Definition
Change in engineer performance metrics (fuel efficiency, schedule adherence, safety event rate) during manual operation segments compared to a baseline established when ATO/energy management system (EMS) was first deployed or during structured manual-operation assessments.
The Measurement Gap
The Measurement Gap
Railroads deploying energy management systems (EMS) and ATO know these systems improve aggregate fuel economy and schedule performance. What they don't track is what happens to the engineers' manual skills while the automation does the work. When automation fails (sensor outage, system degradation, territory not covered by the model) the engineer takes over manually. If manual proficiency has decayed, degraded-mode performance is worse than anyone planned for.
This gap is well-documented in aviation human factors research. Airline pilots lose hand-flying proficiency under autopilot, which is why airlines mandate periodic manual-flying assessments. Rail has the same dynamic but no equivalent measurement or assessment program. Engineers who operated EMS-equipped locomotives for years may not have manually managed throttle and dynamic brakes on challenging territory in months.
Decision Supported:
Is ATO or the EMS degrading engineers' manual train handling proficiency? Does the railroad need structured manual-operation training to maintain degraded-mode capability? EMPDR provides the longitudinal data to detect skill atrophy before it causes a degraded-mode performance failure.
Formula
Plain Language: Measure engineer performance on manual-operation segments (EMS disengaged, ATO off). Compare to the same engineers' manual-operation performance in the baseline period. A widening gap indicates skill atrophy.
Notation:
EMPDR = (Performance_manual_current − Performance_manual_baseline) / Performance_manual_baseline
Variables:
| Variable | Definition | Unit | Source | Classification |
|---|---|---|---|---|
| Performance_manual_current | Composite score of fuel efficiency, schedule adherence, and safety events during manual operation in the current period | Composite index | Locomotive telemetry | DERIVED |
| Performance_manual_baseline | Same composite score during baseline period (first 6 months of ATO deployment, or structured manual assessment) | Composite index | Historical data | NATIVE (if retained) |
Sign Convention: 0% = no change from baseline. Positive = degradation (worse performance on manual segments). Negative = improvement (rare but possible if manual training programs are effective).
Edge Cases:
- Involuntary manual segments (the EMS disengages due to system failure) may occur on difficult territory. Control for territory difficulty, or use structured assessment windows on standard territory.
- Engineer turnover: new engineers who never operated pre-ATO have no personal baseline. Use fleet average or cohort comparison.
- Minimum sample: 30 manual-operation segments per engineer per assessment period. If manual segments are too rare, structured assessment windows are necessary.
Proposed formula, pending validation pilots.