The implementation of the double prevention mechanism is an effective approach to addressing prominent issues such as"failure to recognize"and"failure to anticipate"in production safety. However, the construction and operation of the double prevention mechanism are severely impacted by issues such as lagging data informatiza-tion, data gaps, and difficulties in verifying data authenticity during the implementation of risk-based control and hazard identification processes within organizations. To tackle these challenges, we propose a solution based on AR ( augmented reality) technology, integrating Web application development, Android application development, and WebRTC technology. Our primary focus is on the double prevention mechanism of risk-based control and hazard i-dentification. An information-driven platform, comprising both Web and AR glasses applications, has been devel-oped to facilitate real-time data management. The platform has undergone preliminary validation within enterprises, demonstrating its capability to activate the operation of the double prevention mechanism in a timely and effective manner. Our approach improves the analytical and decision-making efficiency of organizations in dealing with safety risks and should help to ensure the implementation of safety responsibilities.