Improving Decision-Making System in Infrastructure Projects: A Framework to Propose Infrastructure Building Information Modeling (I-BIM) with Artificial Neural Networks

In the construction industry, countries are trying to be more careful with projects that they want to invest in due to the more restricted budget allocation that hit the world after COVID-19. Information and Communications Technology (ICT) brought a new vision to the industry under the light of Building Information Modeling (BIM) to save more time and money on huge projects, but there are many challenges in the path of utilization of this technology for other sectors of the construction industry, such as infrastructure. According to the World Economic Forum (WEF), there is enormous growth in the infrastructure budget, which is going to be needed by 2040. Now the main question is: "Does the BIM philosophy, which has been used for more than two decades, come to any sort of adoption for infrastructure projects?" The foundation of this dissertation has been built on top of this question, and it has been followed up to support it by implementing this technology for infrastructure projects and reforming a framework precisely for this sector of the construction industry. To do that, a comprehensive ontology and literature review have been done to identify the adoption of BIM for infrastructure projects and find the main categories and variables to define a new category in information modeling terms as Infrastructure Building Information Modeling (I-BIM). Moreover, a tri-axial model as a conventional BIM method regarding the Policy-Technology-Process (PTP) model has been discussed parallel to competency tire establishment in accordance with five main categories and 26 defined variables in total. Therefore, a questionnaire survey has been conducted to distribute among two target countries, the United States of America and Turkey. Afterwards, in Core Competency, detailed data mining has been performed on each variable in each category for each target country to identify those variables as benefits, neutrals, or barriers to adopting I-BIM terminology. Furthermore, the Domain Competency analysis has been completed by defining meta variables and introducing the Total Number of Meta Variables (TNMV) factor at three different levels to distinguish each meta category more precisely and accurately for each target country individually. The third and last tire was Execution Competency, which has been concluded with the assistance of an Artificial Neural Network (ANN) to train, validate, and test collected data from previous tires in order to find the best performance according to the Q-Factor (Quality) for each country and the E-Factor (Experimental) for each variable to significantly upsurge the accuracy and impact of all variables within the proposed I BIM framework Keywords: Building Information Modeling, Project Management, I-BIM, Infrastructure Project, Construction Management, Digital Transformation, Meta Variables, Artificial Neural Network, ANN.