Autonomous mining through cooperative driving and operations enabled by parallel intelligence

Autonomous mining is promising to address several current issues in the mining sector, such as low productivity, safety concerns, and labor shortages. Although partial automation has been achieved in some mining operations, fully autonomous mining remains challenging due to its complexity and scalability in field environments. Here we propose an autonomous mining framework based on the parallel intelligence methodology, employing self-evolving digital twins to model and guide mining processes in the real world. Our framework features a virtual mining subsystem that learns from simulating real-world scenarios and generates new ones, allowing for low-cost training and testing of the integrated autonomous mining system. Through initial validation and extensive testing, particularly in open-pit mining scenarios, our framework has demonstrated stable and efficient autonomous operations. We've since deployed it across more than 30 mines, resulting in the extraction of over 30 million tons of minerals. This implementation effectively eliminates the exposure of human operators to hazardous conditions while ensuring 24-hour uninterrupted operation.

Long Chen, Yuting Xie, Yuhang He, Yunfeng Ai, Bin Tian, Lingxi Li, Shirong Ge, Fei-Yue Wang