AEC-Q100, Q101, Q200: automotive component qualification
Guide · Automotive qualification
The AEC-Q label has been printed on nearly every automotive bill of materials for thirty years, while no European, US or Japanese regulatory text turns it into a direct legal obligation. This family of documents, issued by the Automotive Electronics Council, describes the qualification tests applied to electronic components installed in vehicles. Three core documents shape the framework: AEC-Q100 for integrated circuits, AEC-Q101 for discrete semiconductors, AEC-Q200 for passive components. This guide presents how they fit together, the temperature grades, the test groups, the relationship with ISO 26262 and with the PPAP process inherited from IATF 16949, plus the common traps encountered during component selection for an automotive project.
Origin and legal status of the AEC framework
Section titled “Origin and legal status of the AEC framework”The Automotive Electronics Council was set up in the early 1990s by three US automakers (Chrysler, Ford, General Motors) and their main semiconductor suppliers. The context was the multiplication of embedded controllers, sensors and electronic actuators, and the lack of a common qualification language across vendors. Each OEM published its own specifications at the time, forcing suppliers to perform redundant and expensive testing.
The consortium released the first edition of AEC-Q100 in 1994, followed by AEC-Q101 (discrete semiconductors) and AEC-Q200 (passives). The documents are now maintained collectively, with broad representation across global semiconductor manufacturers and tier-1 suppliers. Successive revisions have extended scope to new families (Q103 for MEMS, Q104 for multichip modules, Q006 for algorithm-based inference components).
The legal status of AEC-Q documents remains advisory and contractual. No European directive, no US federal regulation, no international vehicle type approval (UN-ECE) cites AEC-Q as a normative requirement. The framework propagates through OEM Customer Specific Requirements (CSR), which mandate the supply of a signed AEC-Q report in their specifications. Absence of a report leads, in practice, to component refusal, without any regulatory text being invoked.
This contractual nature explains two key properties of the framework. AEC-Q qualification is self-declared: the component supplier runs the tests and publishes the report, generally without third-party intervention. And it is non-enforceable in an administrative sense: a dispute over the AEC-Q conformity of a component is resolved between parties through contract, not by a market surveillance authority. For the broader CE marking mechanics, see the CE marking guide.
The AEC document family
Section titled “The AEC document family”| Document | Family covered | Typical examples |
|---|---|---|
| AEC-Q100 | Integrated circuits and actively driven semiconductors | Microcontrollers, ASICs, memories, drivers, transceivers, CMOS sensors |
| AEC-Q101 | Discrete semiconductors | Diodes, bipolar transistors, MOSFETs, IGBTs, rectifiers, TVS |
| AEC-Q200 | Passive components | Resistors, capacitors (ceramic, electrolytic, film), inductors, fuses, resonators, LC filters |
| AEC-Q103 | MEMS sensors | Accelerometers, gyroscopes, pressure sensors, microphones |
| AEC-Q104 | Multichip modules (MCM) | Power modules, SiP, isolated power blocks |
| AEC-Q006 | Algorithm-based inference components (AI/ML) | Neuromorphic accelerators, automotive NPUs |
The split of the family reflects the failure physics of the components in scope. The dominant failure modes of a CMOS integrated circuit (electromigration, NBTI, gate-oxide breakdown) are radically different from those of a multilayer ceramic capacitor (mechanical cracking, ESD, dielectric degradation under bias). Each AEC-Q document therefore defines tests and durations matched to the family at hand, while sharing a common structure organised by test groups.
AEC-Q100 temperature grades
Section titled “AEC-Q100 temperature grades”AEC-Q100 distinguishes four qualification grades based on ambient operating temperature. This parameter is central: it determines where the component is allowed to sit in the vehicle, the accelerated stress profile applied during qualification, and the final cost of the component. The table below summarises the grades defined by Revision H of the document.
| Grade | Ambient temperature | Typical location | Application examples |
|---|---|---|---|
| Grade 0 | -40 to +150 °C | Under-hood close to engine, transmission | Engine ECU, knock sensor, injection control |
| Grade 1 | -40 to +125 °C | Under-hood general, hot luggage compartment | ABS module, automatic transmission ECU, smart alternator |
| Grade 2 | -40 to +105 °C | Sun-exposed cabin, dashboard | Instrument cluster, airbag sensor, heated infotainment ECU |
| Grade 3 | -40 to +85 °C | Standard cabin, mild trunk | Door module, seat control, cabin telematics |
The grade dictates the stress range applied during accelerated tests. A Grade 0 component undergoes HTOL at appreciably higher junction temperatures than a Grade 3 component, which translates into a higher acceleration factor under the Arrhenius equation. Test duration and sample sizes are however similar: it is the severity of conditions, not the cumulative hours, that defines the grade.
In an automotive project, the grade is not chosen at the component level: it follows from the mission profile of the ECU in which the component is integrated. Once the enclosure is placed in the under-hood zone, all critical components in that enclosure must hold a Grade 0 or Grade 1 qualification, unless a documented thermal mitigation device is in place. Grade 2 and Grade 3 fit the cabin. A Grade 3 used under-hood is a specification defect, regardless of the statistical quality of the component.
AEC-Q100 test groups
Section titled “AEC-Q100 test groups”AEC-Q100 organises tests into seven groups, A through G. This structure is mirrored, with adaptations, in AEC-Q101 and AEC-Q200. The table below summarises the role and contents of each group for integrated circuits.
| Group | Object | Main tests | Pass criterion |
|---|---|---|---|
| A | Accelerated environmental stress | HTOL, THB, TC, autoclave (uHAST), PTC, HAST | No parametric or catastrophic defect after stress |
| B | Package integrity | Solderability, thermal flux resistance, mechanical (shock, vibration, drop), visual inspection | Visual and electrical conformity after stress |
| C | Die-level reliability | EM (electromigration), TDDB, HCI, NBTI, metal stress | Extrapolated lifetime above mission target |
| D | Extended electrical verification | Characterisation at limits, parametric distribution across multiple lots | Functional margins on three independent lots |
| E | Defect screening | ESD HBM (Human Body Model), CDM (Charged Device Model), latch-up | ESD/latch-up levels conforming to declared thresholds |
| F | Cavity package integrity | Hermeticity, gross-leak, fine-leak (sealed ceramic packages) | Tightness across pressure and temperature range |
| G | High voltage and high current | SOA tests, power cycling, oxide robustness against transients | No failure within the safe operating area |
Group A acronyms unpacked: HTOL (High Temperature Operating Life), THB (Temperature Humidity Bias), TC (Temperature Cycling), HAST (Highly Accelerated Stress Test, biased humidity under pressure), PTC (Power Temperature Cycling), uHAST (unbiased HAST, equivalent to autoclave).
Each test imposes a minimum sample size, a minimum duration and a number of independent lots (typically three, from three different production weeks). Statistical criteria rely on a reliability demonstration of the LTPD (Lot Tolerance Percent Defective) type, at a defined confidence level. The statistical method itself is not specified as a test but governs how results are read.
For the practical setup of testing and the lab accreditation chain, see the general certification costs guide, which addresses typical budgets for product qualification.
AEC-Q101 and AEC-Q200: specifics of discretes and passives
Section titled “AEC-Q101 and AEC-Q200: specifics of discretes and passives”AEC-Q101 reuses the group structure of AEC-Q100, adapted to discrete semiconductors. Critical tests differ: for a MOSFET, SOA (Safe Operating Area) analysis, power cycling and repetitive avalanche become structuring tests. Temperature grades are also defined (Grade 0 to 3), with profiles aligned on AEC-Q100 but biasing conditions specific to the family.
AEC-Q200 covers passives and structures testing around dielectric and mechanical reliability. For a multilayer ceramic capacitor (MLCC), critical tests include thermomechanical cycling degradation, PCB cracking, insulation resistance after humid stress. For an inductor, the focus is magnetic saturation versus temperature, dielectric strength between turns and mechanical resistance of the core. AEC-Q200 includes sub-revisions for the more demanding families (fuses, crystal resonators, resistive sensors).
The distinction matters during selection: an AEC-Q200-qualified capacitor is not automatically equivalent to its commercial counterpart. Within the same series, the AEC-Q200-qualified part numbers are often restricted to certain dielectrics (X7R, X8R preferred for thermal stability), to specific capacitance and voltage ranges, and to a subset of tolerances. Choosing an MLCC for a Grade 1 under-hood application typically excludes Y5V and Z5U dielectrics.
AEC-Q vs ISO 26262: two distinct logics
Section titled “AEC-Q vs ISO 26262: two distinct logics”A common confusion is to assume that an AEC-Q100-qualified component is fit for a safety-critical application under ISO 26262. It is not. The two frameworks cover complementary but disjoint dimensions of automotive component reliability. The table below summarises their respective scope.
| Dimension | AEC-Q (100/101/200) | ISO 26262 (parts 5, 9, 11) |
|---|---|---|
| Nature | Industrial qualification technical document | International functional safety standard |
| Issued by | Automotive Electronics Council (private consortium) | ISO (international standards body) |
| Object | Physical reliability under stress (lifetime, temperature, humidity, ESD) | Functional safety: control of dangerous failures |
| Metrics | LTPD, raw FIT, parametric margins | ASIL, SPFM, LFM, PMHF, FMEDA |
| Granularity | Single component, production lots | Component embedded in a safety element, up to vehicle level |
| Status | Contractual (OEM CSR) | ISO standard, required via CSR and state of the art |
| Recognition | Global automotive industry | Global automotive industry, adjacent safety-critical sectors |
A microcontroller targeted at a braking function rated ASIL D must satisfy both AEC-Q100 Grade 1 (at minimum) and a complete ISO 26262-11 safety dossier, including FMEDA, calculation of diagnostic coverage metrics, and the documentation of Safety Element Out Of Context (SEooC). Neither condition is sufficient on its own. For the detail of ISO 26262 requirements, see the ISO 26262 automotive functional safety guide.
Conversely, AEC-Q does not address the real-time detection of dangerous failures. An AEC-Q200 capacitor may develop a mechanically induced internal crack and remain in that state for several hours before dielectric breakdown. AEC-Q200 demonstrates that the probability of that mode is low; ISO 26262 requires the system to detect and manage the failure before it becomes dangerous to the driver.
Linkage with IATF 16949 and the PPAP process
Section titled “Linkage with IATF 16949 and the PPAP process”IATF 16949 is the automotive-sector quality management standard, derived from ISO 9001 and published by the International Automotive Task Force. It requires suppliers to operate a management system covering the full cycle, from development through serial production. AEC-Q is not named explicitly, but the standard requires demonstration of reliability of purchased components, which in practice flows through AEC-Q qualification.
The PPAP (Production Part Approval Process), defined by the AIAG manual, is the part approval workflow before serial production launch. It consists of an eighteen-element dossier submitted to the customer for validation. Elements concerned by component qualification typically include:
- Element 5: design records, design FMEA (DFMEA).
- Element 7: process design, process FMEA (PFMEA).
- Element 10: dimensional results.
- Element 11: material and performance test results.
- Element 14: initial process studies (Cpk).
- Element 17: warrant (PSW, Part Submission Warrant).
AEC-Q reports are filed under element 11 (performance testing) and referenced in element 14 when they bear on process stability. Without an AEC-Q report, the PSW generally cannot be signed by the customer. PPAP is not the end of the process: any later change triggers a PCN (Product Change Notification) and, depending on scope, partial or full requalification and a new PPAP.
For suppliers not native to the automotive supply chain (industrial vendors with components reusable in automotive), the IATF + PPAP documentation typically represents a higher organisational effort than AEC-Q qualification itself. See the companion IATF 16949 automotive quality guide for the system requirement detail.
Traceability, counterfeit risk and change notifications
Section titled “Traceability, counterfeit risk and change notifications”The automotive ecosystem requires end-to-end traceability of components installed on vehicles. Three mechanisms converge to this end.
Lot traceability relies on the lot code marked on the package and on the retention of production data (wafer, site, equipment) at the manufacturer for the duration of the vehicle warranty (typically fifteen years). When a field failure is recorded at customer level, the chain must be able to trace from the vehicle back to the production equipment.
PCN management requires the supplier to notify customers in writing of any change likely to affect form, fit, function or reliability. For an AEC-Q-qualified component, a PCN on package, back-end process, factory site, or even a strategic raw material change triggers requalification, partial or full, according to a documented decision matrix. Customer notice is generally six to twelve months, with transition samples made available.
Counterfeit mitigation has been a major topic since the 2020-2022 shortages. Counterfeit components introduced into the automotive channel via the grey market carry a double risk: premature vehicle failure and OEM liability exposure. DPI (Direct Part Identification, 2D Data Matrix laser marking), systematic queries against manufacturer databases, and exclusive procurement through authorised distributors are the operational countermeasures. The certification timeline guide discusses supply lead times and project impact.
Automotive component vs commercial component: what really changes
Section titled “Automotive component vs commercial component: what really changes”Beyond the AEC-Q report itself, automotive qualification of a component drives several observable differences against its commercial or industrial equivalent. The main ones are:
- Selection and binning: AEC-qualified lots are produced on dedicated lines or specific bin selections, retaining parts best centred in the parametric distribution. Commercial components may come from the same wafers but from less demanding bins.
- Parametric stability over life: drift of an AEC-Q100 voltage reference (Vref) is typically lower than its commercial counterpart, at identical technology, because drift has been characterised and bounded.
- ESD and EOS resilience: declared HBM and CDM thresholds are higher on qualified components, and resilience to field EOS (Electrical Overstress) events is documented by the fab.
- Thermal cycling endurance: cycling lifetime is demonstrated over hundreds to thousands of cycles depending on grade, which exceeds commercial requirements by one or two orders of magnitude.
- Declared failure rate (FIT): in-service FIT for a qualified component is typically lower and supported by a statistical dossier. For commercial parts, the published FIT is often indicative.
- Longevity commitment: for automotive components, the supplier commits to a longevity of availability (typically ten to fifteen years), with long-notice end-of-life PCN. Commercial components may be replaced or discontinued within a few years.
These differences do not always translate into a change of internal reference: the same die can be marketed under two part numbers (commercial and automotive) with different post-fabrication binning, a stricter test program and an adapted package.
A consequence often overlooked at the design stage is the gap in field return rates between commercial and automotive grades. The FIT figure published in a commercial datasheet is generally a vendor estimate derived from limited HTOL data, not from a fleet of millions of parts observed in operating conditions. The FIT figure declared in an automotive qualification report, by contrast, draws on extensive field data and survives external audit. For a designer choosing between a commercial part at a fraction of the cost and an automotive-grade equivalent, the cost differential masks a difference of one or two orders of magnitude in observed reliability, particularly during the early infant-mortality window.
A second consequence sits at the boundary with electrical overstress (EOS) events. EOS, by contrast with ESD, refers to dissipation events exceeding the absolute maximum ratings during normal operation, generally caused by induction, ground bounce or improper power sequencing. EOS is not part of AEC-Q test groups as such, yet automotive-grade components typically embed margins on input clamps, body diodes and supply ramps that significantly improve EOS resilience in field conditions. Two parts identical on paper can behave very differently under a real-world load dump or jump-start transient.
AEC-Q006 and qualification of inference components
Section titled “AEC-Q006 and qualification of inference components”The introduction of algorithm-based inference functions in vehicles (perception, sensor fusion, ADAS, assisted driving) has raised a difficulty: the targeted components (NPUs, neuromorphic accelerators, dedicated AI SoCs) exhibit failure modes and software dependencies that the A-to-G groups of AEC-Q100 did not fully cover.
AEC-Q006 is the consortium's recent response. The document, still young at initial release, complements AEC-Q100 for these components by adding tests for:
- Algorithmic stability under thermal stress and ageing (drift of weights and activations).
- Robustness to digital noise induced by ECC memory failures.
- In-service surveillance of accelerators (activation sampling, digital signatures).
AEC-Q006 does not replace ISO 26262. For inference functions integrated into an ASIL chain, the safety analysis remains under ISO 26262 and SOTIF (ISO 21448, Safety Of The Intended Functionality), which addresses hazards not linked to a hardware failure but to inadequate intended behaviour.
Common traps during component selection
Section titled “Common traps during component selection”| Trap | Consequence | Action |
|---|---|---|
| Confusing grade and package (Grade 1 vs Grade 2 inside the same series) | Component under-qualified for the ECU zone | Check the exact AEC suffix on the PN, not the family |
| Treating an AEC-Q100 report as sufficient for ASIL B+ | ISO 26262 non-conformity, product return | Request ISO 26262-11 documentation from the supplier (SEooC) |
| Buying a qualified component via an unauthorised distributor | Counterfeit risk, broken traceability | Mandate sourcing through manufacturer-authorised distributors |
| Ignoring a PCN on factory site change | Unanticipated requalification, PPAP delay | Set up PCN tracking in the PLM/ERP system |
| Selecting a Y5V MLCC for a Grade 1 application | Effective capacitance collapses at -40 °C or +85 °C | Prefer X7R or X8R for under-hood use |
| Skipping Initial Process Studies (PPAP element 14) | Customer PSW refusal | Align part Cpk with customer CSR requirements |
| Treating AEC-Q200 as a homogeneous quality | Non-qualified sub-references within the same series | Verify the exact qualified PN, not the commercial series |
Component selection in an automotive environment remains a discipline in its own right, sitting at the intersection of physical reliability constraints (AEC-Q), functional safety (ISO 26262), system quality (IATF 16949 + PPAP) and vehicle cybersecurity (ISO/SAE 21434). The spilma glossary gathers the key terms (AEC, HTOL, THB, FMEDA, ASIL, PPAP, PCN, FIT) with their reference definitions.
See also
Section titled “See also”- IATF 16949: automotive quality management
- ISO 26262: automotive functional safety
- EV charging: IEC 61851, ISO 15118 and OCPP conformity
- EN 50128 and EN 50657: railway software assurance
Sources & references
- Automotive Electronics Council, official AEC documents portal , AEC www.aecouncil.com/AECDocuments.html
- AEC-Q100 Rev H, stress test qualification for integrated circuits , AEC www.aecouncil.com/Documents/AEC_Q100_Rev_H.pdf
- AEC-Q101 Rev D1, stress test qualification for discrete semiconductors , AEC www.aecouncil.com/Documents/AEC_Q101_Rev_D1.pdf
- AEC-Q200 Rev E, stress test qualification for passive components , AEC www.aecouncil.com/Documents/AEC_Q200_Rev_E.pdf
- ISO 26262, road vehicles, functional safety (all parts) , ISO www.iso.org/standard/68383.html
- AIAG, Production Part Approval Process (PPAP), 4th edition , Automotive Industry Action Group www.aiag.org/quality/automotive-core-tools/ppap