Fuzzy, neural network, and expert system hybrid method development framework for cost contingency estimation in subsea gas pipeline construction

Gas infrastructure development will be accelerated as part of the National Strategic Project since gas will be utilized as an intermediary energy source during the energy transition period toward net zero emissions. Risk-based cost contingency is important to be estimated in the early project to anticipate uncertain risk events in the subsea gas pipeline construction that may cause a cost overrun situation. The application of fuzzy, ANN, expert system, and hybrid approaches as tools to estimate cost contingency, have increased in the last decades. This study aims to identify the framework of cost contingency with Fuzzy - ANN current trend development. A thorough analysis of the literature from publications between 1993 and 2023 is used as a basis for theoretical, conceptual framework development and relevant risk factor identification. Our findings reveal that hybrid fuzzy neural network methods are modern mathematical with artificial intelligence tools using simulation processes to determine risk management cost contingency models. It is also identified with the concern of the interrelated factor of risk-based that causes cost contingency in the construction industry, which can be improved using Bayesian. In addition, the genetic algorithm is proposed to be integrated to optimize the simulation process.