Integrating Bond Graphs and Bayesian Networks: A Novel Decision Model for Bidders in Public Procurement

Abstract

Public procurement is a pivotal aspect of government operations, particularly for construction firms, where the decision to bid (d2b) is a recurrent and strategic choice. This decision-making process involves selecting projects for competitive bidding, which significantly influences firms’ success and reputation. However, the d2b process is complex, as contractors must navigate the delicate balance between pursuing profitable projects and avoiding the costs and risks associated with unsuccessful bids. This paper endeavors to explore the integration of BGs with BNs to develop a novel decision model tailored for companies engaged in public procurement. Our methodology includes two primary steps: Firstly, leveraging expert insights, we quantify the uncertainty associated with various variables in the bidding process using Bayesian Networks (BNs). These networks enable us to analyze expert opinions, historical data, and market trends to assess the likelihood of bid success and evaluate risks accurately. Secondly, we model the dynamic nature of factors influencing the d2b process using Bond Graphs (BGs). By capturing interactions between components over time based on expert knowledge, BGs provide insights into the evolving dynamics of market conditions, bidder readiness, buyer’s governance, tendering transparency and project characteristics. Through the fusion of BGs and BNs, this study aims to provide construction firms with a robust analytical framework to navigate the complexities of public procurement, ultimately enhancing their competitiveness and success in the marketplace.