Cloud-based factory simulation systems have become technically affordable solutions for the effective and efficient simulation of manufacturing systems. However, a simulation cloud (SC) may be incapable of implementing uncommon sequencing and scheduling methods. In addition, ensuring the logical correctness of the scheduling method adopted on an SC is difficult. To resolve these problems, the concept of scheduling intelligence, a special type of humanized intelligence, is proposed in this study. Scheduling intelligence accounts for the subjective beliefs and judgment in scheduling jobs in a factory and can be encapsulated into extensible markup language (XML) files based on the standard data structure defined on a cloud service provider (CSP). Then, an SC uses a dedicated XML parser to convert the XML files into modules to be called by the main program of the factory simulation system adopted on the SC.