This project developed a new reliability prediction framework capable of accommodating recent experience with electronic components (e.g. effects of atmospheric radiation and reduced feature sizes of high density semiconductors), and includes provisions to more accurately address the impacts of the physical failure mechanisms on predicted reliability.
The goal of the project was to assure that updated data models and other information are accessible, useful and widely accepted by suppliers and users of aerospace and high performance electronic systems. The results were based on input and collaboration among a large community of reliability practitioners in all levels of the supply chain.
The effort produced a framework for reliability prediction/assessment based on the results of ongoing research and in-service experience and data addressing
- The impact of atmospheric radiation
- The impact of sub-100 nanometer CMOS technology
The new reliability prediction module addressing the impact of sub-100 nanometer CMOS technology provided a simplified prediction approach and a Weibull to CFR Conversion Method.
The module was given to the Naval Surface Warfare Center (NSWC) Crane to be added to a future update to MIL-HDBK-217.
The project produced a roadmap and evaluation of available reliability prediction methodologies and data sources including their feasibility, availability, effectiveness and long-term supportability.
The roadmap delivered a detailed plan of action cooperatively developed with reliability industry researchers collaborating with NSWC Crane, CALCE, RiAC and AMSAA.
The project studied current modeling practices with recommendations on how safety, software, reliability and human factors can be integrated to provide a holistic systems reliability approach and issued a report on the findings.
Project participants used the roadmap to provide the organization, coordination and oversight to create an integrated reliability prediction methodology and deliver a detailed framework for the reliability modeling approach, including many of the features of the roadmap that are common to all reliability modeling approaches, such as common standards for establishing models, application of models, testing, data collection and validation.
The reliability prediction methodology framework included:
- Establishing New Reliability Models
- Standards for the progress of subprojects
- Typical progression of tasks
- Common rules for engaging and proposing a model
- Checklist for Subproject launch
- Application of Reliability Models
- Common rules for using models
- Calibration
- Levels of detail needed for different applications
- Criteria for modeling environmental effects
- Address complexities in the Natural Environment
- Validation
- Define what it means to be “validated” (versus demonstrated)
- Standards for testing
- How much field data is enough (agree on statistical tests)
- Mechanism for review and update of models
- Ongoing maintenance of models
- Ground rules for periodic updates
- Use of field data
- Electronic based methodology
- Issues to resolve (e.g. configuration control) to achieve an accelerated (over paper publication) but still deliberate process
- Vetting of new contributions
- Processes for updating
- Usage standards, user policy
- Defaults
