🎓 Lesson 20
D5
Pilbara Fleet Harmonization Workshop
Pilbara Fleet Harmonization is the process of aligning different autonomous haul trucks, loaders, and support systems so they work together smoothly and safely at iron ore mines in Western Australia.
🎯 Learning Objectives
- ✓ Analyze fleet heterogeneity using OEM telemetry data formats and identify interoperability gaps
- ✓ Design a harmonized dispatch policy that maintains ≥92% fleet utilization across mixed-AHS operations
- ✓ Calculate and validate inter-fleet communication latency thresholds for safe collision-avoidance response (<120 ms end-to-end)
- ✓ Apply ISO 15143-3 AEMP Level 2 data schema to normalize payload, fuel, and health metrics across three vendor platforms
- ✓ Evaluate fleet harmonization maturity using the Autonomous Mining Maturity Model (AMMM) Stage 3–4 criteria
📖 Why This Matters
In the Pilbara — home to >70% of Australia’s iron ore exports — autonomous haulage fleets now exceed 400+ trucks per major site. Yet most operations run mixed OEM fleets due to phased procurement, legacy contracts, and technology refresh cycles. Without harmonization, these fleets suffer from inconsistent cycle times (+18% variance), dispatch system conflicts, duplicated maintenance workflows, and near-miss incidents caused by non-uniform perception-response delays. This workshop bridges the gap between theoretical AHS integration and real-world Pilbara deployment — where geology, dust, heat (45°C+), and 1,200 km supply chains demand robust, field-proven harmonization.
📘 Core Principles
Harmonization rests on three interdependent pillars: (1) Data Interoperability — enforcing standardized telemetry exchange (AEMP/ISO 15143-3) and semantic alignment (e.g., ‘payload’ must mean gross payload minus tare, not bucket weight); (2) Operational Synchrony — aligning dispatch logic (e.g., priority rules, queuing thresholds, speed profiles) across vendor-specific fleet management systems (FMS); and (3) Physical Integration — ensuring consistent vehicle dynamics (acceleration/deceleration rates, braking distance tolerances, GPS-RTK correction latency) so traffic management systems (TMS) can predict behavior within ±0.3 m positional uncertainty. Crucially, harmonization is *not* homogenization: it preserves OEM innovation while constraining variability to pre-approved tolerance bands defined in the site’s Autonomous Operations Protocol (AOP).
📐 End-to-End Communication Latency Budget
Safe AHS coordination requires all sensor-to-action loops (e.g., obstacle detection → path replanning → brake command execution) to complete within strict latency budgets. This formula allocates allowable delay across network, processing, and actuation layers — critical when integrating third-party TMS with OEM-specific vehicle controllers.
Total System Latency (TSL)
TSL = L_sensor + L_compute + L_network + L_actuateMaximum allowable end-to-end delay for safety-critical AHS control loops.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| L_sensor | Sensor fusion latency | ms | Time from raw sensor capture (LiDAR, radar, camera) to fused object list |
| L_compute | Onboard compute latency | ms | Time for path planning, risk assessment, and command generation |
| L_network | Network transmission latency | ms | Round-trip time for V2X or cellular uplink/downlink of critical commands |
| L_actuate | Actuator response latency | ms | Time from command receipt to physical action (e.g., brake apply, steering torque) |
Typical Ranges:
Pilbara high-speed haul (≥55 km/h): 95 – 120 ms
Low-speed loading/dumping: 60 – 90 ms
💡 Worked Example
Problem: A Pilbara site integrates Komatsu Haul Trucks (with OEM TMS) with a third-party AI-based collision avoidance module. Measured sensor fusion latency = 42 ms, onboard compute delay = 28 ms, wireless V2X transmission (Wi-Fi 6E) = 15 ms, actuator response (brake apply) = 35 ms. Is total latency within safe limits?
1.
Step 1: Sum all measured latencies: 42 + 28 + 15 + 35 = 120 ms
2.
Step 2: Compare against Pilbara industry benchmark: ≤120 ms for <100 km/h operation (Rio Tinto AOP v4.2, Sec. 7.3.1)
3.
Step 3: Verify margin — no headroom remains; recommend redundant comms path or edge-compute offload to reduce compute latency by ≥10 ms
Answer:
The result is 120 ms, which meets but does not exceed the safe limit of 120 ms for high-speed haulage in open-pit environments.
🏗️ Real-World Application
At BHP’s South Flank mine (2023 deployment), 92 Komatsu 930E-1S and 38 CAT 794AC trucks operated under a single DISPATCH® FMS. Harmonization required: (1) AEMP Level 2 data mapping to unify ‘engine load’, ‘hydraulic pressure’, and ‘tire temperature’ semantics; (2) custom speed profile overlays to match CAT’s higher top speed (64 km/h) with Komatsu’s optimized cruise (58 km/h) without disrupting queue stability; and (3) synchronized RTK-GNSS correction streams via Trimble Pivot™ to ensure <10 cm cross-fleet positioning error. Result: 11% reduction in average cycle time variance and zero TMS-induced near misses over 14 months.
🔧 Interactive Calculator
🔧 Open Autonomous Haulage Systems Integration Calculator📋 Case Connection
📋 Open Pit Iron Ore Mine Fleet Harmonization at Rio Tinto Pilbara
Heterogeneous OEM protocols causing dispatch system latency >12s and payload mismatch errors