AI-BASED FINANCIAL RISK AND COST PREDICTION FRAMEWORK FOR CONSTRUCTION EXPERIMENTS: A CASE STUDY OF STEEL HALL ASSEMBLY
Abstract
<p>This study proposes an artificial intelligence-based framework for financial risk assessment and cost prediction in construction experiments, focusing on steel hall assembly processes. Construction experiments involving real-time monitoring technologies are associated with significant financial uncertainty. Proposed framework combines cost modeling, risk classification and data-driven prediction using a multilayer perceptron (MLP) model. Input variables include the number of structural elements, assembly duration, labor engagement, detected errors and safety-related events. The model predicts costs and associated risk levels, which are compared with actual outcomes from a controlled steel hall assembly experiment. The case study demonstrates the framework’s applicability in identifying cost deviations and financial risk patterns. Results indicate that the approach enables early detection of potential cost overruns and improves decision-making in the planning and execution of construction experiments. Integrating artificial intelligence with financial analysis can improve resource management and provide a basis for applying the framework to a broader range of construction scenarios.</p>