The European Union has introduced the concept of Industry 5.0, which focuses on human-oriented manufacturing. It emphasizes the safe integration of humans, machines, and the environment in manufacturing. This results in a human-on-the-loop and human-in-the-loop smart manufacturing system where the human role is irreplaceable. The safety management of the production site in manufacturing workshops is crucial for ensuring human-oriented manufacturing. Autonomous unmanned workshops include humans in achieving key performance indicators through human–robot–environment collaboration, while ensuring production safety. Digital twin (DT) technology provides an effective way to control human-oriented workshop production safety. However, the production safety management of manufacturing workshops is a knowledge-intensive process, and existing DT workshops need to have autonomous reasoning ability for unsafe states. Unlike traditional safety management methods that rely on human subjective experience, this approach accounts for the complex interactions between humans, machines, and the environment. In light of this, this chapter primarily introduces the ontology-based reasoning method for insecure states of DT workshops and the Human–Robot Collaboration (HRC) security control method of the DT workshop, providing necessary technical support for the production safety control of human-oriented DT workshops.