Reliability engineering - Operational Management and Standards

Reliability Operational Assessment Introduction Reliability operational assessment is the process of monitoring, testing, and analyzing how well systems perform in real-world conditions after they've been deployed. This is fundamentally different from predicting reliability before a system is built—instead, it focuses on collecting actual failure data, understanding why failures occur, and continuously improving the system based on what we learn from the field. This approach ensures that reliability issues are caught and corrected during the system's operational life, not just during design and testing. Monitoring and Data Collection Effective operational reliability assessment begins with systematic monitoring of critical system parameters. During the fault tree analysis phase of reliability design, engineers identify which parameters are most likely to indicate problems. Once the system is deployed, both electronic surveillance (such as sensor data and automatic alerts) and visual inspection are used to track these parameters. Different systems require different data collection approaches. For vehicles, failure logs recorded by onboard diagnostic systems provide detailed information. For consumer products, manufacturers track return rates and failure counts. Systems that sit dormant or on standby—such as emergency backup equipment or stored military assets—require a formal surveillance program where random samples are periodically inspected and tested to detect problems before they become critical. A crucial but often overlooked aspect of data collection is monitoring reliability after any changes to the system. Whenever field upgrades are applied, recall repairs are made, or system modifications are implemented, additional reliability testing must follow. Small changes that seem inconsequential can sometimes have unexpected effects on reliability. Reliability Testing and Modifications No reliability program can anticipate every possible failure mode before a system is deployed. This is especially true when human interaction is involved—users find creative ways to break systems that engineers never imagined. Therefore, failures will inevitably occur, and this is not a sign of failure; it's simply part of operational reality. The key to continuous improvement is systematic root cause analysis. When a failure occurs, reliability engineers investigate not just what failed, but why it failed. This investigation traces causal relationships and identifies the underlying conditions that led to the failure. Once the root cause is understood, corrective actions can be implemented to prevent similar failures. All system failures and the corrective actions taken must be formally reported to the reliability engineering organization. This creates a feedback loop where field experience directly informs design and manufacturing improvements. Failure Reporting, Analysis, and Corrective Action System (FRACAS) The most common tool for operational reliability assessment is the Failure Reporting, Analysis, and Corrective Action System (FRACAS). This is a structured database and workflow system designed specifically to manage reliability in operational systems. Here's how FRACAS works: Failure reporting: When a failure or incident occurs in the field, it is reported into a centralized system with all relevant details (what failed, when, where, how, etc.) Analysis: The system organizes and stores all failure data, creating a comprehensive repository Corrective action: Once failures are analyzed and root causes identified, corrective actions are tracked through completion Statistical analysis: The accumulated data enables statistical analysis that reveals reliability, safety, and quality metrics For FRACAS to be effective, organizations must adopt a single, easy-to-use system across all products and systems. When different teams use different reporting systems, data becomes fragmented and the statistical power of the analysis is lost. Common Outputs from a FRACAS System Once failure data has been collected in a FRACAS system, various reliability metrics can be calculated and analyzed: Field Mean Time Between Failures (MTBF) represents the average time a system operates before experiencing a failure. Unlike MTBF calculated during design, field MTBF reflects real-world operating conditions and actual failure rates. Mean Time To Repair (MTTR) measures how quickly the system can be restored to operation after a failure occurs. This metric is critical because a system that fails frequently but repairs quickly may be more acceptable than a system that fails rarely but takes forever to repair. Spare parts consumption rates track how many replacement parts are being used in the field. Unexpectedly high consumption rates for certain parts can indicate design weaknesses or manufacturing quality issues. Reliability growth trends show whether system reliability is improving, declining, or remaining stable over time. When corrective actions are effective, you should see reliability improving as field experience accumulates. Distribution of failures breaks down where and why failures are occurring—organized by failure type, location in the system, part number, serial number, or symptom. This distribution analysis often reveals patterns that point to systemic issues rather than random failures. Limitations of Using Historical Data While FRACAS provides valuable operational data, it's important to understand the limitations of using past failure data to predict reliability of new systems. Context matters enormously. Reliability depends not just on the system itself, but on how and where it is used. A system with excellent reliability in one environment may fail frequently in a different climate, geographic region, or operating pattern. Historical data from similar systems is useful for developing estimates, but it should never be treated as definitive for a new system in a new context. Additionally, small design or manufacturing changes can significantly affect reliability outcomes. A component supplier might change materials to reduce costs, a manufacturing process might be slightly modified for efficiency, or a design might be slightly simplified. Any of these seemingly minor changes can alter failure modes in unpredictable ways. This is why field data from the old version cannot be directly applied to predict reliability of the new version. Organizational Context for Reliability Engineering <extrainfo> How Reliability Engineering Organizations are Structured The way reliability engineering is organized depends on project size and criticality. Small, non-critical projects may handle reliability informally without a dedicated organization or person. However, larger projects and safety-critical systems typically establish a formal reliability function, often placed within a product assurance organization, systems engineering group, or specialty engineering department. A key advantage of creating an independent reliability organization is that it can be insulated from budget and schedule pressures. When reliability engineers report directly to the project manager, there's sometimes pressure to downplay reliability concerns if they might delay a release. An independent organization can more objectively assess reliability issues without such conflicts of interest. Because reliability decisions are most impactful during early design phases, reliability engineers frequently serve as members of integrated product teams (IPTs) that guide the overall design process. This ensures that reliability is built in from the start rather than added as an afterthought. </extrainfo> <extrainfo> Professional Development in Reliability Engineering Licensing Requirements: Reliability engineers working on systems that directly affect public safety (aircraft, medical devices, nuclear systems, etc.) must be licensed as Professional Engineers (PE) by their state or province, just like structural engineers or electrical engineers. However, not all reliability practitioners need licensure—those in non-safety-critical roles or supporting positions may not require PE certification, though relevant credentials are typically expected for advancing in the field. Career Opportunities: Reliability engineers are essential in aerospace, defense, automotive, medical devices, energy, and any industry where system failure could jeopardize safety or mission success. These industries actively recruit reliability engineers because the cost of failures is so high—a failed spacecraft cannot be recalled, and a failed medical device can cause injury. International Standards and Methodologies Several international standards guide reliability analysis. The French FIDES methodology (UTE-C 80-811) uses physics-of-failure principles combined with test data and field returns to predict reliability. The RDF2000 standard (UTE-C 80-810) is another French methodology based on decades of telecommunications industry experience. </extrainfo>