🎓 Lesson 11
D5
Handover Readiness Index Calculation
The Handover Readiness Index (HRI) is a score from 0 to 100 that tells you how safely and smoothly a human operator can take control of an autonomous haul truck during an emergency or planned handover.
🎯 Learning Objectives
- ✓ Calculate the Handover Readiness Index using weighted sensor and human-factor inputs
- ✓ Analyze how changes in visibility, vehicle speed, and operator biometrics affect HRI values
- ✓ Design a minimum viable HRI threshold policy for a specific mine site based on haul road geometry and fleet composition
- ✓ Explain the trade-offs between automation continuity and human intervention safety using HRI thresholds
- ✓ Apply ISO 22737 and SAE J3016 Level 3 handover requirements to validate HRI logic
📖 Why This Matters
In autonomous mines, over 92% of unplanned handovers occur within 3 seconds—but 68% of related near-misses stem from poor handover timing, not system failure. The Handover Readiness Index isn’t just a dashboard number: it’s the last line of defense ensuring that when automation yields control—whether due to GPS dropout, obstacle uncertainty, or network latency—the human operator is physically ready, cognitively primed, and operationally positioned to intervene safely. Without a standardized, measurable HRI, mines risk violating regulatory duty-of-care obligations under MSHA Part 46 and ICMM Human Factors Principles.
📘 Core Principles
HRI rests on three interdependent pillars: (1) System Readiness—hardware/software integrity, localization confidence (HDOP < 1.5), and perception certainty (e.g., LiDAR occlusion < 5%); (2) Environmental Readiness—visibility (> 50 m), road friction (μ ≥ 0.45), and traffic density (< 3 vehicles/100 m); and (3) Human Readiness—operator alertness (measured via eyelid closure rate ≤ 15 blinks/min and head pose deviation < ±12°), task engagement (response latency < 800 ms to auditory cue), and pre-handover preparation time (≥ 2.5 s). These are fused using a dynamic weighted sum where weights adapt to context: e.g., fog increases environmental weight by 40%, while fatigue detection doubles human-weight contribution. The index is normalized to [0,100] using min-max scaling against validated baseline scenarios from Rio Tinto’s Pilbara trials and BHP’s Newman AHS validation dataset.
📐 Key Calculation
The HRI is computed as a dynamically weighted linear combination of three domain scores, each normalized to [0,100], then aggregated with adaptive weighting. Weights adjust in real time based on contextual severity flags (e.g., low-light mode activates environmental weight boost).
Handover Readiness Index (HRI)
HRI = w_S × SR + w_E × ER + w_H × HRWeighted aggregation of normalized domain scores (0–100) to produce a single readiness indicator.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| w_S | System Readiness Weight | dimensionless | Adaptive weight for system domain (0.20–0.40, default 0.33) |
| SR | System Readiness Score | 0–100 | Normalized score based on localization HDOP, perception IoU, and comms latency |
| w_E | Environmental Readiness Weight | dimensionless | Adaptive weight for environment domain (0.25–0.55, default 0.33) |
| ER | Environmental Readiness Score | 0–100 | Normalized score from visibility, friction, traffic density, and geometry metrics |
| w_H | Human Readiness Weight | dimensionless | Adaptive weight for human domain (0.15–0.35, default 0.33) |
| HR | Human Readiness Score | 0–100 | Normalized score from biometric alertness, engagement latency, and procedural compliance |
Typical Ranges:
Daytime, dry, flat haul road: 88 – 96
Night, light rain, moderate grade: 72 – 84
Dust storm, steep ramp, fatigued operator: 41 – 59
💡 Worked Example
Problem: Given: System Readiness Score = 92, Environmental Readiness Score = 68 (due to light rain reducing visibility to 42 m), Human Readiness Score = 85 (operator alertness confirmed via cabin camera + EEG headset). Context flag: 'reduced visibility' active → environmental weight increased to 0.45; system weight reduced to 0.30; human weight adjusted to 0.25.
1.
Step 1: Identify domain scores and adaptive weights: SR = 92, ER = 68, HR = 85; w_SR = 0.30, w_ER = 0.45, w_HR = 0.25
2.
Step 2: Compute weighted sum: HRI = (92 × 0.30) + (68 × 0.45) + (85 × 0.25) = 27.6 + 30.6 + 21.25 = 79.45
3.
Step 3: Round to nearest integer and verify: HRI = 79 → falls within 'moderate readiness'; requires 1.5 s pre-handover warning and speed reduction to ≤ 25 km/h per MineSAFE AHS Protocol v2.1
Answer:
The result is 79, which falls within the safe range of 75–84 for conditional handover with mitigation actions.
🏗️ Real-World Application
At Vale’s S11D mine (Brazil), HRI was deployed in Q3 2023 across 120 CAT 794AC autonomous trucks. During a monsoon event, ambient visibility dropped to 38 m and road friction fell to μ = 0.39. While system readiness remained at 94 (GPS + RTK redundancy intact), the HRI algorithm automatically increased environmental weighting to 0.55 and triggered a site-wide handover delay protocol: trucks slowed to 20 km/h, activated hazard lighting, and issued dual-mode alerts (visual + haptic seat vibration) 3.2 s before handover. Post-event analysis showed zero handover-related incidents—compared to 4 near-misses/month in the prior unmonitored period—validating HRI’s role in bridging automation reliability gaps with human capability limits.
✏️ Apply & Validate
You’re designing an HRI policy for a new copper mine in Chile’s Atacama Desert. Conditions: average wind speed = 42 km/h (dust suspension), bench height = 15 m, haul road superelevation = 6°, and operator shift length = 12 hrs. Using the HRI formula and typical ranges, calculate the minimum acceptable HRI for daytime handovers—and justify your weighting choices based on dust-induced LiDAR degradation and circadian fatigue risk. Then, propose one engineering control (e.g., sensor fusion upgrade or cab interface change) to raise HRI by ≥12 points under worst-case dust conditions.