Road Transport Vs. Ro-Ro: a Modellistic Approach to Freight Modal Choice
G Bernetti, M Dall'Acqua and G Longo, University of Trieste, IT
This paper deals with the freight transportation system at national and international level. In particular the analysis has been focused on the modal choice step within the transportation planning process. Two transportation modes have been taken into account: the road transport and the road-maritime intermodal one (Ro-Ro). Moreover only medium range trips have been considered. The behaviour of the operators has been considered at an aggregate level and also the modellistic
h4. Approach takes into account average values which could be available in a planning process at strategic level. In particular a binomial Logit model has been chosen.
According to these premises, a lot of data have been collected about both the freight transportation demand and supply. Particular attention has been paid on the attributes to be included within the utility function of both modes. Some attributes have been chosen and determined according to similar models which exist in the literature, while some new elements have been introduced in order to take into consideration particular situations such as for example the administrative process which goes with transport especially in the case of international trips. Some of these new attributes have been included on the basis of an analysis which has been carried out on some existing international Ro-Ro services in the Mediterranean.
Two models have been calibrated and validated on the Adriatic freight transportation system from Southern and Central Italy and Turkey to the Italian north east regions, Slovenia, Austria and southern Germany. The first one refers only to the Italian transport system, while the second takes into consideration the international one. In particular, the study has been focused on one of the most important north Adriatic Italian port (Trieste).
The calibration procedures gives good results for both models (R2>0,80) and also the other statistical indicators take interesting values. These results will be discussed in much more detail within the paper.
These models could be used on one hand to estimate the effects on the demand behaviour caused by some supply modification or, on another hand to evaluate the feasibility of infrastructural projects. Further developments include the increase of the data base, the extension of the choice set also to the railway and the realisation of a multinomial Logit model.
Association for European Transport