A supplier submits samples that pass every dimensional check. Six weeks later, the customer rejects the entire production run because a secondary operation—thermal cycling in the paint oven—caused embedded weld nuts to drift out of position. The supplier lost €400,000 in scrap and missed their Start of Production date. The failure was not a lack of manufacturing capability. The failure was a lack of proof.
The supplier submitted parts. They did not submit systematic, auditable evidence that the process could consistently produce conforming components under real-world conditions. That distinction is the entire purpose of the Production Part Approval Process (PPAP). When organizations treat PPAP as a bureaucratic hurdle rather than a verification methodology, they ship unproven processes and absorb the resulting warranty costs.
PPAP is a structured conversation between a supplier and a customer that answers one question: can this process reliably produce parts that meet every requirement, every time? Codified by the Automotive Industry Action Group (AIAG), the process defines eighteen elements of evidence. Missing any one of them weakens the chain of proof and increases the probability of a field failure.
The 18 Elements Form a Narrative, Not a Checklist
Most plants treat the 18 PPAP elements as a compliance checklist. They fill out forms, attach legacy documents, and submit the package. This approach misses the structural logic of the manual. The elements tell a sequential story of how a part moves from a design intent to a proven, repeatable manufacturing process. Each element feeds the next.
The foundation is the Design Record and Engineering Change Documents. These establish exactly what the part must be, capturing every modification that altered the original design intent. If a change is not documented, it did not officially happen. Building to an unapproved revision is a direct path to a rejected submission.
The Process Flow Diagram maps every step from raw material receipt to shipment. The PFMEA takes that map and identifies what could go wrong at every step, documenting the controls used to mitigate those risks. The Control Plan then translates those risk assessments into actionable shop-floor instructions, specifying what to measure, how to measure it, and how to react to variation.
Measurement System Analysis (MSA) proves the gauges can reliably distinguish between conforming and non-conforming parts. Initial Process Studies use capability indices (Cpk, Ppk) to demonstrate statistical control. A Cpk of 1.67—the standard automotive threshold—provides a mathematical margin that absorbs normal variation as tooling wears. The Part Submission Warrant closes the package, signing the formal declaration that all evidence is complete.
Connecting the Links in the Evidence Chain
The value of PPAP lies in internal consistency. The Process Flow feeds the PFMEA. The PFMEA feeds the Control Plan. The Control Plan dictates the Dimensional Results. Every element must reference and align with every other element. A disconnect between documents is a symptom of an uncontrolled process.
If the Control Plan specifies a critical dimension measured with a CMM, but the Process Flow lacks a CMM inspection step, the documentation is internally inconsistent. If the PFMEA identifies a specific failure mode that the Control Plan does not address, there is a gap in the manufacturing controls. These gaps are cheap to fix before production starts. They cause rejects and recalls after production ramps up.
The PPAP Documentation Sequence
- 01Process Flow DiagramMaps every manufacturing step from raw material to shipment.
- 02Process FMEAIdentifies failure modes and risks at every step mapped in the flow.
- 03Control PlanTranslates FMEA risk mitigation into specific measurement and reaction plans.
- 04MSA & Dimensional ResultsProves the measurement system is valid and parts conform to the control plan.
Submission Levels and Trigger Events
Not every situation requires the full eighteen-element package. The AIAG manual defines five submission levels. Level 3 is the default for new parts and requires all 18 elements. Level 1 requires only the Part Submission Warrant and might apply to minor, low-risk changes. Level 5 applies when the customer wants to review the complete evidence package at the supplier's facility.
The customer specifies the level based on risk. The critical discipline lies in ensuring the evidence matches the risk. Submitting a Level 1 package for a safety-critical component is negligence. Furthermore, PPAP is not a one-time launch event. A re-submission is required whenever a change occurs that could affect part conformance.
I have audited plants that missed their Start of Production dates because they changed a sub-supplier, did not notify the customer, and did not re-submit. The new material passed lab tests but failed field validation. Triggers for re-submission include new tooling, manufacturing location changes, material source changes, and product reactivation after twelve months of inactivity.
Identifying the Standard Failure Modes
After two decades of managing quality systems, I can confirm that PPAP failure patterns are highly predictable. The Copy-Paste Submission is the most common. An engineer takes a previous package, changes the part numbers, and resubmits it. The PFMEA lists irrelevant failure modes and misses actual process risks. The package looks complete on paper, but it is a work of fiction.
The Golden Sample Syndrome occurs when a supplier carefully selects the best parts—or specifically adjusts the process for the sample run—to guarantee dimensional results pass. The submission reflects a capability that does not exist in normal production. When the customer audits the line or runs a production trial, the actual process performance collapses.
Compliance Mentality vs. Proof Methodology
Compliance Mentality
- Asks: What do we need to submit to get approved?
- Copies previous packages and changes part numbers.
- Adjusts the process specifically to pass the sample run.
- Treats the 18 elements as a disjointed bureaucratic checklist.
Proof Methodology
- Asks: What do we need to prove to know our process works?
- Builds FMEAs and control plans from current process data.
- Runs samples under normal production conditions and cycle times.
- Connects the process flow to the dimensional results.
The Capability Mirage relies on short-term data. The initial process study shows a Cpk of 2.0, but it is based on thirty consecutive parts from a single shift, on a freshly calibrated machine, using a hand-picked material batch. The long-term capability, which accounts for tool wear and material variation, remains unknown and untested.
The Measurement System Blind Spot destroys data integrity. The dimensional results indicate every specification is met, but the gauge has a Gauge Repeatability and Reproducibility (GR&R) of 40%. Nearly half the measurement variation comes from the gauge itself. The parts might be conforming, or they might be scrap. Without a valid MSA, the data is pure noise.
The cost of proof is always lower than the cost of failure. If you cannot prove the process works, you are gambling with customer trust and safety.
Statistical Rigor and Real Production Conditions
Doing PPAP right requires honesty and intellectual rigor. If the initial process study shows a Cpk of 1.1 against a requirement of 1.67, report it. Explain the variation. Develop a plan to improve the capability. Customers respect honesty and a corrective action plan far more than they respect fabricated data. Cpk targets exist to ensure the process spread fits comfortably within the tolerance band.
Sample parts must be produced using production tooling, production materials, production operators, and production cycle times. If any of these conditions differ from what will be used in normal manufacturing, the PPAP does not represent reality. Building a PPAP using prototype tooling or hand-selected samples invalidates the dimensional results and the capability studies.
Before submission, a cross-functional team must review the entire package as a system. The manufacturing engineer verifies that the process flow matches the actual shop floor layout. The quality engineer verifies the statistical validity of the MSA and capability studies. The supply chain representative confirms material traceability. This systemic review catches internal inconsistencies before the customer finds them.
Applying PPAP Principles Across Industries
While PPAP originated in automotive under IATF 16949, its principles apply to any industry where the consequences of a non-conforming part are severe. In aerospace, the AS9100 standard incorporates similar evidence requirements under the First Article Inspection (FAI) process defined in AS9102. The rigor is higher because the consequences of failure are catastrophic.
In medical devices, the FDA's quality system regulation (21 CFR 820) requires process validation that mirrors PPAP elements. The focus remains identical: proving the process is validated, not just verified. In electronics manufacturing, major OEMs require their suppliers to submit PPAP-like packages for critical components. The terminology differs, but the evidence chain is exactly the same.
A proper PPAP submission for a moderately complex part might cost €5,000–€15,000 in engineering time and measurement resources. A rejected production run costs ten times that amount. A field failure costs exponentially more, especially if it triggers a recall. When the cost of failure is high, the standard of proof must be equally high.