MEASURING TIME SPENT IN VALUE-ADDING WORKSPACES USING SMARTWATCHES

This study addresses the lack of procedures for automatically measuring the share of time that construction workers spend on value-adding activities as a way to automate the work sampling technique. While previous studies aimed to automate this technique by focusing on activity recognition using sensors or video-based technologies, this research is concerned with identifying workers' locations on job sites using location-based sensors embedded in smartwatches. For this, the authors conducted a case study, which aims to measure the share of time workers spent in different outdoor workspaces. The study was carried out on a renovation project and involved five steps: (1) clarifying the workspace categories (production, preparation, and transportation); (2) data collection of carpenters' locations using geographic data points collected by smartwatches during 7 days; (3) data extraction and data aggregation; (4) data cleaning; and (5) data analysis using a Python script to automatize the classification of the data points into workspaces. The main contribution is a visual tool to visualize workers' positions on the job site in 2D. This information can be useful to indicate how many hours per day they spend in different workspaces and to understand the nature of a given construction activity.