📋 Case Study

Chilean Copper Mine: Autonomous Haul Fleet Deployment

Achieving safe, reliable, and productive autonomous haulage under extreme environmental conditions (high altitude: 3,200 m ASL; ambient temperatures from −5°C to 42°C; abrasive dust; and steep, winding haul roads) while maintaining compatibility with legacy infrastructure and minimizing disruption to ongoing production.

🏗️ Project Overview

A Tier-1 copper mine in the Atacama Desert, northern Chile, deployed an autonomous haul fleet across its open-pit operation. The site processes ~450 ktpd of ore and waste, with a 2.8-km average haul distance and 320-m vertical lift. The project involved retrofitting and integrating 42 autonomous 290-tonne CAT 794 AC electric drive haul trucks into existing dispatch and traffic management systems.

🎯 Challenge

Achieving safe, reliable, and productive autonomous haulage under extreme environmental conditions (high altitude: 3,200 m ASL; ambient temperatures from −5°C to 42°C; abrasive dust; and steep, winding haul roads) while maintaining compatibility with legacy infrastructure and minimizing disruption to ongoing production.

🔧 Design Approach

Systems engineering lifecycle approach: (1) Digital twin–based route simulation and risk modeling; (2) Multi-layered sensor fusion architecture (LiDAR, radar, GNSS-RTK, IMU, thermal cameras); (3) Edge-computing–enabled real-time path planning with dynamic obstacle avoidance; (4) Phased commissioning—starting with unladen truck validation, progressing to loaded haulage in segregated zones, then full integration with human-operated equipment using ISO 26262–aligned functional safety certification.

📐 Design Diagram

Chilean Copper Mine: Autonomous Haul Fleet DeploymentDTDigital TwinSFSensor FusionECEdge ComputePCPhased Commissioningd = 187.3 mBraking distanceA = 22.6 dBLiDAR attenuationσ_pos = 0.17 mGNSS-RTK (3D RMS)Extreme EnvironmentAltitude: 3200 m ASL • Temp: −5°C to 42°C • Dust: ρ = 1200 μg/m³ • Steep/winding roads

AI-generated project design illustration

📐 Key Calculations

Required braking distance at max grade

d = v² / (2 × g × (μ × cosθ + sinθ))
Result: 187.3 m
Informed road widening and emergency stopping zone placement on 10% grade sections; ensured compliance with ISO 21627 safety thresholds for autonomous emergency stops.

Dust-induced LiDAR attenuation margin

A = 10 × log₁₀(P_in / P_out) = σ × ρ × L
Result: 22.6 dB at 150-m range (ρ = 1200 μg/m³ avg PM10)
Drove selection of 1550-nm wavelength LiDAR over 905-nm to maintain detection reliability in high-dust conditions.

GNSS-RTK positioning uncertainty at altitude

σ_pos = √(σ_horiz² + σ_vert²) ≈ 0.08 m (horizontal), 0.15 m (vertical)
Result: 0.17 m (3D RMS)
Validated sub-20-cm lateral accuracy required for lane-keeping on 12-m-wide haul roads with 0.5-m safety margins.

📊 Results

Metrics: Fuel consumption reduced by 12.4%, Average payload utilization increased from 91.2% to 97.8%, Haul cycle time variance decreased by 63%, Zero autonomous-related lost-time injuries over 32 months
The autonomous fleet achieved 94.7% scheduled availability (exceeding target of 92%), contributed to a 17% increase in annual waste movement capacity, and enabled consistent 24/7 operations despite labor shortages—without compromising safety or maintenance integrity.

💡 Lessons Learned

  • High-altitude GNSS signal degradation requires redundant inertial navigation calibration routines every 4 hours
  • Dust ingress mitigation must be co-designed with OEMs—not retrofitted—especially for cooling intakes and sensor housings
  • Cross-functional operator–technician shift handover protocols are critical for anomaly resolution during mixed-fleet transitions

Key Takeaways

  • 1Autonomous haulage success hinges less on AI sophistication and more on robust, environment-hardened sensing and deterministic control architecture