Article ID: | iaor200969554 |
Country: | United States |
Volume: | 55 |
Issue: | 8 |
Start Page Number: | 737 |
End Page Number: | 746 |
Publication Date: | Dec 2008 |
Journal: | Naval Research Logistics |
Authors: | Suzuki Yoshinori |
Keywords: | energy, artificial intelligence: decision support |
Fuel optimizers are decision models (software products) that are increasingly recognized as effective fuel management tools by U.S. truckload carriers. Using the latest price data of every truck stop, these models calculate the optimal fueling schedule for each route that indicates: (i) which truck stop(s) to use, and (ii) how much fuel to buy at the chosen truck stop(s) to minimize the refueling cost. In the current form, however, these models minimize only the fuel cost, and ignore or underestimate other costs that are affected by the models' decision variables. On the basis of the interviews with carrier managers, truck drivers, and fuel-optimizer vendors, this article proposes a comprehensive model of motor-carrier fuel optimization that considers all of the costs that are affected by the model's decision variables. Simulation results imply that the proposed model not only attains lower vehicle operating costs than the commercial fuel optimizers, but also gives solutions that are more desirable from the drivers' viewpoint.