A PROBABILISTIC MULTI-PARAMETRIC BID EVALUATION SYSTEM FOR CONSTRUCTION PROJECTS USING BELIEF NETWORKS TECHNIQUE

Abstract

One of the most difficult decisions in the construction industry is selecting the most suitable contractor to construct the project under consideration. Despite of the shortcomings associated with the low-bid system, the majority of construction contracts abroad and in Egypt are awarded using such system, where the bid price is the sole criterion in determining the successful contractor. To improve the quality of the bidder selection process, this research introduces a

. newly developed system that benefits from the intelligent inference capabilities of Bayesian Belief Networks (BBN). The main concept of this system follows the general concept adopted in earlier multi-parametric bid evaluation systems, which is based on the calculation of equivalent monetary values of various bidder parameters, e.g., time, quality, safety, and work attitude and practices, while taking into account the necessary alterations and adjustments needed to overcome the pitfalls of the existing bid evaluation systems. These monetary values are then added to the bid price after performing the necessary adjustment for the advance down payment to calculate the total combined bid value (TCBV), which can be used to identifY the most appropriate bidder for the project. For the research implementation, several surveys were conducted for data acquisition from several, industry experts in three sectors, i.e., residential buildings, industrial projects, and roads and highways. An automated system was then developed for this probabilistic multi-parametric

bid evaluation approach. This automated system, known as "BN-BID", is primarily composed of four integrated modules that handle adjustment of bid price, calculation of the equivalent monetary values of various evaluation parameters, prioritization of the evaluation parameters. and calculation of the total combined bid value (TCBV). BN-BID is entirely developed via Microsoft Access™ and the programming capabilities of Visual Basic Application VBA™. All four modules of BN-BID are encompassed in a uSer-friendly interface to facilitate the utilization of the developed system The system is finally tested and validated for proper adaptability and functionality using two case studies. The first case study represents a residential building project while the second represents an industrial project.