Vehicle routing plan is usually provide optimal or near optimal way to serving client which several constraint, such as delivery time windows. The dynamic fleet environment in city logistic has many unexpected events when delivery was executed. We will conduct a research to develop a near real time vehicle operation system support. The future result of this research is an intelligent dynamic fleet management support system. Architecture of system will consider integrating communication and positioning technology, decision support system module and data management. The system will monitor and accept report from delivery vehicle, adjusts routing plan and calculate total cost that affected. Evaluation for this system will use simulation environment and real case study.