Recently, the Shangwan Coal Mine under Shendong Coal has built the Company’s first AI-powered gas board identification system, achieving automatic and intelligent gas monitoring data collection, recognition, storage, and report generation. This innovative system has improved the efficiency of gas inspections and ensured the timeliness and accuracy of data.
The intelligent report generation function of the gas inspection system is integrated into the Mine’s intelligent ventilation and prevention system, allowing for gas inspection data queries, personal gas inspection ticket outputs, and historical data searches, along with various report outputs. The system is equipped with various functions, including real-time data display, over-limit alarms, fault reminders, logical diagnostics of abnormal data, accident statistics, and intelligent operation and maintenance & control capabilities.
Leveraging multimodal large model technology, the Mine has developed a fine-tuned dataset for gas board identification, achieving a 95% accuracy rate for gas board image recognition and a 98% accuracy rate for structural recognition. Inspectors can simply take photos or read gas inspection board images via a mobile app to receive terminal identification results and verify data, automatically uploading the confirmed information to the database.
