Estimating the length of trunk tours for environmental and cost evaluation of distribution systems

Summary

During the last 10 years the road freight traffic in Germany, measured by vehicle miles, increased dramatically by about 41%. About 40–50% of this mileage is estimated to be caused by the distribution of consumer goods. Thus, the decisions of the single industrial companies on their distribution systems have a strong influence on the total freight traffic and on the environment. In a distribution network the majority of the traffic (about 70–80%) is caused by the delivery from the warehouses to the customers, commonly served in tours. Therefore, we distinguish between local delivery tours for small order sizes over short distances and direct deliveries of orders with a large size in trunk tours. Usually, in the framework of strategical distribution network design the assignment of the customers to the warehouses or transshipment points is optimized on the base of direct distances between warehouse and customers; the real length of the tours are not regarded. Considering that most of the environmental parameters for evaluating the traffic in distribution systems (e.g. fuel consumption, emissions) behave nearly proportional to the vehicle miles, the estimation of the tour length is very important. Futhermore transportation costs are strongly dependent on the tour length. The investigation aims at estimating the length of direct deliveries in trunk tours. We focus on the presentation of an estimation model, and first computational results are shown comparing the accuracy of the model with real tour data of a consumer goods manufacturer. In addition, we show numerical examples for evaluating the trunk traffic by environmental measures and by costs.