Advanced Muck-Pile Characterization: UAV, PCA and AI Synergy in Optimizing Blast Design and Excavator Loading Efficiency

Abstract Muck-pile behavior plays a pivotal role in optimizing mining operations, as post-blast throw, drop, and lateral dispersion directly influence equipment selection and loading efficiency. The present study conducted a series of novel blasting trials to systematically evaluate the effects of different blast design parameters on key muck-pile characteristics. Principal Component Analysis (PCA) was employed to determine the most influential design variables for developing optimized blast configurations. The investigation comprised multiple combination blasts executed in four distinct phases at the OC1 RGIII mine of SCCL. Blast layouts were precisely developed using advanced blasting software and implemented in accordance with the parameters identified through PCA. Muck-pile properties were quantified using advanced AI-based analytical tools, yielding detailed and reliable insights.The findings revealed that a spacing-to-burden ratio of 1.35, stemming length equal to 0.9 times the burden, 1 m decking, and a V-pattern initiation sequence produced superior muck-pile conditions. This optimized configuration resulted in a reduction of drop by 3 m, a decrease in throw by 5.9 m, and an increased lateral spread of 19.3 m. These improvements contributed to smoother loader manoeuvrability, enhanced loading efficiency, and overall operational optimization. Keywords AI–Muckpile, Blast fragmentation, Muck-pile parameters and Unmanned Aerial Vehicle