| Truckload transportation is an important field of business that supports almost all other industrial and commercial activities in any country. Empty miles are non-revenue generating miles where the vehicle is traveling without carrying a load. These miles are expensive for the carriers and eventually for their customers. Another important issue for truckload carriers is driver turnover. Carriers who cannot retain their drivers suffer high recruiting and training costs as well as indirect costs like safety problems and customer dissatisfaction. The objective of this research is to develop an optimization model that is capable of minimizing the empty miles traveled by a carrier's fleet, and positively affect driver retention. The research problem is a multi-depot multiple traveling salesman problem with time windows. A unique attribute of this research is formulating driver satisfaction into model constraints to impact driver retention. The proposed model is solved using a pure genetic algorithm coded in Visual Basic for Applications (VBA). This algorithm uses a parameter that defines a selection pool size for the solution construction process. This parameter increases the probability of finding feasible solutions. Other operators used are: single point crossover operator and mutation rate. Full factorial design analysis was conducted on randomly generated problems to find the significant parameters. The methodology reached the optimal solution for the 20 loads test problem and showed low variation between different replications. The proposed methodology has a high potential for future study and development, however, more effort needs to be focused on the efficiency and execution time to reach a practically applicable solution. |
