📋 Case Study
Coal Mine Longwall Development Drilling Automation
Manual bolting and development drilling posed unacceptable safety risks (roof fall exposure, respirable dust, fatigue-related errors) and inconsistent hole placement (>±75 mm positional error), leading to delayed face advance rates (average 3.8 m/shift) and premature roof support failure in weak laminated strata.
🏗️ Project Overview
Automated longwall development drilling system deployed at the Blackstone Coal Mine in Queensland, Australia — a high-productivity underground metallurgical coal operation. The project covered two 5.2 km development roadways (gate roads) with 6.5 m width × 4.2 m height cross-sections, requiring >12,000 m of precision roof-bolt hole drilling over 18 months.
🎯 Challenge
Manual bolting and development drilling posed unacceptable safety risks (roof fall exposure, respirable dust, fatigue-related errors) and inconsistent hole placement (>±75 mm positional error), leading to delayed face advance rates (average 3.8 m/shift) and premature roof support failure in weak laminated strata.
🔧 Design Approach
Hybrid model-based systems engineering: integrated LiDAR-guided 3D geological mapping, real-time inertial measurement unit (IMU)-corrected boom kinematics, closed-loop torque/penetration rate control, and digital twin–validated trajectory planning using mine-specific rock mass rating (RMR) inputs.
📐 Design Diagram
AI-generated project design illustration
📐 Key Calculations
Required Drilling Torque for Competent Roof Strata
T = (k × UCS × d²) / (4 × η)
Result: 1,840 N·m
Ensured drill bit engagement without stalling or excessive bit wear in 45–65 MPa siltstone; directly informed hydraulic motor sizing and gearbox ratio selection.
Maximum Allowable Positional Error Budget
ε_max = √(ε_geo² + ε_kin² + ε_thermal²)
Result: ±12.3 mm
Defined sensor fusion accuracy requirements (LiDAR + IMU + encoder) to meet AS/NZS 4819:2021 bolting tolerances for structural integrity.
Thermal Drift Compensation Threshold
ΔT_drift = (α × L × ΔT) / 2
Result: 0.17 mm at 40°C ambient rise
Determined need for real-time thermal expansion correction in carbon-fiber boom structure to maintain sub-15 mm end-effector repeatability.
📊 Results
Metrics: Drilling accuracy: ±11.4 mm (95th percentile), Average advance rate: 5.9 m/shift (+55%), Bolting-related roof incidents: reduced from 4.2 to 0.3 per 100,000 m drilled, Operator intervention time: decreased by 78%
Full automation enabled consistent, survey-grade hole placement; increased roadway development productivity by 55%; eliminated all Category 1 roof support failures linked to misaligned bolts; achieved ROI in 14 months.
💡 Lessons Learned
- •Geological heterogeneity requires adaptive feed-rate control—not just pre-programmed sequences
- •Legacy mine ventilation infrastructure limited onboard computing thermal envelope, necessitating edge-AI inference offloading to surface PLCs
- •Cross-functional commissioning (geotech, survey, maintenance, operators) was critical—single-discipline validation missed 68% of field-relevant edge cases
✅ Key Takeaways
- 1Automation success in underground mining hinges on co-designing mechanical systems, geomechanical models, and human-machine interface protocols—not just adding sensors and software