Failure, life, mission, and architecture workbench.
Start with the data boundary, then select the model that matches the failure behavior and decision—not the metric that happens to be familiar.
MTBF
Repairable interval between failures.
MaintainabilityMTTR
Active repair duration.
LifeMTTF
Observed non-repairable life.
RateFailure rate
Failures per unit exposure.
ForecastExpected failures
Population workload projection.
MissionReliability probability
Exponential survival.
RequirementReliability to MTBF
Target translation under exponential assumptions.
Life modelWeibull reliability
Age-aware survival probability.
PercentileB10 life
Ten-percent failure life.
Age effectHazard rate
Conditional failure intensity.
ArchitectureSeries system
All required elements succeed.
RedundancyParallel system
At least one path succeeds.
Voting logicK-out-of-N
Required minimum successful paths.
IntrinsicInherent availability
MTBF and MTTR relationship.
ReadinessOperational availability
Includes support delay.
Use probability models with care.
Architecture tools assume matching mission duration and independent paths. Field systems often have shared utilities, common-cause failures, changing duty, or incomplete detection. Document the judgment around the math. Use the RBD guide and Weibull reference when appropriate.