Coal Subsurface Mapping for Open Cut Selective Mining

A reliable coal seam sensing system is required to improve the productivity of selective mining in open cut mining operations. Currently, operators are required to manually adjust the material extraction depth based on various cues, such as sight or machine vibration as the implement cuts the strata. Manual seam selection can be complex or tedious which often leads to a diluted coal product. Furthermore, the ability to visually observe changes in the geology at night time or in poor weather conditions can be near impossible if the coal looks similar to the waste material. Although assistive guidance systems exist for shovels and excavators, these systems are typically uploaded with extraction surfaces generated from existing geophysical logs which only provide sparse detail about the geology.

This project developed a prototype subsurface scanning system that can measure coal thickness from the top of the exposed coal surface down to a coal-interburden interface. The system consists of a sled that houses ground penetrating radar and position sensors, towed by a remotely controlled robotic vehicle. The system scans the open cut surface in a grid-like fashion after the surface has been prepared, but prior to mining. A digital surface model of the underlying coal to interburden interface is generated which can be uploaded to an in-cab guidance system.

The key research outcomes include the following:

• A workflow was developed where information generated by a GPR-based sensing system that scans the exposed seam roof surface can be processed and integrated into production to provide guidance information for operators with minimal impact to production.
• The level of interference and clutter that different materials and a robotic tow vehicle introduce into GPR sensor data was investigated.
• A custom sled was designed and manufactured whereby no noticeable interference is introduced into the sensor data.
• Custom software to acquire and process sensor data was developed that allows the generation of a digital surface model within hours of survey completion.
• Three surveys were conducted at a production mine site and the performance of the sensing system was assessed. Whilst the system did not meet the desired accuracy tolerance specification of ±5% requested by site personnel, there are questions regarding the validity of the accuracy assessment method. Furthermore, lessons learnt from these surveys provide insight as to how the performance can be improved.
• The full system was successfully demonstrated at a production mine site with a dozer operator mining using a digital surface model for in-cab machine guidance.

The benefits described by the dozer operator that evaluated the digital surface model during survey #3 provide strong motivation for ongoing technology development. It is recommended that ongoing research and development be pursued with the following activities:

• Conduct additional surveys incorporating the lessons learnt from survey #3 such as using the GPR scanning system to measure the elevation of the mined floor.
• Trial the system in the application of coal roof detection to measure the remaining overburden stand-off thickness after a blast cycle.
• Implement autonomous control for the robotic towing vehicle and develop path planning software to enable mine sites to survey areas automatically.
• Adapt the system so that it can be machine-mounted to provide layer thickness measurements for real-time sensing whilst mining.