Delay response and risk-cost optimization in programs under the employer-furnished resource mode: a risk propagation path perspective

Purpose This study examines delay response and employer risk-cost optimization in programs under the employer-furnished resource mode, where delays affect both schedule performance and resource supply. Design/methodology/approach An integrated decision framework is developed by combining Monte Carlo Simulation (MCS), Graph Neural Network–General Regression Neural Network (GNN-GRNN), and Deep Deterministic Policy Gradient (DDPG). The framework is used to generate uncertain delay scenarios, identify inter-subproject relationships, predict resource demand, and optimize crashing and resource supply decisions in a coordinated manner. The framework is evaluated through a case study of a large engineering program, with the case data providing the baseline inputs for scenario generation and model evaluation. Findings The proposed method reduces average resource demand prediction error by 4.73% and lowers total employer risk cost by 15.0%, including reductions of 14.9% in schedule risk cost and 15.4% in resource risk cost. The results also show that, under the employer-furnished resource mode (EFR), some non-critical subprojects may have higher intervention priority and that critical path-oriented delay response does not necessarily minimize total loss. Originality/value This study extends delay response analysis from a schedule-focused perspective to a coupled schedule–resource risk-cost perspective. It shows that, under the employer-furnished resource mode, managerial attention should move beyond critical path control toward the identification and coordinated control of risk propagation paths.