Real-Time Location Data for Automated Learning Curve Analysis of Linear Repetitive Construction Activities

Abstract

Learning curve analysis has been used for decades in construction industry to study the effect of experience on productivity, especially in repetitive jobs. Accurate and reliable estimate of time and cost are the benefits of performing a learning curve analysis. While learning curves exist in all levels of a project, detailed analysis of workers’ learning curve for short-term activities (that only span for a couple of minutes) require meticulous manual observation. This demand in manual effort overshadows the economic benefits from such analyses. Also, manual observations are time consuming, subjective and prone to human errors. This research outlines how data for automating such learning curve analyses can be gathered from a construction site. For this purpose, real-time location data using Global Positioning System (GPS) technology is collected from the workers as they perform their regular activities. Data acquired from GPS technology is used with occupancy grid analysis to calculate the amount of time spent by workers in specific area, which is used to demonstrate the spatiotemporal analysis of learning curve. A linear construction activity is presented as a case study. Results include automatic generation and visualization of learning curves for workers. The proposed method enables minute analysis of learning curves in activity level which can be directly associated to project level by following work breakdown structure. The method can be used in construction field for improving estimation, scheduling and training by project managers.