This research is divided into two parts. The first part is to identify the hazards of on-site operators through image recognition technology, such as personnel falling abnormally, identifying dangerous areas, and so forth. The second part is to monitor the hazardous gas sensing data in the factory environment. Smart analysis modules with measured data establish smart hazard early warning analysis and notifications through big data or AI technology. In this study, the technology was tested in a domestic small and medium-sized enterprise, and the function of research and development was tested in the field to evaluate the functional practicability and feasibility of this technology in the future application of the small and medium-sized enterprise. Intelligent hazard early warning analysis and notification technology includes personnel detection and abnormal falling of personnel in dangerous areas. In terms of industrial safety, leakage can be discovered in the early stage of abnormal gas leakage, and the manager will be notified immediately of the early warning or the result of the hazard; the popular Line communication APP is used to respond. This study provides workers with the ability to check the working hours in each working environment and the monitoring concentration at the time through smart devices, establish the relationship between the exposure dose of the working environment and possible occupational hazards, It will serve as a reference for the prevention of occupational diseases, and effectively improve the worker self-health management awareness.

Image-based operator hazard identification module architecture