Control Plan: Three-Phase Template for Prototype, Pilot and Production
What is a Control Plan?
A control plan is a core APQP and IATF 16949 document that describes how each manufacturing step controls process parameters and product characteristics, including what is controlled, the control method, sample size and the reaction plan. It is the bridge between PFMEA and work instructions: PFMEA identifies risk, the control plan specifies the controls, and work instructions carry out the operations. The production control plan must match actual shop-floor practice, making it a key document verified during customer process audits.
When to Use It
Use a control plan during new product launch under APQP, when a process, equipment or customer requirement changes, and whenever PFMEA or significant customer complaints are updated. It is required across prototype, pilot and production phases, with the production phase being the strictest. Review and update the plan periodically per customer requirements, and use it as the basis for layered process audits.
How to Use It (Step by Step)
Start from the high-risk items in the PFMEA and list the process steps that need control. Fill in the AIAG-style fields per operation: process step, equipment and tooling, characteristic type (product or process), special characteristic symbols, method (measurement tool, control technique, sample size, frequency) and reaction plan. Define the reaction plan for anomalies, stop and isolate and identify who to notify, then verify it is executable on the floor, train operators, and update the document with version and date on every change.
Key Formulas / Example
The control plan is a structured table rather than a formula-driven tool. Special characteristics are flagged with symbols such as SC for safety or regulatory characteristics and CC for important characteristics affecting function, assembly or customer satisfaction; their sources include customer drawings, DFMEA severity 9-10 and historical complaints. Sample size and frequency follow the risk level of each characteristic, and key characteristics generally get tighter sampling.