Simplified discrete event simulation for heavy construction operations

Although discrete event simulation has high potential as a management tool for construction projects, the application of simulation has been limited in the construction industry. The major reasons for this stem from the difficulty of the complex model-building process required for computer simulation and attaining meaningful simulation-related data. To support the utilization of discrete-event simulation in the analysis and the planning of heavy construction operations, the primary objective of this research is to establish a framework for a more simplified simulation environment. As a new approach for a more simplified simulation modeling environment, the resource operation, cycle-based modeling approach is identified and designed to allow modularized model-building and step-by-step guidance in model building. For this approach, six basic types of state classes are identified and designed. The six state classes embed knowledge about the construction resource behavior that corresponds to the operation structure. According to a given operation structure, they duplicate the behavior of construction resources. To support the behavior of each state class, the operation structure is identified as static and dynamic. The information about the operation structure supports the work of each state type as being self-controlling objects which allow object oriented simulation to occur. Common knowledge for simulation model-building and experimentation in heavy construction operations is identified and consolidated into the knowledge processing module to guide the project manager in the general simulation procedure. The object-oriented database for construction resources is investigated in order to support the preparation of simulation-related data. The user interface is designed to support knowledge and to allow project managers to respond interactively to the guidance of the knowledge-processing module. All of these functions are established in a prototype system, and the interactions among the system components are described. This prototype system was developed to provide a framework for more simplified discrete event simulation in heavy construction operations. It is called Knowledge-embedded, MOdularized Simulation system (KMOS). KMOS is applied and evaluated using real data from construction projects. KMOS' merits, limitations and further improvements are discussed.