Developing a system for construction site layout planning using genetic algorithm and building information modeling

Abstract

Effective construction site layout planning (CSLP) is essential to ensure site safety and enhance work efficiency. The process of CSPL involves identifying temporary facilities (TFs) required to support construction activities, determining their sizes, and optimizing their locations based on pre-determined objectives. The CSLP problem is generally formulated as an optimization problem, where a specific objective can be achieved with a set of constraints. The complexity of construction site layout planning increases due to the management of a significant quantity and variety of temporary facilities, the required areas for each temporary facility, the locations and quantities of construction cranes, and additional constraints. Many previous CSLP models have been developed by optimizing the locations for TFs while assuming that tower cranes are fixed at predetermined locations. Moreover, manually inputting and interpreting positions and areas of TFs from the optimization model can be laborious, time-consuming, and susceptible to human error, which may potentially lead to incorrect placement of TFs. In this study, an automated system is proposed to generate an optimal site layout. Genetic Algorithm (GA) model is used to develop the optimization model that searches all free areas and finds optimal locations for TFs and tower cranes simultaneously. Then, the model is integrated with BIM to develop an automated system. BIM is used to create geometric data for the system input and update the facility locations based on the output from the optimization model. The system is tested on two examples, and the performance is analyzed by each fitness value and closeness relationship between facilities. The proposed system is validated internally by using a case study in a previous study, and then externally validated by two case studies from real construction projects. The comparison between existing site layouts and the layouts generated by the system is conducted. The results indicate that the system can optimize locations for TFs and tower cranes simultaneously. The proposed system will be beneficial for project managers who wish to generate optimal site layouts automatically. Instead of solely relying on past experience and common sense, the system can facilitate project managers to optimize and visualize their optimal site layouts.