In recent years, developing a digital twin for a complex system has been a key research effort in the modeling and simulation (M&S) community. This paper presents one of the digital twin systems for an urban garbage collection system, located in Sejong City, South Korea. This garbage collection system requires constant maintenance and management, and particularly, a rapid collection of distributed garbage is required for urban sanitation and the population's satisfaction. However, the current status-quo is following a routine garbage collection schedule, so the system management becomes rigid. Therefore, we develop a digital twin model to experiment with the garbage collection mechanism and speed to provide a simulator for further study on its operation. Our digital twin model is designed by following the DEVS formalism and calibrated by a large-scale of real-world datasets. This paper delivers the practice of designing, experimenting, and calibrating this urban system.