Freight Transport Demand Modelling

Abstract

In freight transport models, freight generation is the stage which estimates the amount of cargo generated or attracted by establishments or by geographic zones. The literature distinguishes two classes of models: on the one hand Freight Generation (FG) and Freight Attraction (FA) models, which are the production and attraction of cargo measured in tonnage (or volume), on the other hand Freight Trip Production (FTP) and Attraction (FTA) models, which regard the number of vehicle movements (Holguin-Veras et al. 2014).Generation models can be estimated with aggregate or disaggregate data. Disaggregate data is interesting because it avoids aggregation biases. It also allows, in some cases, to investigate the influence of variables which only make sense at the disaggregate level, or the presence of non-linear effects. Finally, disaggregate models can be a good basis to disaggregate aggregate data (for example, regional freight data could be disaggregated to the city level with the appropriate establishment dataset and a reliable disaggregate generation model).The estimation of disaggregate generation models requires disaggregate data at the establishment level. This data is obtained through surveys targeted at business establishments, such as commodity flow surveys. Establishments are typically described by the economic activity sector, economic size (workforce or turnover), location, and type (offices, plant, warehouse, etc.). Variables about production,

Modelling change in supply-chain-structures and its effect on freight transport demand

“…Besides the aggregation level and model focus, the models can also be distinguished by the applied methods. For an overview see for example Thaller et al (2016).…”

Logistics decisions in descriptive freight transportation models: A review

“…From the perspective of the policy maker, an accurate aggregate repre sentation of such decisions is important to anticipate responses of the logistics sector on policy measures. In the past decades, researches have come to recognize the importance of developing descriptive models of logistics decision making to build better freight transport models (Tavasszy et al 1998, Davidsson et al 2005, Tavasszy 2006, Wisetjindawat et al 2007, Liedtke 2009, Roorda et al 2010, Tavasszy and de Jong 2013, Thaller et al 2016). The argument is that models that build on an understanding of normative principles of logistics decisions are better able to describe and predict the aggregate result of firms' behavior.…”