| Article ID: | iaor2003982 |
| Country: | United Kingdom |
| Volume: | 38E |
| Issue: | 1 |
| Start Page Number: | 53 |
| End Page Number: | 63 |
| Publication Date: | Jan 2002 |
| Journal: | Transportation Research. Part E, Logistics and Transportation Review |
| Authors: | Haughton Michael A. |
| Keywords: | statistics: regression, transportation: general, stochastic processes |
A drawback of daily reoptimization of delivery routes as a response tactic to randomly fluctuating customer demands concerns inefficiencies in actual freight deliveries to customers. Underlying these inefficiencies are route reoptimization's potentially erratic day-to-day changes in delivery routes, and hence, highly unstable route assignments for delivery vehicle drivers. This study's primary contributions are that it develops, quantifies, and models a robust metric for assessing the resulting requirements that the instability places on drivers to learn multiple routing assignments. A key research result is that regression analysis can be used to model the behavior of a metric.