Using Fuzzy Inference Systems for Lean Management Strategies in Construction Project Delivery

Abstract

When using lean waste management in construction project delivery, computational methodologies are currently an innovative technology for the implementation of efficient and effective improvement strategies in the development of Industry 4.0 in Chile. Lean models are able to manage data obtained from construction projects along with the data obtained from the knowledge base of professional experts (expert survey). The waste management of construction projects under the lean philosophy requires cooperative efforts, where the opinion of professional experts is completely paramount to analyze multidisciplinary knowledge. Therefore, new protocols and disruptive procedures based on artificial intelligence (AI) tools can help decision makers prioritize activities, minimize uncertainty, and avoid wasteful actions that add no value to the project and thus can be minimized or completely eliminated. The vagueness of subjective human judgment in the degree of application of lean waste management in project delivery is modeled by a fuzzy logic model that includes additional considerations related to the lean implementation. Moreover, multiple linear regression analysis has been implemented in order to verify and validate the previous digital fuzzy model. In this sense, the main aim of this study is to develop new approaches regarding AI systems, using fuzzy sets and multiple linear regression for managing waste in construction project delivery in the metropolitan area of Santiago, Chile. A theorized application of the models reveals that the sample (100 construction projects) can be classified into three lean waste condition levels: high, medium, or low waste effects. The outcomes of this research will contribute to the Chilean construction industry environment and will open new ways for harnessing AI-based technology in the construction industry to the fullest potential, to achieve better time and cost predictability with a client- and end-user-centered world view.