Agent-Based Simulation for Productivity Optimization in Linear Pipeline Projects

Linear pipeline construction projects often face operational inefficiencies due to the sequential nature of tasks and spatial constraints between dependent activities. This study presents an agent-based simulation (ABS) framework to evaluate and compare alternative scheduling strategies for such projects. A ۱۰,۰۰۰-meter pipeline is modeled with ten sequential operations, each characterized by specific productivity rates and spatial dependencies. The simulation enforces a minimum spacing constraint between operations to prevent task overlap and conflicts. Three scheduling approaches are implemented and analyzed: Fixed Minimum Separation, Linear Scheduling Method (LSM), and a Genetic Algorithm (GA)-based optimization. The GA encodes each activity's start day, spatial buffer, and optional wait period into a chromosome, which is then evaluated using a fitness function based on total project duration, crew idle time, and work fragmentation. These strategies are applied without modifying operation productivity rates or relying on crew overtime, ensuring that improvements stem solely from scheduling logic. Simulation results demonstrate that the GA-based method achieves the most balanced trade-off between minimizing idle time and reducing project duration. The proposed model minimizes crew idle times, providing contractors and project managers with a robust decision-support tool to optimize scheduling and resource allocation in large-scale, linear construction projects.

Keywords:

Genetic Algorithm , Agent-Based Simulation , Pipeline Scheduling , Construction Optimization , Linear Projects