## Pannon Newspapers: Development of mathematical model and solution method for the optimization of single day, multiple endpoint transportation problem

The Pannon Newspapers Company delivers five types of daily newspapers, numerous marketing leaflet, and several weekly newspapers in four counties. The transportation routes were created manually based on past experience.

The production occurs at a single location, the Veszprém print. Usually, it is carried out between 10 pm and 2 am at the night. The Veszprém Diary is printed at last because it is to be transported for the shortest distance. The transportation of the daily newspaper is the most important, because according to the business policy of the company, the daily newspaper should be at the local distributors at 4 am, and should be delivered to the customers until 6 am. The leaflets are usually transported together with the newspapers, but the exact time of the delivery is not critical. Penalty is due only after 2 or 3 day delay. It means if the delivery is not possible on the first day there is still possibility to make it up.

The delivery are performed by contracted transporters but the management of the transport is the responsibility of the Pannon Newspapers. It makes possible that the transport routes and the full delivery system can be recreated flexibly. For example, new routes can be created or existing routes can be modified. To do this, the new route must be traveled to check it in real life, because the unloading points might be problematic. Consequently, although the routes can be modified, such a system is not practical which change the routes dynamically day by day.

The delivery happens in two levels. The newspapers are delivered directly to the nearby settlements from Veszprém. In case of the faraway settlements, a middle point is used for transshipment. A larger truck is used to transport the newspapers to the transshipment point where the load is distributed among several smaller trucks. The transshipment points are usually areas with roof, for example a gasoline station.

The aim of the project was to develop a mathematical model and a solver for such a Vehicle routing problem which involves a single source and hundreds of endpoints. The solution of the problem with standard methods is not possible because of the size of the problem. Consequently, a tailored solution method had to be developed. The efficacy of the mathematical model and the solution method was verified with the solution of a reference problem. The project ended in April 2008.