In recent years, frequent mining accidents have exposed numerous issues across traditional mining operations. With key coal-producing provinces across China successively issuing relevant policy documents, the pace of intelligent mine construction has accelerated. Upgrading conveyor belt transmission systems is a critical component of mine intelligence and a key direction for cost reduction and efficiency improvement.

△ Policies on intelligent mining from various provinces

Conveyor belt transmission systems face several industry-wide challenges. For example, low automation levels, inspection blind spots, and prolonged worker exposure to high-dust environments increase both safety and occupational health risks. To address these issues, a "Smart Mining All-in-One Machine" has been developed based on SOPHON computing products, establishing a "Conveyor Belt Intelligent Video Monitoring System." This system has been deployed in Shanxi Aluminum’s intelligent transformation project, significantly advancing its journey toward intelligence, automation, and unmanned operations.

01丨 Efficient and Accurate Algorithm Recognition

To enhance the safety, stability, and controllability of conveyor belt systems in mining operations, over 10 AI vision algorithms have been developed:

Conveyor Belt Operational Status Monitoring

Including belt blockage, material leakage, tearing, misalignment, empty running (uniform material), slippage, and foreign objects

Health Monitoring of Belt Drive Equipment

Including motors, gearboxes, couplings or hydraulic couplers, drive drums, idlers, and belt frames

Personnel and Environmental Safety Monitoring

Including open flames and smoke, safety helmet wearing, and personnel intrusion beyond boundaries

△ Application results of conveyor belt series algorithms

02丨 Flexible and Rapid Deployment with All-in-One Machine

SOPHON edge computing products offer high computational power, low power consumption, and compact size, with strong environmental adaptability, making them ideal for complex edge scenarios. The algorithm-integrated all-in-one machine built on SOPHON computing products enables fast deployment and flexible handling of high-concurrency application scenarios, delivering powerful edge computing capabilities for diverse smart applications.

△ SOPHON computing products

By leveraging self-developed AI vision algorithms and SOPHON high-density servers, the "Smart Mining All-in-One Machine" and the "Conveyor Belt Intelligent Video Monitoring System" have been co-built. The system enables remote backend monitoring, allowing real-time oversight of the entire conveyor belt system, timely judgment and response to alarm-triggering anomalies, and achieving round-the-clock real-time monitoring with centralized command and control.

03丨 Deployment and Application at Shanxi Aluminum

In Shanxi Aluminum’s intelligent transformation project, a series of AI vision algorithms for conveyor belt monitoring are deployed on SOPHON high-density servers SE6/8, covering nearly 100 cameras across the plant’s conveyor belt areas. The system monitors the operational status of over 10 conveyor belts in the raw material workshop and the health status of associated drive equipment, performing real-time algorithmic analysis via edge computing.

△ Conveyor Belt Intelligent Video Monitoring System

This project has not only elevated the level of intelligence and automation of Shanxi Aluminum’s conveyor belt transmission systems, significantly reducing human resource requirements, but also met the needs for cost-effective, efficient, and safe mining operations. Furthermore, it demonstrates the capability for full-chain algorithm development on heterogeneous platforms and the ability to deploy and infer algorithms across different computing platforms.

Shanxi Aluminum

Shanxi Aluminum Co., Ltd., a subsidiary of State Power Investment Corporation (SPIC), is a key enterprise in SPIC’s alumina production and operation. The company's main business is alumina, with an annual production capacity of 2.9 million tons. It operates its own thermal power plant and dedicated railway line, and currently manages four self-owned mines with bauxite geological reserves of approximately 100 million tons.

SOPHON

Founded in 2016, SOPHON is committed to becoming a global leader in general-purpose computing power. The company focuses on the research, development, and application of computing products such as TPU and RISC-V processors. Adhering to a fully open and open-source ecosystem philosophy, SOPHON collaborates with ecosystem partners to build a full-scenario product matrix covering cloud, edge, and endpoint devices, providing computing products and comprehensive solutions for applications in internet, urban operations, intelligent computing centers, general security, smart manufacturing, AIGC, and more.