State Grid Hubei Information & Telecommunication Company Empowers Smart Safety Supervision with AI

Recently, to further advance the automatic identification of safety violations and real-time risk warnings, and to effectively improve on-site safety management, State Grid Hubei Information & Telecommunication Company has conducted a comprehensive review and systematic upgrade of its AI-powered intelligent violation detection system. Leveraging the integrated reasoning capabilities of the Guangming Large Model and driven by a dual-engine approach of "business + technology," the company continues to optimize the functions of its smart safety supervision platform.

Screenshot of the safety supervision video

The platform enables intelligent applications in scenarios such as typical violation detection, intelligent video diagnostics, high-risk worksite screening, and safety qualification access control. Currently deployed in Wuhan, Jingmen, Enshi, Shennongjia, and other regions, it can accurately identify common violations such as not wearing a safety helmet, improper attire, and working on ladders without assistance.

The platform also covers 7,140 cameras across 1,672 substations in Hubei Province, enabling intelligent inspections for unplanned operations and strengthening work plan compliance management. In addition, it features video quality diagnostics and intelligent camera angle correction, which automatically alert and adjust cameras that are not properly aligned with the worksite. The system simultaneously generates video quality reports to support the continuous optimization and efficient operation of equipment.

Screenshot of high-altitude work supervision at a substation

In the next phase, State Grid Hubei Information & Telecommunication Company will continue to deepen the application of AI in violation detection, safety access control, and work plan management. The company aims to promote the deep integration of artificial intelligence with its safety management system, expand large-scale application of typical scenarios, and build a stronger digital defense line for production safety.