In enterprises, many processes are implemented to achieve the defined business goals. These are mainly cyclical processes, that is, processes that repeat and have a designated course. They usually differ in details, such as the source and destination, or where they appear in the system. When modeling in simulation programs, these differences significantly affect the advancement of the model. In the process of reflecting processes in programs, there is no possibility of automatic operation – “artificial intelligence” responsible for cyclicality. The aim of this work is to present an approach to modeling – the so-called reference process that allows one to simplify and shorten the design time of logistics processes in simulation models. A solution was proposed in which only the rules for selecting the so-called locations acting as buffers, storage fields, and modeled process activities are generated automatically. The proposed rule allows for modeling according to random rules, indicating further directions for research related to quantitative and time rules. The article also presents the implementation of this solution in the LogABS simulation program.