🎓 Lesson 5
D3
Fatigue Risk Modeling for ROC Shift Schedules
Fatigue risk modeling predicts how tired ROC operators will get during their shifts so we can design schedules that keep them alert and safe.
🎯 Learning Objectives
- ✓ Calculate fatigue risk scores for ROC shift patterns using the SAFTE-FAST hybrid algorithm
- ✓ Design compliant ROC shift schedules that maintain average alertness above 75% over 28-day cycles
- ✓ Analyze historical incident data to correlate fatigue risk peaks with near-miss events in ROC logs
- ✓ Apply regulatory thresholds (e.g., Australian Standard AS/NZS 4360:2019 fatigue limits) to evaluate schedule safety
📖 Why This Matters
In 2022, a major iron ore ROC recorded three uncommanded conveyor stoppages during 3:00–5:00 AM shifts—each coinciding with predicted alertness dips below 60%. Fatigue isn’t just about feeling tired; it degrades situational awareness, slows response time by up to 50%, and increases automation monitoring failure rates by 3.2× (Mine Health and Safety Council, 2023). For ROC engineers, modeling fatigue isn’t optional—it’s a frontline safety control.
📘 Core Principles
Fatigue risk arises from three interacting domains: (1) Homeostatic pressure—the longer awake, the greater the drive to sleep; (2) Circadian rhythm—a ~24-hour endogenous alertness cycle peaking at ~10 AM and troughing at ~4 AM; and (3) Task load—ROC monitoring tasks impose high vigilance demand with low event frequency, accelerating attentional drift. FRM synthesizes these using physiological models: SAFTE tracks sleep-wake history and predicts effectiveness (0–100%), while FAST adds real-time task demand weighting. Critically, ROC-specific adaptations include accounting for screen glare, seated posture, and delayed auditory feedback—all validated in CSIRO’s 2021 ROC Ergonomics Study.
📐 SAFTE-FAST Alertness Score
The SAFTE-FAST hybrid model computes a time-weighted effectiveness score (E_t) reflecting operator alertness at any hour. It combines homeostatic sleep pressure (H), circadian amplitude (C), and task demand multiplier (D). Used daily to benchmark shift designs against fatigue thresholds.
SAFTE-FAST Effectiveness Score
E_t = [(H_t + C_t)/2] × DTime-specific alertness score (0–100%) predicting operator effectiveness for safety-critical ROC tasks.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| E_t | Effectiveness Score | % | Predicted cognitive alertness level at time t |
| H_t | Homeostatic Pressure | % | Sleep pressure component derived from wake duration and prior sleep |
| C_t | Circadian Component | % | Circadian-driven alertness amplitude at circadian time t_c |
| D | Task Demand Multiplier | unitless | ROC-task-specific amplification factor (1.0 = baseline office work) |
Typical Ranges:
ROC day shift (06:00–14:00): 82–94%
ROC night shift (22:00–06:00), after 2 nights: 58–71%
💡 Worked Example
Problem: A ROC operator works a rotating shift: Day (06:00–14:00), Late (14:00–22:00), Night (22:00–06:00). Their last full sleep was 8 hrs ending at 06:00 on Day 1. Calculate E_t at 03:00 on Night 3 using SAFTE-FAST baseline parameters.
1.
Step 1: Compute homeostatic pressure H = 100 × (1 − e^(−t/16)), where t = hours since last sleep onset = 45 hrs → H ≈ 93.2
2.
Step 2: Apply circadian component C = 50 + 35 × cos(2π(t_c − 14)/24), where t_c = circadian time (03:00 ≈ CT 5.5) → C ≈ 28.1
3.
Step 3: Apply task demand multiplier D = 1.2 for ROC monitoring (per CSIRO validation); then E_t = (H + C)/2 × D = (93.2 + 28.1)/2 × 1.2 = 72.8
4.
Step 4: Compare to threshold: E_t < 75 indicates elevated fatigue risk per AS/NZS 4360:2019 Annex F
Answer:
The result is 72.8%, which falls below the recommended minimum alertness threshold of 75% and signals high fatigue risk requiring schedule adjustment.
🏗️ Real-World Application
At Rio Tinto’s Nammuldi ROC, engineers applied SAFTE-FAST to redesign 12-hour rotating shifts. Original pattern (Day-Late-Night-Off) produced 37% of night-shift hours below E_t = 75. After shifting to a slow-rotating ‘Day-Night-Off-Off’ sequence with mandatory 10-hr post-night rest, mean E_t rose from 68% to 81%, and ROC-initiated emergency stops dropped 64% over six months—verified via independent audit by the WA Department of Mines, Industry Regulation and Safety.