Abstract:Slag pot transportation in smelting plants, which involves moving hot molten slag from furnaces to cooling yards and subsequently transporting cooled slag to dumping stations, is traditionally plagued by harsh operating conditions such as poor lighting, high temperatures, and high dust concentrations. These conditions pose significant manual operational risks. The advent of intelligent and 5G technologies has created the technical prerequisites for remote slag pot transportation, presenting substantial market potential and critical safety significance. As teleoperation represents the future trend for hot slag handling, this paper presents the retrofitting of a slag pot carrier’s operational process based on machine learning. By leveraging remote driving technology, we achieve beyond-visual-line-of-sight (BVLOS) operation of the carrier, thereby mitigating operational risks and improving the working environment for personnel. The study lays a solid foundation for enhancing the automation and intelligence of the entire smelting production process and demonstrates significant engineering application value.