Mining is a high-risk industry, so efficiency and safety are key priorities. As mines continue to go deeper and exploit low-grade deposits, bulk mining methods, such as sublevel caving (SLC), have become increasingly important. SLC is suitable for massive steeply dipping ore bodies and is known for its high degree of mechanisation, productivity, and low operational cost. Moreover, technological developments and mechanisation have allowed these methods to be applied at greater depths. In modern mechanised mines Load haul dump (LHD) machines are central to achieving the desired productivity. Therefore, automation of LHDs and their increasing use in mines make it crucial to understand the performance of these machines in actual mining environments. The aim of this research was to understand the differences in the productivity of semiautonomous and manual LHDs and identify how external factors impact the performance of these machines in SLC operations. The research also investigated how LHD operator training could improve the loading efficiency.

Performance data for semi-autonomous and manual LHDs were collected from LKAB’s Kiirunavaara mine’s central database, GIRON. These data were used to compare cycle times and payloads of semi-autonomous and manual LHDs. The data were filtered and sorted so that only data where both machine types were operating in the same area (crosscut, ring, and ore pass) were used. To understand the impact of external factors, data on the occurrence of boulders were collected from LKAB’s Malmberget mine by recording videos of LHD buckets, while the data on operator training were obtained by performing baseline mapping and conducting a questionnaire study with the LHD operators at LKAB’s Kiirunavaara mine.

The results of the comparative analysis of manual and semi-autonomous LHDs showed the mean payload was 0.34 tonnes higher for manual LHD machines. However, the differences were not consistent across different areas of the mine. Similarly, when comparing the cycle times, in 57% of the studied area, manual LHDs had lower cycle time, while the opposite was true in the remaining 43% of the areas. Therefore, the differences in cycle time and payload due to mode of operation are not conclusive, meaning that one machine type does not completely outperform the other. This highlights the importance of understanding the external factors that cause such differences. Moreover, the findings emphasize the need to upgrade LHD operator training based on pedagogical principles and the inclusion of new technologies to enhance loading efficiency and increase overall productivity.

The mining sector has evolved over the years, increasingly adopting automation and digitalisation to improve safety, reduce carbon footprint, and enhance productivity. The integration of digital technologies and automation continues to change traditional mining practices and the nature of work. Load haul dump (LHD) machines remain integral to the automation of underground hauling operations. Additionally, in mines that utilise the density difference of ore and waste, the bucket weight from these machines is also used to determine the grade of the ore. Consequently, the automation of LHDs and their growing use in mines necessitate a comprehensive understanding of their performance and impact on loading control and dilution. 

The aim of this research was to investigate the impact of digitalisation and automation on future LHD operations. It explored the differences in productivity due to mode of operation, its impact on iron grade calculation and future training and competence of mining personnel.

Performance data for semi-autonomous and manual LHDs were collected from LKAB’s Kiirunavaara mine’s central database, GIRON. These data were used to compare cycle times and payloads of semi-autonomous and manual LHDs. The data were filtered and sorted so that only data where both machine types were operating in the same area (crosscut, ring, and ore pass) were used. To evaluate the sensitivity of density-based Fe grade calculation the data were simulated and analysed using global sensitivity analysis. Moreover, the data on operator training were collected through baseline mapping and conducting a questionnaire study with the LHD operators at LKAB’s Kiirunavaara mine. Whereas the data on end-users perspective of digitalisation and automation was based on questionnaire study at LKAB, and workshops conducted with production workers from Aitik and Garpenberg mines at Boliden. 

The comparative analysis of manual and semi-autonomous LHDs showed the mean payload was 0.34 tonnes higher for manual LHD machines. However, these differences were not consistent across different areas of the mine. Similarly, when comparing the cycle times, in 57% of the studied areas, manual LHDs had lower cycle time, while the opposite was true in the remaining 43% of the areas. Therefore, the differences in cycle time and payload due to mode of operation are not conclusive, meaning that one machine type does not completely outperform the other. This highlights the importance of understanding the external factors that cause such differences. In terms of sensitivity of density-based iron grade calculation, the bucket weight, followed by void ratio and fill factor were identified as the most significant input parameters. Moreover, the findings from the survey conducted with operators and production workers anticipate an increased transition towards autonomous operations. They believed the impacts of digitalisaiton and automation are positive, but a small proportion had negative perceptions. In terms of education they identify the need to upgrade training and emphasise the understanding of mining processes along with computer skills will remain crucial competencies in the future to facilitate digitalisation and automation.