The mining industry, long defined by harsh conditions and heavy labour, is undergoing a tech-driven transformation. Such technologies are not only enhancing efficiency but also addressing critical challenges such as safety, sustainability, and cost management.
By leveraging Internet of Things (IoT) sensing devices and AI, mines can implement sophisticated machine-to-machine communication and remote operation of autonomous vehicles and equipment; technologies that can improve productivity and safety by reducing human exposure to hazards and supporting continuous, around-the-clock operations.
What is Huawei’s Pangu AI?
At the forefront of this revolution is Huawei’s Pangu – a family of AI foundation models built to drive industrial innovation. The latest iteration, Huawei Cloud Pangu 5.5, was unveiled in 2025 as “a 718-billion-parameter deep thinking model for industrial applications”[i]. Pangu has already been implemented in over 500 scenarios across more than 30 industries, with its capabilities spanning natural language, scientific computing, and computer vision, making it a versatile engine for smart mining operations. The Pangu architecture includes a layered structure that ranges from general foundation models to industry-and task-specific models, allowing for flexible deployment across cloud, edge, and device levels.
Pangu is transforming mining in three crucial ways:
- Inspection and safety (Pangu CV) – Using cameras, infrared, lidar and radar, to detect cracks, wear, belt damage, hot bearings, leaks, loose rock or support failures.
- Process and energy optimisation (Pangu Prediction Model) – Using data analytics to predict throughput, recovery, quality, energy use and failures, and make recommendations.
- Autonomy, robotics, and harsh-environment work prevention (Pangu Multimodal and World Model) – Using IoT to safely guide and control autonomous trucks, loaders, drilling and service vehicles, via cameras and lidar data.
Mining companies can now also finetune domain models on Pangu via ModelArts – combining geology, operations, maintenance and safety data – creating the pioneering “Mining large model” for decision support, technical Q&A, and planning.
This latest innovation comes on the back of the Pangu Mine Model – the world’s first commercial AI model for the energy and mining sector developed by Huawei in 2023[ii]. Connecting cloud and edge with central AI resources handling model training and coordination, edge devices on-site apply real-time AI inference. By requiring only small amounts of mine data to learn, the Pangu model can autonomously improve safety and efficiency across key mining activities: from drilling and digging, to equipment control, haulage, and even ventilation.
Yimin Mine’s autonomous fleet
These advances in AI are in evidence at the Yimin open-pit coal mine in Inner Mongolia, which has become a showcase for intelligent mining. In May 2025 Yimin made history by deploying a fleet of 100 autonomous, all-electric mining trucks – the world’s first large-scale autonomous haulage fleet. A joint innovation by Huaneng Inner Mongolia Eastern Energy Co., Ltd. and Huawei, the project aimed to create a “zero-carbon, autonomous, and intelligent” mining transport system.
The trucks can operate 24/7, even under extreme winter conditions, which can plunge to -48.5°C and often include dust storms, fog, and blizzards. Each 45-tonne truck is custom-built by Huawei and XCMG with its driver’s cab replaced with a huge battery pack. At least two cameras, lidar units, and millimetre-wave radars feed a 360° field of vision to the truck’s AI brain to maintain a 40-metre area of visibility even in complete darkness. The AI can also avoid soft ground with special anti-sink controls keeping the heavy trucks from getting bogged down, and can detect roadway damage, automatically alerting maintenance crews.
Meanwhile, the accuracy of the truck’s reversing and docking for unloading, surpasses most human drivers. There are around 800 sensors at the mine to monitor equipment and site security, while AI-powered LiDAR and BeiDou satellite positioning continually assess the stability of steep slopes. If any ground movement is detected, the system can react quickly to prevent accidents, replacing reactive with predictive safety management. Additionally, automating the fleet has freed up over 350 employees for safer, more skilled supervisory and maintenance roles.
Award-winning Pangu 5.5 AI is helping make mines safer and more efficient
Huawei’s Pangu 5.5 AI has been awarded the Mining Technology Excellence Awards 2025 for supporting coal mining operations, with outstanding applications in digging, machinery, and transport. Special consideration was given to the Pangu CV Model for its use in detecting faults like cracks or heat anomalies in mines. The award also recognises Huawei and Huaneng Inner Mongolia Eastern Energy Co., Ltd. for Pangu’s transformative impact on mining safety, efficiency, and sustainability.
The example of Yimin Mine shows how innovation in autonomous trucks and intelligent infrastructure can transform an age-old industry – to make it safer, more efficient, and ready to meet the challenges of the future.
[i] https://www.huaweicloud.com/eu/news/20250620192415143.html
[ii] https://e.huawei.com/en/case-studies/industries/mining/shandong-energy-pangu-2023
