An Integrated Sensor Data and Machine Learning System for Work Progress Monitoring in Industrialized Construction

Abstract: Industrialized Construction (IC) is an approach to address ongoing challenges in the construction industry. Barriers affecting IC adoption include manufacturing inefficiencies caused by inaccurate production monitoring on shop floors. This paper presents the integrated use of machine learning and sensor data collected from IMUs and load cells installed on manufacturing platforms to monitor work progress. The research method includes the development of data collection, processing, and data analysis methods in a controlled laboratory, followed by validation on a real-world shop floor. Results present a 77% F1-score accurate monitoring through temporal windows and weighted deltas.