FMEDA: Quantitative Hardware Safety Analysis
Master the quantitative engine of ISO 26262-5: turn per-mode failure rates in FIT, failure-mode distributions, and diagnostic coverage claims into the SPFM, LFM, and PMHF metrics that decide whether a hardware design meets its ASIL target.
- Chapters
- 10
- Chapters
- Hardware Metrics
- 3
- Hardware Metrics
- Reliability Unit
- FIT
- Reliability Unit
- Interactive Visuals
- 6
- Interactive Visuals
- 01What FMEDA Is
- 02Failure Rates and FIT
- 03Failure-Mode Distribution
- 04SPF, RF, and MPF Faults
- 05Diagnostic Coverage
Why it pays for itself
Compute the three hardware metrics
Roll per-mode failure rates up into SPFM, LFM and PMHF using the Annex C equations, demonstrated on a compact numeric example, and check the results against the ASIL B, C and D targets.
Classify every fault correctly
Apply the decision tree that sorts each failure mode into safe, single-point, residual or multiple-point fault - the classification step where most FMEDAs quietly go wrong and metrics become indefensible.
Defend diagnostic coverage claims
Match safety mechanisms such as range checks, watchdogs and lockstep to the low, medium and high coverage tiers, and back each claimed percentage with fault-injection evidence instead of data-sheet optimism.
What you’ll be able to do
Rate every failure mode in FIT
Pull justified base failure rates from approved handbooks and correct them for the vehicle mission profile.
Classify faults correctly
Apply the decision tree to separate safe, single-point, residual, and latent multiple-point faults against a defined safety goal.
Defend diagnostic coverage claims
Match safety mechanisms to coverage tiers and back each percentage with fault-injection evidence rather than assumption.
Compute the three hardware metrics
Aggregate FMEDA data into SPFM, LFM, and PMHF and check the results against the relevant ASIL targets.
Build and review an FMEDA spreadsheet
Lay out columns, run the analysis workflow, and audit a completed FMEDA for the common metric-invalidating errors.
Chapter by chapter
- 01
What FMEDA Is
Positions Failure Modes, Effects and Diagnostic Analysis (FMEDA) as the quantitative, coverage-aware extension of FMEA that feeds the mandatory hardware metrics of ISO 26262-5.
- FMEDA versus FMEA: qualitative ranking versus quantitative metric inputs
- Inductive (bottom-up) analysis complementing deductive Fault Tree Analysis (FTA)
- Living document refined from hardware concept through verification and maintenance
- 02
Failure Rates and FIT
Defines the FIT unit (one failure per 10^9 operating hours), the recognised reliability handbooks, mission-profile correction, and the split of total lambda into safe and dangerous fractions.
- FIT to lambda conversion and the constant hazard rate assumption
- Source handbooks: SN 29500, IEC TR 62380, FIDES, MIL-HDBK-217F, supplier data
- Temperature and stress derating tied to the vehicle mission profile
- 03
Failure-Mode Distribution
Shows how a component lambda is apportioned across its individual failure modes by percentage, then each mode classified as safe or dangerous against the safety goal in scope.
- Distribution percentages applied to a worked current-sense resistor example
- Standard failure-mode libraries by component family
- Handling shared hardware across multiple safety goals
- 04
SPF, RF, and MPF Faults
Classifies every failure mode as Safe, Single-Point Fault, Residual Fault, or Multiple-Point Fault using a decision tree, and explains the latent multiple-point case.
- The four fault classes and a step-by-step classification decision tree
- Single-Point Fault versus Residual Fault distinguished by safety mechanism presence
- Latent multiple-point faults driving the Latent Fault Metric (LFM)
- 05
Diagnostic Coverage
Covers how a safety mechanism reduces the dangerous undetected failure rate, the low, medium, and high coverage tiers, and the evidence needed to defend a diagnostic coverage claim.
- Diagnostic coverage defined as the detected fraction of a dangerous failure rate
- Indicative coverage values for common mechanisms (range check, watchdog, lockstep)
- Credible evidence: fault injection over optimistic data-sheet assertions
- 06
Deriving SPFM, LFM, PMHF
Rolls FMEDA data up into the three mandatory metrics using the Annex C equations, then demonstrates the aggregation on a compact three-component numeric example.
- Single-Point Fault Metric and Latent Fault Metric from per-mode lambda sums
- Probabilistic Metric for random Hardware Failures (PMHF) in FIT per hour
- Metric targets keyed to ASIL B, C, and D
- 07
Building the FMEDA
Lays out the FMEDA spreadsheet column structure, the analysis workflow from bill of materials to frozen metrics, and an annotated table for an EPS phase-monitoring circuit.
- Column structure from component and lambda through mode, class, and coverage
- Annotated example table for an Electric Power Steering monitoring circuit
- Build process: enumerate, rate, distribute, classify, claim, aggregate
- 08
Worked Example
Walks an end-to-end FMEDA for a small ADAS perception subsystem (microcontroller, sensor, power supply), computing the metrics against an ASIL B safety goal.
- Roughly 260 FIT split across microcontroller, sensor, and power supply
- SPFM near 96.8 percent and LFM near 97.3 percent against ASIL B and C targets
- Observations on which modes dominate the residual fault budget
- 09
Pitfalls and Review
Catalogues the common errors that invalidate FMEDA results, from optimistic coverage claims to under-counted latent faults, and supplies a practical review checklist.
- Optimistic diagnostic coverage and incomplete failure-mode enumeration
- Wrong distributions, under-counted latent faults, double-counted coverage
- Reviewer checklist for auditing a completed FMEDA
- 10
Data Sources and Tools
Surveys the reliability prediction handbooks, automotive mission-profile standards, FMEDA tooling, and how the per-mode failure rates feed quantitative Fault Tree Analysis.
- Reliability handbooks and mission-profile definitions for passenger cars
- Dedicated FMEDA tools versus spreadsheet-based analysis
- Per-mode FIT values supplying cut-set probabilities for quantitative FTA
Not just text: the visual toolkit
FIT Rate Explorer
Interactive view of how temperature, stress, and mission profile move a component failure rate across its useful-life region.
Failure-Mode Distribution Pie
Splits a single component lambda across its failure modes and shades the safe versus dangerous fractions.
Fault Classification Decision Tree
Branching flow that sorts each failure mode into Safe, Single-Point, Residual, or Multiple-Point fault.
Diagnostic Coverage Visualizer
Shows a safety mechanism carving the detected portion out of a dangerous failure rate at low, medium, and high tiers.
Metrics Roll-up Calculator
Aggregates per-mode rates into Single-Point Fault Metric, Latent Fault Metric, and PMHF against ASIL targets.
PMHF Budget Timeline
Tracks how single-point and latent contributions accumulate into the probabilistic hardware failure budget over operating life.
ADAS Perception Subsystem FMEDA
A complete FMEDA for a small automatic emergency braking perception subsystem (microcontroller, sensor, power supply) evaluated against an ASIL B safety goal, carried from per-component rates through to final metrics.
- Total dangerous-relevant rate of roughly 260 FIT distributed across the three elements
- Each component lambda split into failure modes and classified as safe, single-point, residual, or latent
- Diagnostic coverage claimed per mechanism (range checks, watchdog, lockstep) and applied to dangerous modes
- Single-Point Fault Metric computed near 96.8 percent against the ASIL B target
- Latent Fault Metric computed near 97.3 percent with the ASIL C target as reference
- PMHF rolled up in FIT per hour and compared to the safety goal budget
Unlock the full FMEDA table and PMHF rollup
Who this guide is for
- Hardware safety engineers building their first FMEDA from a bill of materials
- Reliability engineers moving from qualitative FMEA into quantitative ISO 26262-5 analysis
- Reviewers auditing a supplier FMEDA for optimistic coverage or under-counted latent faults
- Engineers whose SPFM or PMHF misses its ASIL target and who need to find out why
Frequently Asked Questions
Common questions about FMEDA: Quantitative Hardware Safety Analysis
Start the course today
A free account unlocks one full concept guide, 3 work product templates, 1 guided process, the Markov simulator, and 5 practice exams per month. The Pro and Expert plans unlock more of the 77-guide library. No credit card required.