Between 2009 and 2019, defective Takata airbag inflators killed at least 27 people and injured hundreds more. The devices designed to save lives instead became fragmentation grenades, spraying metal shrapnel into vehicle cabins during deployment.

The recall eventually encompassed over 100 million inflators across virtually every major automaker. Takata filed for bankruptcy. Criminal charges were filed. And the root cause? A manufacturing process that wasn't adequately analyzed for failure modes.

The inflators used ammonium nitrate as a propellant—a cost-effective choice that degraded when exposed to moisture and temperature cycling. The manufacturing process didn't adequately control moisture ingress. The failure mode was predictable, preventable, and catastrophic.

FMEA Was Born for Exactly This Scenario

Failure Mode and Effects Analysis originated in manufacturing contexts. When the U.S. military developed the methodology in the 1940s, they were concerned with production reliability—ensuring manufactured items performed as designed.

The automotive industry perfected FMEA. Every major automaker requires Process FMEA (PFMEA) for manufacturing operations. The AIAG-VDA methodology now represents industry best practice, with a seven-step process refined over decades:

Planning and Preparation – Define the manufacturing process scope and assemble the analysis team
Structure Analysis – Map process steps, equipment, and material flow
Function Analysis – Document what each process step must accomplish
Failure Analysis – Identify how each step can fail and the resulting effects
Risk Analysis – Assess severity, occurrence, and detection for each failure mode
Optimization – Implement controls for high-priority risks
Results Documentation – Create traceable records enabling continuous improvement

Dissecting the Takata Failure Through FMEA Lens

The Takata inflator manufacturing process had clear failure modes that proper PFMEA would have surfaced.

Structure Analysis

The inflator manufacturing line included propellant tablet production, tablet loading into inflator housings, seal welding, and final testing. Each step introduced potential failure modes.

Function Analysis

Propellant tablets must maintain chemical stability throughout product lifetime. Housing seals must prevent moisture ingress for 15+ years. Assembly processes must not introduce contamination or damage.

Failure Analysis

Here's where systematic analysis would have raised alarms:

Moisture Absorption During Manufacturing

Failure Mode: Propellant tablets absorb ambient moisture during manufacturing
Effect: Degraded propellant chemistry causing overly aggressive burn rate
Cause: Inadequate humidity control in production environment

Incomplete Seal Welding

Failure Mode: Seal welding creates incomplete hermetic closure
Effect: Moisture ingress over product lifetime
Cause: Welding parameter variation, contamination at seal interface

Propellant Composition Variation

Failure Mode: Propellant composition varies between batches
Effect: Inconsistent burn characteristics across production
Cause: Raw material variation, inadequate incoming inspection

Risk Assessment

The severity was always 10—airbag failures during deployment are potentially fatal. The question was occurrence and detection. With proper PFMEA, the team would have asked:

What is our process capability for moisture control?
How do we verify seal integrity?
What variation exists in propellant properties?

These questions would have revealed inadequate controls long before 100 million defective units shipped.

Manufacturing Failure Modes Every Plant Should Analyze

Tool Wear Progression

Failure Mode: Cutting tool wears beyond tolerance before scheduled replacement
Effect: Parts produced outside specification
Current Controls: Time-based replacement schedule
Detection: 4 (some in-process measurement exists)
Action: Tool wear monitoring, statistical process control on critical dimensions

Setup Parameter Drift

Failure Mode: Machine settings drift from nominal over production run
Effect: Gradual quality degradation, specification escapes
Current Controls: First-article inspection only
Detection: 6 (discovered at end-of-run inspection or customer complaint)
Action: Statistical process control, in-line measurement, closed-loop control

Material Contamination

Failure Mode: Foreign material introduced during handling or processing
Effect: Product defects, customer returns, potential safety issues
Current Controls: Visual inspection, cleanliness procedures
Detection: 5 (small contamination easily missed)
Action: Covered containers, positive airflow, enhanced lighting at inspection

Operator Error in Manual Operations

Failure Mode: Assembly step performed incorrectly or omitted
Effect: Defective product, rework required, potential field failures
Current Controls: Work instructions, training
Detection: 6 (some errors caught at test, others escape)
Action: Error-proofing (poka-yoke), sequence verification, assembly verification systems

Environmental Excursions

Failure Mode: Temperature or humidity exceeds material storage requirements
Effect: Material degradation, latent defects in finished product
Current Controls: Environmental monitoring, documented limits
Detection: 5 (excursions may occur between monitoring intervals)
Action: Continuous monitoring with alarms, material segregation pending review

Building Effective PFMEA Practice

Manufacturing organizations often have legacy PFMEA documents—created during product launch, then filed away. Effective practice treats PFMEA as living documentation that evolves with processes and products.

During Process Development

Before equipment is specified and layouts finalized, analyze failure modes. Decisions made at this stage—equipment selection, control strategies, inspection points—determine whether failure modes will be managed or merely documented.

At Production Launch

Update PFMEA with actual process capability data. Theoretical occurrence ratings become empirical. Detection effectiveness is validated or revised. This is when paper analysis meets physical reality.

During Production

Every quality escape, every customer complaint, every near-miss updates your understanding:

Did the failure mode exist in your PFMEA?
Was the detection rating accurate?
Are controls working as intended?

Continuous improvement requires continuous PFMEA refinement.

When Changes Occur

New materials, new equipment, new suppliers, new operators—each change introduces potential failure modes. Effective change management triggers PFMEA review. This discipline prevents the slow accumulation of risk that characterizes many quality problems.

The Automotive Standard for Manufacturing Quality

IATF 16949, the automotive quality management standard, mandates PFMEA for production processes. Suppliers to automotive OEMs undergo rigorous audits verifying PFMEA practice. The result is extraordinary quality—modern vehicles contain thousands of components, yet defect rates are measured in parts per million.

This level of performance isn't automatic. It results from disciplined application of failure analysis at every manufacturing step. Organizations outside automotive can adopt these practices without automotive overhead.

The Takata tragedy demonstrates what happens when PFMEA practice fails. The failure modes were knowable. The consequences were foreseeable. The controls were achievable. What was missing was the organizational commitment to rigorous analysis and follow-through.

One hundred million recalls later, the lesson should be clear. Manufacturing excellence requires systematic failure analysis—not as a documentation exercise, but as a genuine commitment to understanding what can go wrong and ensuring it doesn't.

Your manufacturing process contains failure modes. The only question is whether you'll discover them through analysis or through crisis.

FMEA in Manufacturing: Learning from the Takata Airbag Crisis