Abstract: At present, safety monitoring of the assembly process of the detonation device of the weapon system is highly dependent on manual labor, leading to issues such as incomplete coverage of hazard identification and untimely hazard warnings. This conflicts with the human-centric concept in Industry 5.0. To solve the above problems, a safety warning decision-making method based on digital twin was proposed for the assembly workshop of detonation device. Firstly, taking the assembly workshop of the weapon system detonation device as a specific object, a digital twin-based safety warning decision-making framework of the detonation device assembly workshop was proposed, and the system architecture of the safety monitoring and warning decision-making for the detonation device assembly workshop was developed. Secondly, three key technologies of the proposed method were elaborated in detail: real-time information-driven synchronous modeling technology of detonation device assembly workshop, data acquisition and monitoring technology of detonation device assembly workshop, and warning and decision-making technology based on data mining and deep learning. On the basis of the above methods and technologies, a safety monitoring and warning decision-making system for the detonation device assembly workshop was developed and applied in a detonation device assembly workshop to verify the feasibility of the proposed method and system, providing new ways for the application of digital twin technology in the field of workshop safety monitoring and warning.