Skip to main content
Reliability

Reliability Engineering

Reliatic's reliability module converts raw failure event data into quantitative metrics, Weibull models, and RCM maintenance strategies. Every calculation is traceable to source records and updates in real time as new failure events are logged.

Core Reliability Metrics

All metrics are calculated from structured failure event records. Inputs come from closed work orders, inspection findings, and operator-logged failure events—not from manual spreadsheet entries.

MTBFMean Time Between Failures
Total Operating Time ÷ Number of Failures

Primary metric for repairable systems. Reliatic computes MTBF per asset class, asset tag, and failure mode. Updated automatically on each failure event closure.

MTTFMean Time To Failure
Total Operating Time ÷ Number of Failed Items

For non-repairable components. Used in spare parts analysis and replacement interval optimization. Sourced from manufacturer data or field failure history.

MTTRMean Time To Repair
Total Repair Time ÷ Number of Repairs

Repair efficiency metric. Reliatic captures repair start/end timestamps on work orders and calculates MTTR per asset type and maintenance team.

AAvailability
MTBF ÷ (MTBF + MTTR)

Operational availability derived from live MTBF and MTTR data. Displayed on the Reliability Dashboard and included in monthly KPI reports.

Weibull Analysis

Reliatic fits a two-parameter Weibull distribution to failure time data for each asset class with sufficient history (minimum 5 failure events). The shape parameter β characterizes the failure rate trend; the scale parameter η represents the characteristic life at which 63.2% of the population has failed.

Weibull fit — V-2041 HP Separator (seal population)
asset_class: Centrifugal_Pump_Seal
sample_size: 23 failure events
shape_β: 2.4        // wear-out regime
scale_η: 8,200 hrs  // characteristic life
B10_life: 3,640 hrs // 10% will fail by this time
B50_life: 7,720 hrs // median failure time
strategy_recommendation: TIME_DIRECTED
suggested_interval: 6,500 hrs (0.79 × η)
β < 1
Infant Mortality

Decreasing failure rate. Indicates design defects, poor installation, or substandard components.

Strategy: Review installation procedures, incoming QC, and burn-in testing.

β = 1
Random Failures

Constant failure rate. Exponential distribution. Failures occur randomly, independent of age.

Strategy: PM strategy will not reduce failures. Focus on redundancy and fast detection.

β > 1
Wear-Out

Increasing failure rate with age. Classic wear, corrosion, or fatigue-driven degradation.

Strategy: Time-directed PM is effective. Optimize replacement interval from Weibull characteristic life η.

Reliability Centered Maintenance (RCM)

RCM analysis in Reliatic links each failure mode from the FMEA to an optimized maintenance task type. The task type is recommended based on the failure mode's β value, detectability, and consequence severity. All recommendations require engineer sign-off before being added to the PM schedule.

TDTime-Directed

Performed at fixed intervals regardless of condition. Appropriate when β > 1 and replacement interval is known.

CDCondition-Directed

Triggered by measurement crossing a threshold (wall thickness, vibration, temperature). Appropriate when degradation is monitorable.

FFFailure-Finding

Verifies hidden functions (standby equipment, safety devices) are still operable. Required for protective devices under PSM.

RTFRun-to-Failure

Economically justified when failure consequence is low and restoration cost is less than prevention cost. Must be explicitly accepted in the FMEA.

Run-to-Failure discipline: RTF must be explicitly justified in the FMEA with a completed consequence evaluation. The platform will not allow RTF selection for assets where the failure consequence category is Safety or Environmental.

Failure Event Recording

All MTBF, MTTF, and Weibull calculations depend on structured failure event records. Each event must capture sufficient data for statistical analysis—free-text descriptions alone are insufficient.

Required failure event fields
{
  asset_id:        "V-2041",
  failure_date:    "2026-01-14T06:32:00Z",
  failure_mode:    "SEAL_LEAK",             // from FMEA taxonomy
  failure_cause:   "CAVITATION",            // root cause code
  detection_method: "OPERATOR_OBSERVATION",
  time_to_repair:  420,                     // minutes
  parts_replaced:  ["SEAL_KIT_P/N-44X"],
  rca_required:    true,                    // auto-set if MTBF threshold breached
  rca_id:          "RCA-2026-0041"
}

Predictive Maintenance Triggers

Reliatic can generate governance events automatically when reliability data crosses configured thresholds. These triggers connect the reliability module directly to the Reliability-to-Action Loop.

Trigger
Corrosion Rate Exceeded
Source
Thickness reading analysis
Platform Outcome
Inspection interval reduced, RBI score recalculated
Trigger
MTBF Trend Declining
Source
Failure event history
Platform Outcome
RCM strategy review queued, PM interval candidate flagged
Trigger
Remaining Life < 20%
Source
Wall loss + corrosion rate
Platform Outcome
Risk event created, fitness-for-service assessment required
Trigger
Overdue Failure Finding Task
Source
PM scheduler
Platform Outcome
Governance event fired, asset compliance state → OVERDUE
Trigger
Weibull β Change Detected
Source
Statistical re-fit on new data
Platform Outcome
Maintenance strategy recommendation updated
Reliability Engineering — Reliatic Documentation — Reliatic