Artificial Intelligence for Developing Accurate Preliminary Cost Estimates for Composite Flooring Systems of Multi-Storey Buildings

• Published in: Journal of Asian architecture and building engineering Vol. 21; no. 1; pp. 120 - 132
• Main Authors: Elhegazy, Hosam, Chakraborty, Debaditya, Elzarka, Hazem, Ebid, Ahmed M., Mahdi, Ibrahim M., Aboul Haggag, Said Y., Abdel Rashid, Ibrahim
• Format: Journal Article
• Language: English
• Published: Taylor & Francis 02-01-2022; Taylor & Francis Group
• Subjects: artificial neural networks; cost estimation; decision-making; floor systems; preliminary design; structural optimization; Value engineering
• ISSN: 1346-7581; 1347-2852
• DOI: 10.1080/13467581.2020.1838288

In a decision-making study, design alternatives are compared with respect to cost, performance, and reliability, and the best is selected. Often, due to the time constraints imposed by the design schedule and budget restrictions, the number of design alternatives considered is limited. This research focuses on composite flooring systems for multistory buildings and the application of value techniques to the construction industry. The study uses data that obtained from the RSMeans Assemblies Books for the period 1997-2019. The data obtained from RSMeans consists of assembly cost ($/sf of floor) as the dependent variable; and the structural span (ft.), superimposed load (p.s.f), unit cost of sheet metal ($/LF of sheet), unit cost of concrete ($/CY), unit cost of steel structural ($/ton), and total load (p.s.f) as the independent variables. A simple computer model is designed for recommending the optimal composite flooring system of a multistory building, during the preliminary design stage. The value engineering (VE) team to achieve the VE goals mentioned above can use the model.