Thread Content
AI protects safe production and empowers safe intelligent upgrades Author/Source: Huaqiang processing date: 2023-04-11 Click rate: 61 Recently, the company's safety AI real-time analysis system (Phase I) was put into operation. The system takes the creation of safety standardization as the core and the dual prevention mechanism as the main line. It covers professional modules such as safety management, production management, and equipment management. It integrates online monitoring, AI intelligent monitoring and other technologies to build hierarchical control and centralized management. The closed-loop management and control system realizes real-time monitoring of work areas and high-risk operations, unified management of personnel, equipment, environment, etc., and automatic identification, intelligent capture, and voice warnings of unsafe behaviors or states, effectively promoting the implementation of corporate safety production main responsibilities and escorting safe production. Scenario-driven AI empowers the system to use existing and newly built video surveillance equipment to intelligently identify people, objects, behaviors, environment and other elements in the video, automatically discover "three violations" of personnel in production, abnormal equipment status, hidden risks in the production operating environment and other anomalies, and proactively trigger alarms. The first technological innovation of digital intelligence integration is to realize the closed-loop safety supervision application of "automatic early warning push". Based on real-time intelligent analysis of surveillance videos in the production area, the system can alarm nine safety events such as abnormal smoke, personnel falling to the ground, illegal smoking, and oil leakage in key production devices and areas, and push the information to the mobile phones of relevant managers. At the same time, warning messages will pop up on the large screen in the central dispatch room. The second is to realize the "four-color early warning" grid management of safety supervision. Different production scenarios are divided into four-color early warning grid areas of "red, orange, yellow, and blue" according to the safety risk management level. Intelligent alarms for "three violations" behaviors are integrated with time and space. The risk levels of the same type of events are related to the grids in which they are located, achieving hierarchical and classified management of "three violations" behaviors and risk levels. The third is to deeply integrate intelligent vision and data analysis. Comprehensive analysis from multiple dimensions such as equipment, algorithms, scenarios, behaviors, effects, and spatio-temporal distribution can realize situational awareness and trend prediction of enterprise safety production risk management, and provide digital and precise decision-making basis for safety production management.