A framework for offsite construction manufacturing process improvement through digitalization and automated real-time data collection

Abstract

This thesis proposes methods to apply real-time data to improve manufacturing operations in offsite construction using a combination of tools such as simulation, machine learning, and exploratory data analysis. These methods are applied to increase flexibility in operations and digitalize the design, planning, and manufacturing phases in offsite construction. Through empirical implementations, the proposed methods provide applications of digitalization to improve offsite construction manufacturing operations by integrating existing information systems while combining and applying design-related and real-time data gathered from shop floors. The proposed framework includes implementation of digitalization strategies in the design, bidding, and procurement phases of offsite construction. The proposed framework also integrates building information modelling with lean-based metrics based on real-time data gathered using radio-frequency identifier sensors logistically installed on a shop floor The proposed method addresses improvement measures using a combination of expert’s knowledge and statistical analysis, thus reducing the impact of personal opinions with respect to proposed changes to the shop floor in favor of empirical analysis supported by actual production data. A simulation-based analysis is also introduced to address the impact of the proposed framework in terms of increasing labour flexibility and reducing the average cycle time of projects in offsite construction. The key contribution of this research is the development of a framework and methods to improve manufacturing operations in offsite construction by leveraging digitalization and real-time data.