Development and Validation of an Integrated LiDAR-Camera System for Real-Time Monitoring of Underground Longwall Operations

Pasindu Ranasinghe and colleagues at UNSW have designed and validated a compact LiDAR-camera monitoring system purpose-built for underground longwall coal mining operations. The system addresses critical challenges like methane explosion risks, poor visibility, high heat, and limited bandwidth by housing a solid-state LiDAR, industrial RGB camera, and onboard processor inside a certified flameproof enclosure. In laboratory tests, the polycarbonate dome introduced minimal optical distortion, and careful heat sinking and passive conduction lowered peak surface temperature from 106°C to 70°C, with internal temperatures stabilizing at 57°C. A simulated longwall environment demonstrated that over 97% of LiDAR points fell within the camera's field of view, enabling reliable colorization of point clouds. The system transmitted processed, colorized 3D data at up to 10 Hz while keeping bandwidth below 25 Mb/s—well within typical underground network constraints.

The real-world impact is significant: current underground monitoring relies on cameras that lack depth information or on periodic, monochromatic LiDAR scans. This fusion system provides continuous, real-time spatial awareness with rich visual context, allowing operators to detect hazards, track equipment, and monitor strata movement with far greater precision. By combining the strengths of both sensors in a safe, bandwidth-efficient package, the research paves the way for safer, more efficient longwall mining operations. The paper, published on arXiv (2605.02516), details the enclosure-aware calibration and thermal management techniques that make the system viable for permanent deployment in explosive atmospheres.