Overview of Reliability Testing
Reliability testing is a rigorously structured, statistically grounded engineering discipline dedicated to quantifying, predicting, and validating the probability that a product, component, system, or material will perform its intended function without failure under specified conditions for a defined period of time. Unlike simple functional verification or pass/fail quality inspection, reliability testing is inherently probabilistic, time-dependent, and context-sensitive—designed not merely to detect defects, but to model degradation mechanisms, estimate lifetime distributions, identify latent failure modes, and quantify confidence intervals around critical performance metrics such as mean time to failure (MTTF), mean time between failures (MTBF), failure rate (λ), and B10 or B50 life (the time at which 10% or 50% of a population is expected to fail). In the scientific instrumentation and laboratory services ecosystem, reliability testing constitutes a foundational pillar of assurance infrastructure—serving as both a pre-market validation requirement and a post-deployment surveillance mechanism across high-stakes domains including medical device development, aerospace avionics, semiconductor fabrication, automotive electronics, nuclear instrumentation, and pharmaceutical manufacturing equipment.
The strategic significance of reliability testing extends far beyond compliance. In mission-critical applications—such as implantable cardiac pacemakers, radiation therapy control systems, or deep-space telemetry modules—a single unanticipated failure can result in catastrophic human, financial, or environmental consequences. Consequently, reliability testing functions as a quantitative risk mitigation protocol, enabling organizations to translate abstract safety requirements into empirically verifiable design margins. From a business perspective, it directly influences total cost of ownership (TCO) by reducing warranty liabilities, minimizing field recalls, optimizing preventive maintenance schedules, and enhancing brand reputation through demonstrable robustness. Moreover, in regulated environments governed by agencies such as the U.S. Food and Drug Administration (FDA), European Medicines Agency (EMA), or International Electrotechnical Commission (IEC), documented reliability evidence is not optional—it is a statutory prerequisite for regulatory submission packages (e.g., FDA 510(k), De Novo, or PMA dossiers) and forms an integral component of Design History Files (DHF) and Quality System Records (QSR).
Within the broader taxonomy of laboratory services, reliability testing occupies a distinct epistemological niche: it bridges theoretical reliability physics with empirical metrology. While disciplines like materials science or thermal analysis focus on static property characterization, reliability testing interrogates dynamic behavior—how stress interactions (thermal, mechanical, electrical, chemical, radiative) accelerate degradation pathways over time. It therefore demands specialized instrumentation capable of precise, synchronized, multi-parameter stimulus application; real-time parametric monitoring at microsecond resolution; fault injection with controlled fidelity; and statistical data acquisition across hundreds—or even thousands—of concurrent test units. Critically, reliability testing is never conducted in isolation; it is embedded within a systems engineering framework that includes failure mode and effects analysis (FMEA), fault tree analysis (FTA), accelerated life testing (ALT) modeling, Weibull analysis, and reliability growth tracking via Duane or Crow-AMSAA methodologies. As such, the instruments deployed in this domain are not standalone tools but integrated nodes within a cyber-physical test architecture—interfacing with enterprise quality management systems (QMS), product lifecycle management (PLM) platforms, and cloud-based analytics engines to enable closed-loop design feedback.
The scope of reliability testing encompasses both intrinsic and extrinsic dimensions. Intrinsic reliability addresses inherent design and manufacturing variability—arising from material microstructure heterogeneity, process-induced residual stresses, interfacial delamination in multilayer substrates, or dopant segregation in semiconductor junctions. Extrinsic reliability, conversely, pertains to environmental and operational stressors imposed during use: temperature cycling inducing solder joint fatigue, humidity-driven electrochemical migration, voltage overstress causing gate oxide breakdown, or vibration spectra replicating road-induced resonance in automotive ECUs. Modern reliability paradigms increasingly emphasize “physics-of-failure” (PoF) approaches—where test protocols are explicitly derived from first-principles degradation models (e.g., Black’s equation for electromigration, Coffin-Manson for thermal fatigue, Eyring’s model for chemical reaction kinetics)—rather than purely empirical, statistically driven stress screening. This paradigm shift has fundamentally redefined instrument requirements: today’s reliability test systems must support not only high-fidelity stimulus replication but also in situ, non-destructive diagnostics—including real-time impedance spectroscopy, thermography-based hot-spot localization, acoustic emission sensing for crack propagation detection, and time-resolved electroluminescence imaging for LED degradation mapping.
Furthermore, reliability testing has evolved from a late-stage, “test-to-failure” activity into an integral, concurrent engineering practice. Contemporary development workflows embed reliability assessment throughout the design cycle—from early concept validation using virtual reliability simulation (e.g., ANSYS Sherlock, Synopsys SaberRD) to design-for-reliability (DFR) workshops, prototype HALT (highly accelerated life testing), production lot sampling per MIL-STD-781 or ISO 16249, and field return analysis using Weibull++ or JMP Reliability platforms. This lifecycle integration necessitates instrument interoperability: chambers must export timestamped sensor logs in IEEE 1584-compliant formats; power supplies must support programmable transient profiles synchronized to digital I/O triggers; data acquisition systems must ingest mixed-signal streams (analog voltage/current, digital bus traffic, thermal IR frames) with sub-microsecond jitter alignment. Ultimately, reliability testing serves as the empirical bedrock upon which scientific credibility, regulatory legitimacy, and commercial trust are jointly constructed—transforming probabilistic uncertainty into actionable engineering intelligence.
Key Sub-categories & Core Technologies
The instrumentation landscape for reliability testing is highly stratified, reflecting the multidimensional nature of failure physics and the stringent metrological demands of modern validation protocols. Rather than constituting a monolithic category, reliability test equipment comprises several interdependent sub-systems—each engineered to address specific stress domains, failure mechanisms, and analytical objectives. These sub-categories are not merely differentiated by form factor or price point; they embody distinct physical principles, calibration traceability hierarchies, and domain-specific validation requirements. A comprehensive understanding of these technologies is essential for designing statistically valid test plans and interpreting results with appropriate uncertainty budgets.
Environmental Stress Screening (ESS) & Accelerated Life Test (ALT) Chambers
At the core of physical reliability validation lie environmental stress screening (ESS) and accelerated life test (ALT) chambers—precision-engineered enclosures capable of imposing controlled, repeatable, and metrologically traceable combinations of thermal, humidity, pressure, and mechanical stimuli. Modern ESS/ALT chambers transcend basic temperature-humidity cycling; they integrate multi-zone thermal gradient control (±0.1°C uniformity across 1 m³ volume), rapid thermal transients (up to 60°C/min ramp rates with overshoot < ±0.3°C), dew-point controlled humidity modulation (5–95% RH at ±0.5% RH accuracy), and combined environmental profiles (e.g., temperature-humidity-vibration synergy per MIL-STD-810H Method 514.7). High-end systems incorporate inert atmosphere purge capabilities (N₂, Ar, or synthetic air with O₂ < 1 ppm) for oxidation-sensitive components and vacuum-assisted moisture desorption protocols aligned with JEDEC J-STD-020 moisture sensitivity level (MSL) standards.
Technologically, these chambers rely on cascaded refrigeration circuits (dual-stage or cascade CO₂/ammonia systems), PID-controlled resistive heating elements with distributed thermal mass compensation, ultrasonic humidification with dissolved solids filtration, and piezoelectric or electromagnetic shakers integrated into chamber floors. Calibration is performed per ISO/IEC 17025-accredited procedures using NIST-traceable platinum resistance thermometers (PRTs), chilled-mirror hygrometers, and laser Doppler vibrometers. Data integrity is ensured via redundant sensor arrays, real-time deviation alarms, and encrypted audit trails compliant with 21 CFR Part 11. Notably, chamber selection requires rigorous evaluation of thermal inertia characteristics—critical for simulating transient thermal shocks experienced during reflow soldering or power-on surges—and airflow dynamics, as laminar versus turbulent flow patterns significantly influence convective heat transfer coefficients and thus failure acceleration factors.
Highly Accelerated Life Testing (HALT) & Highly Accelerated Stress Screening (HASS) Systems
HALT and HASS represent a paradigmatic departure from traditional qualification testing. Whereas conventional ALT assumes known failure modes and applies statistically derived stress levels, HALT is a discovery-oriented, iterative stress profiling methodology designed to expose design weaknesses and operational limits—not to validate against specifications, but to provoke failures deliberately and rapidly. HALT systems integrate six-degree-of-freedom (6DOF) multi-axis vibration exciters (capable of random, sine, and shock profiles up to 100 g RMS), ultra-rapid thermal chambers (−100°C to +200°C, 60°C/min ramps), and programmable power supplies—all operating synchronously under closed-loop control. The instrumentation stack includes high-bandwidth accelerometers (up to 50 kHz bandwidth), thermocouple multiplexers with cold-junction compensation, and real-time signal analyzers capable of performing Fast Fourier Transform (FFT), envelope spectrum analysis, and modal identification during stress application.
HASS, the production-line counterpart to HALT, leverages the operational limits discovered during HALT to establish tightly constrained, high-throughput screen profiles that eliminate infant mortality without inducing wear-out. HASS systems emphasize throughput optimization: automated handler interfaces (pick-and-place robotics), parallel test station architectures (supporting 256+ DUTs simultaneously), and machine vision-based optical inspection for solder joint cracking or package warpage detection. Core technologies include resonant frequency tracking algorithms that dynamically adjust vibration spectra to maintain constant displacement amplitude across aging DUTs, and adaptive thermal profiling that modulates ramp rates based on real-time thermal impedance measurements. Calibration of HALT/HASS systems follows ASTM E1820 for fracture toughness validation and ISO 10816-3 for vibration severity classification—ensuring that stress intensities are quantifiably linked to mechanical damage accumulation models.
Electrical Stress & Parametric Monitoring Systems
Electrical reliability testing focuses on degradation induced by voltage, current, power, and electromagnetic interference (EMI) stressors. Instruments in this sub-category include ultra-stable, low-noise DC power supplies (< 10 µVrms noise, 1 ppm/h stability), high-voltage insulation testers (up to 10 kV DC with leakage current resolution of 1 fA), electrostatic discharge (ESD) simulators compliant with IEC 61000-4-2 (Human Body Model, Machine Model, Charged Device Model), and surge immunity test systems per IEC 61000-4-5. Advanced systems integrate real-time parametric monitoring: four-quadrant source-measure units (SMUs) capable of simultaneous sourcing and measuring voltage/current with 16-bit resolution and 1 MS/s sampling, transient digitizers with 12-bit vertical resolution and 1 GS/s sampling for capturing nanosecond-scale voltage spikes, and bit-error-rate testers (BERTs) for high-speed serial link reliability assessment (e.g., PCIe Gen5, USB4, SATA).
A particularly sophisticated subset involves bias-temperature instability (BTI) and time-dependent dielectric breakdown (TDDB) test platforms used extensively in semiconductor process development. These systems apply precisely controlled gate voltages (±10 V, 100 µV resolution) while monitoring threshold voltage shifts (ΔVth) and gate leakage currents (Ig) over timescales ranging from milliseconds to years—extrapolated via Arrhenius and E-model acceleration. Instrumentation includes cryogenic probe stations (4K–400K operation), ultra-low-current picoammeters (femtoamp sensitivity), and lock-in amplifiers for extracting small-signal conductance changes amidst thermal noise. Calibration traceability extends to NIST’s Josephson voltage standard and quantum Hall resistance standard, ensuring metrological integrity for reliability predictions influencing multi-billion-dollar fab investments.
Mechanical Fatigue & Structural Integrity Testers
Mechanical reliability assessment targets failure modes rooted in cyclic loading, creep, impact, and interfacial adhesion. Key instruments include servo-hydraulic and electrodynamic fatigue test systems (load capacity 1 N–2,500 kN, frequency range 0.001–100 Hz), thermomechanical analyzers (TMA) for coefficient of thermal expansion (CTE) mismatch quantification, dynamic mechanical analyzers (DMA) for viscoelastic property mapping, and nanoindentation systems for localized hardness and modulus profiling at sub-micron scales. For microelectromechanical systems (MEMS), specialized resonant frequency trackers monitor stiffness degradation in cantilevers subjected to humidity cycling, while shear-lag bond testers quantify die-attach integrity under thermal cycling per IPC-9708 standards.
Modern mechanical reliability testers integrate digital image correlation (DIC) for full-field strain mapping, acoustic emission sensors for detecting microcrack initiation, and synchrotron-compatible stages for in situ X-ray diffraction during load application. Calibration adheres to ISO 7500-1 for static force measurement and ISO 4965 for dynamic force verification, with uncertainty budgets accounting for cross-axis sensitivity, phase lag compensation, and environmental thermal drift. Notably, the emergence of additive manufacturing has spurred demand for residual stress analyzers—using neutron diffraction or X-ray residual stress mapping—to correlate build parameters with long-term structural reliability in safety-critical aerospace components.
Corrosion, Chemical, & Electrochemical Reliability Analyzers
For products exposed to aggressive chemical environments—implantable medical devices, marine electronics, battery enclosures, or catalytic converters—corrosion and electrochemical reliability testing is indispensable. Instruments include potentiostats/galvanostats with ±10 A current range and 100 fA resolution for electrochemical impedance spectroscopy (EIS), zero-resistance ammeters (ZRA) for galvanic corrosion current measurement, salt spray chambers compliant with ASTM B117 and ISO 9227 (including cyclic corrosion variants per GMW14872), and environmental SEMs equipped with in situ electrochemical cells. Advanced platforms incorporate microfluidic corrosion cells for localized pitting studies, Raman spectroelectrochemistry modules for real-time passive film composition analysis, and scanning vibrating electrode technique (SVET) systems for mapping ionic current densities above corroding surfaces with micron-scale resolution.
Calibration protocols follow ASTM G59 for potentiostat accuracy verification and ISO 16773 for EIS measurement uncertainty quantification. Critical considerations include reference electrode stability (Ag/AgCl, saturated calomel), solution resistance compensation algorithms, and frequency response analysis to distinguish true electrochemical processes from capacitive artifacts. In battery reliability testing, differential voltage analysis (DVA) and incremental capacity analysis (ICA) systems integrate with cycler hardware to detect lithium plating onset, SEI growth kinetics, and cathode structural degradation—parameters directly linked to thermal runaway propensity and calendar life prediction.
Optical, Thermal, & Non-Destructive Evaluation (NDE) Diagnostics
As reliability testing shifts toward prognostics and health management (PHM), in situ, non-invasive diagnostics have become central. Thermal imaging systems—microbolometer-based (uncooled) and quantum-well infrared photodetector (QWIP)-based (cooled)—provide spatially resolved temperature maps with NETD < 20 mK and frame rates > 1,000 Hz, enabling hotspot detection during power cycling. Lock-in thermography (LIT) systems modulate electrical excitation at specific frequencies and extract phase-resolved thermal responses to isolate subsurface defects (delaminations, voids) invisible to conventional IR. Similarly, acoustic microscopy (SAM) uses focused ultrasound (15–200 MHz) to generate C-scan images revealing bond line integrity in stacked-die packages, while terahertz time-domain spectroscopy (THz-TDS) characterizes coating thickness uniformity and moisture ingress in pharmaceutical blister packs.
Optical coherence tomography (OCT) systems now achieve axial resolutions < 5 µm for inspecting MEMS mirror actuation dynamics, and hyperspectral imaging platforms correlate spectral signatures with polymer oxidation states during UV aging tests. All NDE instruments require rigorous geometric and radiometric calibration: NIST-traceable blackbody sources for thermal cameras, calibrated step wedges and line-pair targets for resolution verification, and certified reference materials (CRMs) for spectral response validation. Integration with reliability test chambers demands electromagnetic compatibility (EMC) hardening and fiber-optic data transmission to avoid interference—particularly critical for THz and OCT systems operating at GHz–THz frequencies.
Major Applications & Industry Standards
Reliability testing is not a generic capability but a domain-specific competency—its methodologies, instrumentation configurations, and acceptance criteria are meticulously tailored to the unique failure physics, regulatory expectations, and operational risk profiles of each end-use industry. Understanding these contextual constraints is paramount: deploying aerospace-grade vibration profiles on consumer IoT devices introduces unnecessary cost and schedule overhead, while applying automotive thermal cycling standards to medical implants fails to address biocompatibility-related degradation pathways. This section details the principal industrial sectors leveraging reliability testing, their defining application challenges, and the normative frameworks governing test execution and reporting.
Aerospace & Defense
In aerospace and defense, reliability is synonymous with mission assurance. Components aboard commercial aircraft (FAA Part 25-certified), military platforms (MIL-STD-810G/H), or spacecraft (ECSS-Q-ST-30C) must sustain functionality under extreme thermal gradients (−65°C to +125°C), high-altitude vacuum, ionizing radiation (total ionizing dose > 100 krad(Si)), and multi-axis random vibration spectra exceeding 0.1 g²/Hz. Key applications include avionics box qualification (DO-160G Section 22 for lightning-induced transient susceptibility), satellite solar array deployment mechanism endurance testing, and turbine blade thermal barrier coating spallation assessment. Instrumentation must meet stringent electromagnetic compatibility (EMC) requirements (MIL-STD-461G) and operate within classified environments requiring TEMPEST shielding.
Standards governing aerospace reliability are among the most rigorous globally. RTCA DO-160G defines environmental test conditions for airborne equipment, with Section 25 (Temperature Variation) mandating 100+ cycles between extremes with dwell times calibrated to atmospheric lapse rates. MIL-STD-781H prescribes reliability prediction methodologies (e.g., MIL-HDBK-217F, though increasingly supplemented by physics-of-failure models), while SAE ARP4754A and ARP4761 mandate systematic safety assessment integrating reliability data into fault tree analyses. Crucially, aerospace reliability testing requires full configuration item (FCI) testing—meaning assemblies must be tested in flight-representative configurations, including harness routing, grounding schemes, and thermal interface materials—not just bare PCBs. Data reporting follows AS9100 Rev D requirements for traceability, with all calibration certificates, raw sensor logs, and failure root cause analyses archived for the product’s entire service life (often 30+ years).
Medical Devices & Diagnostics
Medical device reliability intersects clinical safety, regulatory compliance, and ethical imperatives. The FDA’s Guidance on General Principles of Software Validation and the EU MDR Annex I General Safety and Performance Requirements (GSPR) explicitly require manufacturers to demonstrate reliability for software-as-a-medical-device (SaMD) and hardware components affecting diagnostic accuracy or therapeutic delivery. Applications span implantable neurostimulators (ISO 14708-3 for active implantable medical devices), infusion pumps (IEC 60601-2-24 for electromagnetic compatibility and mechanical durability), and next-generation sequencing (NGS) instruments where reagent stability and optical path contamination directly impact variant calling error rates. Failure modes here are often subtle: gradual drift in photodiode responsivity compromising fluorescence quantification, or microchannel fouling in lab-on-a-chip devices altering laminar flow profiles.
Regulatory frameworks impose hierarchical reliability requirements. ISO 14971:2019 mandates risk management processes where reliability data feeds directly into severity-probability matrices, while IEC 62304:2015 governs software lifecycle reliability through unit testing coverage metrics (>90% MC/DC for Class C software). For sterilizable devices, ISO 11135 (ethylene oxide) and ISO 11137 (gamma irradiation) specify reliability validation of packaging integrity post-sterilization. Notably, the FDA’s “Total Product Life Cycle” (TPLC) approach requires post-market reliability surveillance—leveraging field failure databases (MAUDE), Bayesian updating of prior reliability estimates, and proactive recall triggers based on Weibull shape parameter shifts. Instrumentation used must comply with IEC 61010-1 for electrical safety and undergo biocompatibility testing per ISO 10993 when in contact with biological samples.
Automotive Electronics & ADAS Systems
The automotive industry’s transition to electric vehicles (EVs), autonomous driving (SAE Level 3–5), and vehicle-to-everything (V2X) connectivity has exponentially increased electronic content per vehicle—from ~100 ECUs in 2010 to >3,000 in modern platforms—making reliability testing a systemic priority. Key challenges include wide temperature operation (−40°C to +125°C ambient, +150°C junction), high-voltage battery management system (BMS) isolation integrity (>1 kV DC), electromagnetic coexistence in dense RF environments (5G, DSRC, UWB), and functional safety compliance per ISO 26262 ASIL-D. Applications range from radar sensor module thermal cycling (AEC-Q200 Grade 0 qualification), infotainment SoC burn-in (JEDEC JESD22-A108F), to lidar optical window abrasion resistance testing per ISO 16232-C.
Automotive reliability standards are harmonized globally yet technically demanding. AEC-Q200 (passive components) and AEC-Q100 (integrated circuits) define stress test profiles with tighter tolerances than generic JEDEC standards—for example, requiring 1,000-hour high-temperature operating life (HTOL) tests at Tj = 125°C with real-time parametric monitoring every 100 hours. ISO 16750 series specifies environmental stresses (electrical loads, mechanical vibration, chemical exposure), while ISO 26262 mandates reliability evidence for hardware elements supporting safety goals, including FMEDA (failure modes, effects, and diagnostic analysis) reports validated against actual field failure data. Recent trends emphasize cybersecurity reliability: ISO/SAE 21434 requires threat analysis and risk assessment (TARA) integrated with hardware reliability models to evaluate attack surface resilience—demanding instruments capable of injecting controlled fault injections (e.g., clock glitching, voltage faulting) while monitoring cryptographic key extraction success rates.
Semiconductors & Microelectronics
Semiconductor reliability testing operates at the atomic scale, where failure mechanisms manifest as single-event upsets (SEUs), hot-carrier injection (HCI), negative-bias temperature instability (NBTI), or time-dependent dielectric breakdown (TDDB). Foundries and OSATs deploy massively parallel reliability test systems—often custom-built—to screen wafers and packaged devices across thousands of test sites simultaneously. Applications include qualification of 3nm-node logic transistors (requiring TDDB testing at 1.2× nominal Vdd for 1,000 hours), GaN power HEMTs for EV inverters (high-temperature reverse bias testing per JEDEC JEP180), and DRAM retention time validation under elevated temperature/voltage stress.
Industry standards are exceptionally granular. JEDEC JESD22 series defines test methods: JESD22-A104 (temperature cycling), JESD22-A108 (HTOL), JESD22-A110 (mechanical shock), and JESD22-A118 (electrostatic discharge). The JEDEC Solid State Technology Association maintains a reliability qualification database (RQDB) where failure data from member companies informs updated test conditions. Crucially, semiconductor reliability relies on acceleration models with strict validity boundaries: the Eyring model for chemical reactions, Arrhenius for thermally activated processes, and inverse power law for voltage-dependent mechanisms. Instrumentation must therefore provide metrologically defensible acceleration factor calculations—requiring traceable temperature sensors at the die surface (not ambient chamber air), calibrated voltage sources with ppm-level accuracy, and statistical process control (SPC) integration for real-time outlier detection across wafer lots.
Energy & Industrial Equipment
Renewable energy infrastructure—offshore wind turbines, utility-scale solar farms, grid-scale battery storage—demands unprecedented reliability due to inaccessibility and high replacement costs. Wind turbine pitch control systems undergo 10-million-cycle actuator testing per IEC 61400-23, while solar inverter electrolytic capacitors are qualified per IEC 61727 for 25-year field life extrapolation. Grid-scale lithium-ion battery racks require UL 1973 and IEC 62619 certification, involving sequential thermal runaway propagation testing, mechanical crush simulations, and fire containment validation. Industrial robotics reliability focuses on harmonic drive gear fatigue (ISO 9409-1), encoder resolution drift under continuous motion, and IP67-rated connector mating durability (>5,000 cycles).
Standards here emphasize system-level integration. IEC 61850 governs reliability of substation automation communications, requiring deterministic latency bounds under electromagnetic disturbances per IEC 61000-4-30. IEEE 1680.2 addresses environmental and reliability criteria for electronic displays, while NEMA MG-1 sets motor insulation class requirements linked to thermal aging models. A critical trend is digital twin-enabled reliability: Siemens Desigo CC or Schneider EcoStruxure platforms ingest real-time sensor data from fielded equipment to update physics-based degradation models, enabling predictive maintenance scheduling that minimizes unplanned downtime. This necessitates reliability test instruments with OPC UA and MQTT connectivity for seamless data federation into industrial IoT ecosystems.
Technological Evolution & History
The historical trajectory of reliability testing instrumentation reflects a profound evolution—from rudimentary empirical observation to a mathematically rigorous, metrologically anchored engineering science. Its development is inextricably linked to technological inflection points: the advent of vacuum tubes, the transistor revolution, the integrated circuit era, and the rise of complex cyber-physical systems. Each epoch introduced new failure mechanisms, demanded higher precision measurement, and catalyzed instrument innovation—transforming reliability from an afterthought into a first-class design constraint.
Foundational Era (1940s–1960s): Empirical Qualification & Military Imperatives
Reliability testing emerged formally during World War II, driven by alarming failure rates in vacuum tube-based radar and communication systems. The U.S. Department of Defense commissioned the first systematic study, culminating in the 1949 “Reliability Handbook” published by the Advisory Group on Reliability of Electronic Equipment (
