Forecasting slipform labor productivity in the construction of reinforced concrete chimneys

Abstract

The aim of this study is to identify the factors that influence productivity in the construction of reinforced concrete chimney (RCC) and to develop a predictive model for forecasting slipform labor productivity in RCC projects. In this scope, 73 “daily site reports” from two RCC constructions were utilized to calculate the slipform labor productivity. The efficiency value’s estimation part for the slipform process considered factors affecting productivity under four main categories: project information, weather conditions, job characteristics, and team-related information. The independent variables were identified as: man-hour value, RCC diameter, shell thickness, daily rising height, height of the working platform, concrete quantity, rebar quantity, minimum and maximum temperatures, maximum wind speed, rain conditions, and slipform quantity. The dependent variable is the efficiency value of slipform process. For this, an ensemble machine-learning technique, Gradient Boosting Machines, was used to predict slipform labor productivity in RCC constructions. According to the analysis, the R2 value is 0.900, indicating a high level of accuracy in the model’s predictions. The study found that the parameters with the highest impact on efficiency were daily rising height and slipform quantity. To validate the accuracy of the prediction model, a survey was conducted with technical experts, and the results were analyzed to investigate the model’s accuracy.

Keywords

• Reinforced concrete chimney construction
• Slipform
• Labor productivity
• Productivity forecast model
• Machine learning
• Gradient boosting method