Coal mines get “AI eyes”: safety and efficiency soar 2025-10-11 09:48:26.0
Guangming Daily, People’s Daily, China.org.cn, China Coal News and other leading outlets have run the story “Coal mines get ‘AI eyes’: safety and efficiency soar”, spotlighting the Vision-AI Integrated Management & Control Platform for coal mines developed independently by CCTEG Tianma Intelligent Control. The coverage is summarised below.
The platform—now in full-scale operation—blends high-definition video, computer vision and artificial intelligence to solve three long-standing underground headaches: hard-to-predict risks, slow response times and low efficiency. It shifts safety management from “human watch” to “AI watch”, raising the bar for coal-mine safety. To date it has been rolled out at dozens of major producers including China Energy, Shandong Energy and Shaanxi Coal Group.
Face conditions are complex and human–machine interactions frequent; traditional walk-by inspections or after-the-fact video reviews miss hazards in real time and cap automation levels. Tianma embeds AI algorithms directly into mine scenes, giving every camera a “smart brain”. HD cameras in critical zones stream video that is analysed on the fly to spot equipment, personnel and environmental anomalies, replacing passive recording with an active “sense-analyse-respond-control” loop.
All underground video is unified on one platform: cameras and streaming servers across the entire mine are centrally orchestrated for unified access, scheduling, storage and service. Whether it is a fixed roadway camera or a dynamic longwall face, managers view crisp, sub-200 ms latency footage from a single screen, breaking down legacy silos.
The system acts as a 24-hour “AI sentinel”, covering people, machines, the environment and management with >96 % recognition accuracy. Even in low-light, high-dust conditions it reliably tracks shearers, shields, coal flow and more—”see clear, judge right, run steady”. When something is wrong, the platform pushes alerts via app, pop-ups or audible/visual alarms and can trigger equipment interlocks, achieving “identify at once, act on the spot”.
So far the platform monitors more than 80 coal faces nationwide, underpinning a new normal of “plan on the surface, execute automatically, zero people inside the face”. At Huanglong No. 1 Mine of Shaanxi Coal Group, AI video analysis has lifted overall face productivity by about 15 % and delivered a step-change in safety.
“Previously we needed three or four people roaming the face at random times; now the surface control room sees everything live and the system handles anomalies itself,” a site engineer said. Since start-up the platform has prevented multiple accidents caused by machine collisions or unauthorised entry.
Tianma notes that the solution piggybacks on existing CCTV infrastructure, so no massive hardware overhaul is required—low cost, high reliability, easy rollout. With its closed-loop sense-and-control design, rapid low-cost deployment and rugged scene adaptation, it is becoming the core system for intelligent perception, linkage and early warning in coal enterprises. Going forward, CCTEG Tianma will deepen the fusion of AI and multimodal large models with mining scenarios, push recognition technology in ever-tougher conditions and keep powering high-quality development for the industry.
