Tech company Micromine announced the launching of a new underground mining precision performance software that uses machine learning to refine loading and haulage processes. The program is part of the company's fleet management and mine control solution, Pitram, as was initially designed by a master's student at the University of Western Australia.
In a press release, Micromine said that by using the processes of computer vision and deep machine learning, onboard cameras are placed on loaders to track variables such as loading time, hauling time, dumping time and travelling empty time. The video feed generated is then processed on the Pitram vehicle computer edge device and the information is transferred to Pitram servers for processing and analyses.
"By capturing images and information via video cameras and analysing that information via comprehensive data models, mine managers can make adjustments to optimise performance and efficiency," the Australian firm's Chief Technology Officer Ivan Zelina said in the media statement. "It also provides underground mine managers with increased business knowledge, so they have more control over loading and hauling processes and can make more informed decisions which, in turn, improves safety in underground mining environments."
According to Zelina, the technology was already tested through pilot programs in Australia, Mongolia, and Russia and the results showed enhancements in mining companies knowledge of their loading processes through automated data collection and analysis.