from equipment can be reviewed. Even in offline environments, up to 500 records are queued and automatically synchronized when the network is restored.

Practical scenario: On a production line measuring automotive brake pad thickness, an X-bar/R control chart is monitored in real time. When 7 consecutive points show an upward trend, a Nelson Rule #3 violation is detected and an immediate alert is sent to the responsible engineer. The engineer records equipment wear (Machine) as the cause in the 6M root cause analysis form and registers a mold replacement action as a CAPA.

SPC deviations lead directly to quality actions

When a deviation is detected on a control chart, it automatically links to the QMS CAPA process. Without manual intervention, the full cycle of root cause analysis (6M: Machine/Material/Man/Environment/Method/Measurement) through corrective action to effectiveness verification proceeds.

• CAPA effectiveness verification: Cpk, defect rate, and alarm frequency are automatically compared before and after corrective action to confirm improvement in quantitative terms
• SPC emergency alarm to production hold event linkage supports the IATF 16949 requirement of "suspect product quarantine"
• Control limit change history is separately preserved, ready for immediate use as evidence during customer audits

Automate process capability evaluation and manage it systematically

Beyond Cp, Cpk, Pp, and Ppk, the system also calculates Cpm/Cpmk (Taguchi method), Cnpk (non-normal distribution), Z.bench, and PPM. Periodic evaluations can be automatically executed through CRON-based timer workflows, ensuring that even hundreds of control characteristics are managed without omission.

• 5 normality tests (Anderson-Darling, Shapiro-Wilk, Kolmogorov-Smirnov, Ryan-Joiner, D'Agostino-Pearson) are performed automatically. Cases where data does not follow a normal distribution are handled appropriately.
• Non-normal data processing path: normality test, distribution identification, transformation (Box-Cox/Johnson), Cnpk-based determination. Even without statistical expertise, the system guides personnel to the appropriate analysis path.
• IATF customer-required Cpk determination criteria are reflected in the workflow. General characteristics use Cpk 1.33 as the threshold, and special characteristics (safety/regulatory) use Cpk 1.67, with automatic determination and improvement action requests issued upon non-compliance.
• Cpm/Cpmk (Taguchi method) is an index reflecting process center deviation from the target value, evaluating not just whether a process is within specification but how close it is to the target. It is useful for processes where precision is critical.
• Quality KPI snapshots: Process capability indices are automatically recorded on daily/weekly/monthly intervals to track trends over time. Visualization is available through 17 dashboard widgets.

Practical scenario: On the 1st of each month, a timer workflow runs to automatically recalculate Cpk for 200 registered control characteristics. When characteristics falling below Cpk 1.33 are found, the responsible process engineer is automatically notified and an improvement action request is generated. When OEM customers require quarterly Cpk reports, Cpk trend data for the relevant period can be immediately exported from the system for submission.

Perform MSA systematically with follow-through to corrective actions

All core study types from the AIAG MSA Manual 4th Edition are supported. Gage R&R (ANOVA and X-bar/R methods), Bias, Stability, Linearity, and Attribute Agreement (Kappa/Fleiss Kappa/Kendall W) analyses can be performed.

• GRR% determination: Automatic 3-way branching in the workflow gateway for below 10% (acceptable), 10-30% (conditional), and above 30% (unacceptable)
• ndc (Number of Distinct Categories) of 5 or higher: automatic determination
• Unacceptable determination triggers follow-up workflow: equipment calibration/repair or operator retraining
• Automatic MSA report generation with file attachment capability

Connect from Design of Experiments (DOE) to control chart application

6 types of DOE (Full Factorial, Fractional Factorial, Taguchi, RSM-CCD, RSM-BBD, Screening) can be performed, and a workflow reflects the derived optimal conditions on control charts. Confirmation experiment Cpk is linked to ensure the flow from experimentation to production application is uninterrupted.

Practical scenario: When shrinkage defects recur in an injection molding process, RSM (Response Surface Methodology) is used to derive the optimal conditions for mold temperature, injection pressure, and holding time. The derived conditions are set as new control limits on the control chart, and when the confirmation experiment achieves Cpk 1.67 or higher, they are applied to production. Control limit change history is automatically recorded, preserving the rationale for why the control limits were changed.

Compared to specialized SPC software (Minitab, InfinityQS, WinSPC)

Specialized SPC software has strengths in the depth of statistical analysis functions. VEXPLOR SPC focuses on connecting SPC data with quality management processes (QMS), manufacturing execution (MES), and resource management (ERP) within a single system. Without transferring data between separate systems, deviations on control charts can be seamlessly processed through to corrective actions.

Faster process anomaly response

• Real-time monitoring delivers immediate alerts upon deviation detection. Response time can be significantly shortened compared to after-the-fact discovery.
• The full cycle from deviation through root cause analysis, corrective action, to effectiveness verification is trackable within the system.

Systematic Cpk management

• Automatic periodic evaluation enables monitoring Cpk trends for hundreds of control characteristics without omission.
• Initial process capability evidence (Cpk 1.67 or higher) required by IATF customers can be output directly from the system.

Measurement system reliability assurance

• Systematic management from MSA execution through determination to follow-up actions (calibration/retraining) enhances measurement data reliability.
• MSA history and reports can be presented immediately during customer audits.

Data-driven process improvement

• The linkage from DOE to optimal condition derivation, control chart application, and Cpk confirmation creates a structure where experimental results are actually applied to production.
• Accumulated deviation history, cause classification (6M), and CAPA effectiveness data can be utilized to identify root causes of recurring process problems.

Control chart operation flow

Control characteristic registration -> Specification (USL/LSL/Target) setting -> Control chart type selection -> Alarm rule configuration (Nelson/WE)
  -> Measurement data collection (POP input or equipment auto-collection) -> Real-time Zone classification (A/B/C)
  -> [Normal] -> Data accumulation, KPI snapshot
  -> [Deviation detected] -> Severity determination -> [Minor] Personnel notification / [Major] Automatic CAPA generation + production hold

Process capability evaluation flow

Measurement data collection -> Normality testing (select from 5 types or automatic)
  -> [Normal] -> Cp/Cpk/Pp/Ppk/Cpm/Cpmk calculation
  -> [Non-normal] -> Distribution identification -> Box-Cox/Johnson transformation -> Cnpk calculation
  -> Cpk determination -> [>=1.67] Special characteristic compliant / [>=1.33] General characteristic compliant / [<1.33] Improvement needed
  -> Result report generation -> KPI dashboard update

MSA execution flow

MSA plan development -> Gage R&R execution (ANOVA/X-bar method)
  -> GRR% calculation -> [<10%] Acceptable / [10-30%] Conditional / [>30%] Unacceptable
  -> ndc determination -> [>=5] Acceptable / [<5] Measurement system improvement needed
  -> [Unacceptable] -> Equipment calibration/repair or operator retraining -> Re-evaluation
  -> Automatic MSA report generation

DOE to production application flow

Process problem identification -> DOE type selection (6 types) -> Experiment design -> Experiment execution -> Results analysis
  -> Optimal conditions derived -> Confirmation experiment (Cpk verification) -> [Compliant] -> Control chart control limit update
  -> Control limit change history automatically recorded -> Production application

Q. We currently use Minitab. Do we need to replace it? Minitab has strengths as a specialized statistical analysis tool, while VEXPLOR SPC focuses on real-time monitoring on the manufacturing floor and integration with the Quality Management System (QMS). Parallel use is possible — continue using Minitab for advanced statistical analyses (regression, reliability analysis, etc.) while running daily control chart operations and process capability management in VEXPLOR SPC. VEXPLOR SPC calculates beyond Cp/Cpk/Pp/Ppk to include Cpm/Cpmk/Cnpk/Z.bench/PPM and supports 6 types of DOE, covering most routine SPC work.

Q. We have hundreds of control characteristics. Can all be monitored in real time? Yes. WebSocket-based real-time streaming settings and timer-based periodic evaluations can be configured differently for each control characteristic. Typically, critical special characteristics (CTQ/CTP) are set for real-time monitoring, while general characteristics are evaluated on daily/weekly periodic schedules.

Q. Can process capability be evaluated for non-normal distribution data? Yes. Normality testing (5 types) is performed automatically, and when data is determined to be non-normal, the system guides through the path of distribution identification, Box-Cox or Johnson transformation, and Cnpk-based determination. Even personnel without statistical expertise can obtain appropriate process capability indices by following the system guidance.