A mixed-integer linear programming approach for temporary haul road design in rough-grading projects

Temporary haul road plays a pivotal part in achieving cost-efficiency and successful project delivery in heavy civil and industrial construction. Temporary haul road layout design has been empirically performed by field managers, superintendents or even truck drivers largely based on experience instead of science. Previous research endeavors in earthmoving research devised analytical algorithms to minimize the total earthmoving cost and developed simulation models to optimize earthmoving resources and processes. In the same domain, this research introduces an optimization methodology for temporary haul road layout design in order to facilitate mass earthmoving operations and improve construction performances on a typical rough-grading site. A Mixed-Integer Linear Programming model, integrated with a cutting plane method, is established for the identified problem. In particular, the cutting plane method is introduced to refine the optimization formulation by maintaining field accessibility and haul road continuity, thus ensuring practical feasibility of the analytical solution. Performances of the newly devised algorithms were benchmarked against genetic algorithms in solving ten test cases. Further, feasibility of the proposed methodology was evaluated based on a real-world case study. In conclusion, the proposed methodology is capable of tackling the temporary haul road layout design problem with high computing efficiency and delivering optimal results that are ready for field implementation.