🤖 AI Summary
Accurate project duration and cost forecasting remains challenging in resource-constrained, highly interdependent task networks. This paper proposes a graph neural network (GNN)-based joint temporal–cost prediction framework. We construct a heterogeneous activity–resource dynamic graph that explicitly encodes task dependencies and resource allocation constraints. To overcome the static assumptions of traditional Critical Path Method (CPM) and Program Evaluation and Review Technique (PERT), we design a resource-aware heterogeneous graph representation learning approach, enabling bottleneck identification and interpretable critical path analysis. The framework integrates GraphSAGE for structural encoding and Temporal Graph Networks (TGN) for modeling evolutionary dynamics. Evaluated on synthetic and benchmark project datasets, our method achieves 23–31% lower absolute error and an R² of 0.91 compared to conventional approaches, demonstrating substantial improvements in predictive accuracy and adaptability to dynamic scheduling scenarios.
📝 Abstract
Accurate prediction of project duration and cost remains one of the most challenging aspects of project management, particularly in resource-constrained and interdependent task networks. Traditional analytical techniques such as the Critical Path Method (CPM) and Program Evaluation and Review Technique (PERT) rely on simplified and often static assumptions regarding task interdependencies and resource performance. This study proposes a novel resource-based predictive framework that integrates network representations of project activities with graph neural networks (GNNs) to capture structural and contextual relationships among tasks, resources, and time-cost dynamics. The model represents the project as a heterogeneous activity-resource graph in which nodes denote activities and resources, and edges encode temporal and resource dependencies. We evaluate multiple learning paradigms, including GraphSAGE and Temporal Graph Networks, on both synthetic and benchmark project datasets. Experimental results show that the proposed GNN framework achieves an average 23 to 31 percent reduction in mean absolute error compared to traditional regression and tree-based methods, while improving the coefficient of determination R2 from approximately 0.78 to 0.91 for large and complex project networks. Furthermore, the learned embeddings provide interpretable insights into resource bottlenecks and critical dependencies, enabling more explainable and adaptive scheduling decisions.