Enhancing stockpile inventory management through UAV- based volume estimation: a case study of salt stockpiles in Hambantota mahalewaya

Abstract

Accurate volume estimation of stockpiles is crucial in industries such as Mining, Construction, salt, and Agriculture to optimize resource utilization. This study evaluates the effectiveness of Unmanned Aerial Vehicles (UAVs) compared to Differential Global Positioning System (DGPS) and Total Station (TS) methods for volume estimation of outdoor salt stockpiles in Hambantota Mahalewaya, Southern province of Sri Lanka. The inventory identified two stockpiles, stockpile 1 and stockpile 2, with volumes of 1832.25 m3 and 819 m3, respectively. An optimal elevation of 55m was utilized for UAV surveys, and the results were compared with DGPS and TS measurements. UAV surveying factors affecting errors, including image resolution, Ground Control Points (GCPs), and image processing software, were assessed for both stockpiles. Survey time and cost for each method were also analyzed. Pix4dMapper and Agisoft Metashape software processed UAV images, while Civil3D software processed DGPS and TS data. Results indicated that increasing UAV survey elevation reduced volume error percentages for both stockpiles, with and without GCPs. For Stockpile 1, UAV volume estimation showed a 0.88% difference from the actual volume, compared to 4.81% for DGPS and 3.35% for TS. Conversely, for Stockpile 2, UAV estimation differed by 0.95%, while DGPS and TS showed differences of 0.56% and 0.10%, respectively. UAV surveys proved efficient in terms of survey time and labor intensity. Despite technological advancements, challenges remain, particularly in addressing topographical variations for accurate volume estimation. To improve UAV-based estimation, addressing bottom elevation discrepancies by establishing fixed benchmarks on flat terrains was suggested. Nonetheless, UAV-based approaches offer fast and relatively reliable results, indicating their potential for widespread adoption.