Abstract
Efficient measurement of labor input is a critical aspect of on-site control and management in construction projects, as labor input serves as the primary and direct determinant of project outcomes. However, conventional manual inspection methods are off-line, tedious, and fail to capture their effectiveness. To address this issue, this research presents a novel method that leverages Inertial Measurement Unit (IMU) sensors attached to hand tools during construction activities to measure labor input in a timely and precise manner. This approach encompasses three steps: temporal-spatial feature extraction, self-similarity matrix calculation, and local specific structure identification. The underlying principle is based on the hypothesis that repetitive use data from hand tools can be systematically collected, analyzed, and converted into quantitative measures of labor input by the automatic recognition of repetition patterns. To validate this concept and assess its feasibility for general construction activities, we developed a preliminary prototype and conducted a pilot study focusing on rotation counting for a screw-connection task. A comparative analysis between the ground truth and the predicted results obtained from the experiments demonstrates the effectiveness and efficiency of measuring labor input using IMU sensors on hand tools, with a relative error of less than 5%. To minimize the measurement error, further work is currently underway for accurate activity segmentation and fast feature extraction, enabling deeper insights into on-site construction behaviors.