First Article Inspection (FAI) – Verifying Process Capability on the First Production Run
This topic is part of the SG Systems Global regulatory & operations glossary.
Updated October 2025 • Qualification & Release Readiness • Quality Engineering / Manufacturing
First Article Inspection (FAI) is a structured, documented confirmation that a manufacturing process can produce a specified item to all applicable requirements using production-intent materials, tooling, equipment settings, and methods. It is performed on the first run of a new product, after significant design or process changes, or when production is transferred to a new line, site, or supplier. Unlike routine in-process checks, FAI aims to verify full conformance—dimensions, materials, special characteristics, labeling/UDI, packaging, and functional attributes—by linking each requirement in the drawing/specification to objective evidence collected from the manufactured “first article.” The core outcome is not only a pass/fail disposition of the inspected unit, but a traceable dossier that proves the process, as set up, can repeatedly meet requirements when operated under the defined controls and tolerances.
In regulated industries, FAI reaches beyond mechanical dimensions. For medical devices and combination products, it ties into DHR content, label/UDI integrity via Barcode Validation, and material identity. In pharmaceuticals and nutraceuticals, “first article” often maps to line trials or initial packaging runs where eBMR steps, weights, reconciliation, and label controls are proven under live conditions. In food, first runs validate changeover effectiveness, allergen segregation, label art accuracy, net content control, and process capability at declared targets. Across sectors, the FAI record captures the as-built configuration, tying the inspected unit back to component lots, machine IDs, tooling, settings, programs, and inspection results so any future deviation can be traced to the baseline.
“FAI isn’t about proving a single part is good—it’s about proving the process is good, with evidence tight enough that any reviewer can see exactly how you made it, how you measured it, and why it will be good again tomorrow.”
1) Purpose & Scope
The purpose of FAI is to reduce the risk of latent nonconformities by verifying, before routine production, that the process and its controls can achieve the specification under realistic conditions. Scope typically includes: (1) confirmation that the latest, approved drawings, specifications, and label masters are in force; (2) verification that materials and components are released via Component Release and traceable to lots; (3) evidence that equipment is qualified and in the correct state (see IQ/OQ/PQ and calibration/cleaning); (4) measurement of all drawing and special characteristics using calibrated methods; (5) verification of labeling, UDI/SSCC, and packaging requirements; and (6) reconciliation of counts and yields, tying back to Batch Genealogy so the first article can serve as a reference. Triggering events include new parts or SKUs, engineering changes affecting form/fit/function, method/tooling changes, moved tooling, supplier changes, and prolonged shutdowns after which drift is plausible.
2) Inputs & Preconditions
An FAI relies on well-controlled inputs. Technical inputs include approved drawings/specs, bills of materials (BOM), manufacturing instructions or eMMR, label artwork and templates managed under Document Control, device programs (CNC/PLC/vision), and gauging/fixture plans. Quality preconditions include equipment qualification status (IQ/OQ/PQ), current calibration, and validated software where electronic records are used (CSV). Material preconditions require identity and status checks at receipt, correct revision and grade, and segregation of “first article” parts to prevent commingling. People preconditions include training to current procedures and, where applicable, Dual Verification for critical observations. Where labels and packaging are in scope, print/apply stations must load the approved template/version and record scan-back events as part of the FAI evidence.
3) Conducting the FAI (Plan → Make → Measure → Reconcile)
Plan. Define the part number/SKU, revision level, special characteristics, sampling (often 100% for criticals), and the measurement methods for each requirement, including gauge IDs and uncertainty. Prepare FAI forms that explicitly map each requirement to a result. Make. Run using production-intent settings, materials, and tooling; capture machine IDs, programs, fixtures, and environmental conditions (EM) if relevant to fit/function. Record any adjustments; significant changes may require restarting the run to keep evidence clean. Measure. Inspect per plan: dimensions via calibrated gauges/vision/CT; functional tests; label scannability and content; packaging integrity; net content checks for weigh-count items using device capture to avoid transcription error. Record raw values, not just pass/fail; link photos or instrument outputs. Reconcile. Confirm counts (start—scrap—yield), tie each measured unit to serial/lot where required, and verify genealogy so the first article can be traced. Nonconformities open a Deviation/NC; only after corrective actions and evidence of effectiveness is the FAI considered complete.
4) What the Dossier Contains
A complete FAI dossier typically includes: a cover sheet naming the part/SKU, revision, and reason for FAI; the applicable drawings/specs and label masters with revision/effective dates; the as-built BOM and routing; device programs and settings (hash or version ID); equipment and gauge IDs with calibration due dates; environmental conditions if relevant; the characteristic-by-characteristic results table with methods and raw measurements; photos or instrument files; label scans and print/apply acknowledgements; reconciliation and yield; approvals and e-signatures with meaning; and links to any CAPA or Change Control records. Where automation is strong, the dossier is rendered from underlying data captured during execution rather than retyped summaries, supporting audit trail review and rapid retrieval.
5) When to Repeat or Partial FAI
Repeat FAI is warranted when changes could affect form/fit/function or labeling/UDI integrity: drawing/spec revisions; material grade changes; tooling refurbishment or replacement; program changes for CNC/vision/print; transfer to a new line or site; supplier changes; or lengthy inactivity. Partial FAI narrows scope to affected characteristics but maintains full rigor on those features. Triggers should be codified in procedure with clear thresholds for “significant change.” Risk-based logic applies—e.g., for label template changes that alter variable-data fields, include print/apply and scan-back checks in scope; for gauge changes, re-establish correlation and measurement uncertainty before relying on new readings.
6) Common Failure Modes & How to Avoid Them
- Using the wrong revision. Drawing or label master out of date. Fix: lock to controlled masters under Document Control; pin version at start; block if superseded.
- “Pass/Fail only” evidence. No raw data or method link. Fix: require raw values and gauge IDs; attach instrument outputs.
- Calibration gaps. Instruments overdue or unverified. Fix: integrate calibration status checks with execution interlocks.
- Label integrity unverified. Template/version not proven. Fix: include label scan-back and print/apply acknowledgements in FAI scope.
- Uncontrolled tweaks. On-the-fly settings changes without trace. Fix: capture reason-for-change with operator/approver identity in the record.
- Genealogy gaps. Can’t trace components to the first article. Fix: enforce scans and genealogy capture at each issuance/assembly step.
- Partial scope after significant change. Critical features omitted. Fix: apply a risk-based trigger matrix defining when full vs partial FAI is required.
7) Data Integrity & Retention
FAI evidence must satisfy ALCOA+: attributable (unique users/devices), legible, contemporaneous, original/true copy, and accurate. Where electronic, apply Part 11-aligned identity, e-signature, and secure audit trails with validated backup/restore and archival. Retention commonly aligns to the product’s device history or batch record period; retrieval should be rapid and complete for inspections, customer approvals, and investigations. Storing the as-built baseline (program hashes, settings, label templates, gauges) alongside results allows efficient comparison when troubleshooting later deviations.
8) How This Fits with V5
V5 by SG Systems Global operationalises FAI by generating, executing, and rendering the dossier from live data rather than after-the-fact paperwork. In V5 MES, an FAI run is instantiated from the approved eMMR with version pinning; steps enforce barcode scans of materials, fixtures, and labels; device integrations capture raw measurements with instrument IDs; Dual Verification governs critical observations; and any out-of-tolerance result opens a Deviation/NC that can escalate to CAPA. In V5 WMS, Directed Picking ensures the correct revision/materials are issued; Bin / Location controls segregate first-article units; and genealogy is built automatically through scans and acknowledgements. The FAI report is rendered with links to drawings/specs under Document Control, calibration states, label template/version, reconciliation, and approvals with e-signatures. Change triggers are codified so engineering changes route through Change Control and, when thresholds are met, automatically require a full or partial FAI before re-releasing to routine production. For analytics, V5 trends common first-article misses by characteristic and line, feeding preventive improvements and supplier development.
9) FAQ
Q1. How is FAI different from routine in-process inspection?
Routine checks monitor a stable process using sampling and SPC. FAI is a one-time (or change-triggered) verification mapping all requirements to evidence on a first run, establishing the baseline and confirming process capability with production-intent conditions.
Q2. When is a partial FAI acceptable?
When a change clearly affects only a subset of characteristics—e.g., a label art revision or a non-critical feature dimension—and risk analysis shows no impact elsewhere. Procedures should define thresholds that force a full FAI for safety-critical or function-critical changes.
Q3. What belongs in the FAI characteristic table?
The requirement identifier, method/gauge, gauge ID and calibration due date, sample size, raw results, acceptance criteria, pass/fail, and any notes/attachments. Avoid pass/fail-only entries; raw data support later investigations.
Q4. How does FAI relate to IQ/OQ/PQ?
IQ/OQ/PQ qualify equipment and process performance in a controlled study. FAI verifies that the actual production setup using qualified equipment, approved masters, and released materials produces conforming product and correct labeling/packaging on the first run.
Q5. How should labels/UDI be handled in FAI?
Treat labels as controlled components: approve templates under Document Control, pin versions at start, record print/apply acknowledgements, and require scan-back verification. Include variable data checks (lot/expiry/UDI/SSCC) in scope where applicable.
Related Reading
• Foundations & Control: Document Control | Change Control | Audit Trail (GxP) | CSV
• Equipment & Readiness: Equipment Qualification (IQ/OQ/PQ) | Asset Calibration Status | Continued Process Verification (CPV)
• Execution & Records: eMMR | Electronic Batch Record (eBMR) | Barcode Validation | Batch Genealogy | Deviation / NC | CAPA