Construction labor productivity: impact of LC-BIM synergy and measurement with machine learning techniques

This thesis investigates the integration of LC-BIM synergy and machine learning methods in calculating and increasing labor productivity for more effective management at construction sites during the construction execution phase. In this context, this study aims to determine and evaluate the effect of LC-BIM synergy on labor productivity in the execution phase of construction works. A workflow model is generated to combine LC-BIM synergy with machine learning methods. For this purpose, first, the LC principles and BIM functions used in the execution phase of construction works were determined, and the effects of the synergies of both techniques on labor productivity were selected through a survey. Later, the information gathered from the sensors attached to the worker and their motions were analyzed using machine learning techniques. Based on the productivity of their motions, the theoretical productivity levels were then computed. Accordingly, it was determined in the study that the synergies between LC's principles of reducing variability, shortening cycle time, and using visual management and BIM system's 4D visualization and real-time construction monitoring and reporting functions are the factors that increase labor productivity the most. The motion data obtained in real-time with the help of sensors were analyzed with the XGB algorithm with an accuracy rate of 94.14%, the activities performed by the workers on the construction site were estimated, and the theoretical productivity of the activities performed was calculated. As a result, it is considered that an automatic productivity calculation based on labor and integrating this calculation into the LC-BIM system will have a significant positive impact on productivity.