2026.07.01 | Closed-Loop Quality Improvement Series
AIQ Process Parameter Digitalization: Using P&Q, SPC, and Big Data to Find Quality Root Causes
Author: Cheng Chi-tang | MiDFUN · Compiled by: MiDFUN Editorial Team
This article is based on Cheng Chi-tang’s presentation on closed-loop quality improvement. It will be updated if the speaker provides further clarification.
One-sentence definition
AIQ process parameter digitalization connects quality results (Q) with process parameters (P) by lot, time, equipment, and operation. Engineers can trace SPC signals back to the process conditions behind quality variation instead of looking only at final inspection results.
This article answers
Why quality results alone are not enough, and how AIQ brings ERP, MES, QRP/SPC, and machine data into one analytical context.
Who should read it
Process engineers, quality engineers, R&D, manufacturing managers, data engineers, and smart-factory teams evaluating process-parameter analytics.
Conclusion first
AIQ helps engineers shorten the path from a quality issue to actionable process clues.
In this article
The hard part of a quality problem is often not seeing the result, but finding the cause hidden in the process. Dimensional shift, yield loss, appearance defects, or inadequate Cpk are outcomes. The clues may lie in temperature, pressure, speed, flow, material lot, equipment state, or the shape of a process curve.
AIQ places quality results and process parameters in the same time-and-lot context, so engineers do not have to manually reconcile ERP, MES, machine databases, SPC reports, and paper records.
Why are quality results alone not enough?
Final inspection can tell engineers that a product has already deteriorated. The process, however, may have started drifting before a result crossed a specification limit: a heating curve may slow, a pressure peak may widen, speed variation may increase, or a material addition time may change. Without connected process data, root-cause analysis depends too heavily on personal experience and manual comparison.
Speaker’s viewpoint in brief
AIQ does not claim that AI can produce the answer by itself. It helps engineers see relationships between quality outcomes and process parameters faster, while turning the analysis into traceable quality knowledge.
AIQ data foundations: ERP, MES, QRP/SPC, and machine data
| Data layer | Typical data | Analytical value |
|---|---|---|
| ERP | Orders, part numbers, lots, suppliers, and cost | Connect quality issues to material and business context |
| MES | Work orders, process stations, equipment status, and output | Connect quality variation to production history |
| QRP/SPC | Inspection values, OOS/OOC, control charts, and quality-issue records | Trace process factors backward from quality outcomes |
| Machine / sensor | Temperature, pressure, speed, flow, and curves | Identify leading indicators of quality |
AIQ depends on connecting these sources through lot, work order, timestamp, equipment, operation, or part number. If those common keys are not governed, AIQ cannot support root-cause analysis reliably.
Engineering curves and dynamic control limits
Process data is often a curve rather than a single value: heating profiles, pressure curves, material-addition timing, equipment speed, or flow. EDA engineering-curve analysis can turn those curves into comparable features such as peak width, time of occurrence, duration, heating slope, or shape differences.
Some process signals change quickly by nature. Fixed upper and lower limits can create false alarms or missed detection. Trend adjustment, error feedback, and asymmetric control limits can make the system more sensitive to relevant patterns and help engineers identify drift earlier.
Key takeaway
P&Q analysis is not about dumping data into a model. It first makes quality outcomes and process conditions comparable within the same lot, time, and equipment context.
How do different roles use AIQ?
| Role | Use case | Decision value |
|---|---|---|
| RD | Compare theoretical curves and parameter combinations during new-product experiments | Identify better recipes or process conditions |
| PE | Monitor production parameters and analyze equipment conditions against quality variation | Stabilize the process and shorten setup time |
| QE | Trace process factors from SPC signals | Support 8D root-cause analysis |
| MGR | Track critical parameters, yield, cost, and improvement results | Prioritize improvements and investment |
AIQ FAQ
What does P&Q mean?
P stands for process parameters and Q for quality outcomes. P&Q analysis connects process conditions with quality results to identify process factors that may influence quality.
Does AIQ require a complete MES first?
Not necessarily, but traceable process data is required. If MES coverage is incomplete, start with critical machines, databases, sensors, or selected process stations.
Connect quality outcomes back to process causes
MiDFUN AIQ Intelligent Quality SystemMiDFUN helps manufacturers integrate SPC, process parameters, and production history into an analyzable, traceable smart-quality architecture.
