The construction sector faces significant uncertainties, including political and social concerns, with infrastructure projects particularly prone to chaos. This study addresses the often-overlooked construction safety hazards in planning methods by integrating safety events and their impacts on project delays.
The study utilizes a combination of stochastic processes and neural network models to develop a novel approach to deal with worker safety in construction planning methods. The proposed method was applied to a real-life database from infrastructure projects. The stochastic model helps in understanding the probabilistic nature of safety events, while neural networks forecast bivariate time series to predict the impact of these events on project delays.
The results demonstrate that neglecting safety variables leads to a misconception that safety occurrences are inevitable. Conversely, by employing the stochastic approach, construction planners gain a better understanding of the implications of safety events on delays. Validation scores indicate that the neural network model outperformed the statistical method, achieving nearly double the accuracy metric.
This study focuses on infrastructure projects, helping construction planners with tools to reduce accidents and fatalities while speeding up construction. Future research could expand the methodology to other types of projects and more diverse datasets.
This research leverages advanced artificial intelligence (AI) approaches, specifically neural networks, to enhance predictive accuracy and provide deeper insights into the complex interactions between safety events and project delays. The integration of these AI techniques represents a significant advancement in construction planning methodologies.
