Evaluation of Project Workflow Control Strategies in Complex Organizations using Agent-Based Simulation

Cross-functional organizational projects are frequently challenged by executional uncertainty, complex interdepartmental dependencies, and constrained availability of role-specific human resources. This study presents a discrete-event, agent-based simulation framework designed to assess the effectiveness of three workflow control strategies—Static, Reactive, and Predictive—in managing task execution within resource-limited organizational environments. The simulation model emulates a structured procurement process comprising ten interdependent tasks executed by five functional roles: Engineering, Procurement, Finance, Legal, and Compliance. Each task is characterized by variable execution durations, precedence constraints, and department-specific resource requirements. Agents represent staff roles with finite availability, and tasks are processed in a non-preemptive, first-come-first-served manner. Control strategies are implemented as dynamic scheduling policies: the Static strategy follows a predetermined execution sequence; the Reactive strategy adapts in response to observed task delays or resource conflicts; and the Predictive strategy utilizes short-term historical performance data to forecast task durations and proactively adjust the schedule. The experimental design spans multiple configurations, varying control strategy type, task duration variability (low, medium, high), resource availability (full, moderate, limited), and project load (single vs. concurrent execution). Each configuration is simulated 100 times using Monte Carlo techniques to capture performance distributions. Key performance indicators include total project duration, average task delay, agent utilization rates, and rescheduling frequency. Simulation results demonstrate that the Predictive control strategy consistently outperforms its counterparts, particularly in high-variability, low-resource scenarios. It achieves this by anticipating future delays and smoothing resource demand, thereby minimizing downstream bottlenecks and task queue congestion. Analysis further reveals that highly central, resource-constrained roles—most notably in Finance and Legal—are primary contributors to systemic delay propagation. The proposed simulation framework offers an extensible methodology for analyzing general workflow processes in organizational systems and provides data-driven insights for optimizing project executions.