European Data Set for Long-Distance Travel (ETIS-BASE)

A reusable methodology for generating a pan-European transportation dataset for long-distance travel demand, level of service, and cost.

IVT Partners

NEA
MK Metric
IWW Institute, University of Karlsruhe Germany
NESTE
ISIS
MDS
SOFRES
Technical Research Center of Finland (VTT)

Project Researchers

Prof. Dr. Kay W. Axhausen
Jeremy Hackney

Sponsors

European Union 5th RTD Framework Programme

Period of Research

05.2003 - 05.2006

Summary

The European Union needs a homogeneous transportation data base for guidance in policy making. This two-part project will develop and test a methodology for establishing a homogeneous data set and a model of the demand for long-distance travel in Europe. This demand data set and methodology are integrated with the second part of the project, which will establish a data set and model of the level of service and the cost of available transportation modes. A reusable methodology will be developed and applied to generate both the data sets and models. The results of the research consist of a report of the methodology developed for synthesizing the databases, CDs of data with applications of the methodology across socioeconomic groups and transportation networks, and users' manuals.

Demand

Aggregate and disaggregate data sources will be used to generate the demand data set, specify and check behavior models, and to calibrate the methodology. Special efforts will be necessary where no or obviously wrong information is available. This methodology cannot be completely automated because of the exceptional data handling requirements which will necessitate expert human judgement. It is therefore not an objective of this project that the methodology be in the form of an automated software product.

As much of the available data from existing EC project sources will be used as possible. The latest survey undertaken by the EC, project DATELINE, has been made available for inclusion into the methodology and data generation process. The data will be analysed to explain consumer behavior and to fill gaps where observations are lacking. Surveys like DATELINE have never been considered in the data generation process. But before such surveys are of any use, they have to be analysed, adjusted, cleaned, corrected and enriched.

1. In-depth data source analysis and preliminary collection. The relevance of currently available data sources to the definition of transport demand parameters will be appraised. The information so far collected include: long-distance travel choices in the DATELINE database; air, road, and rail network models from the BAK project; Labor force data sets and data sets on consumption for estimating weighting parameters; community boundaries and centroids from SABE; supplemental information on rail connections at the community level from national railways.
2. Data model development. A unified data model will be designed on the basis of input from partner SOFRES about the data sets required by the modelling component. This will enable the further provision of the highly organized information for the reference database. Data will be checked for consistency with socio-economic data. A reusable methodology to consistently generate the complete transport demand data set (data, indicators etc.) will be defined at this stage.
3. Final collection of the passenger transport demand data.
4. Generation of a pilot set and recursive check. The methodology will be applied to a set of selected data sources as a proof-of-concept that the methodology works. For example, the Dateline data set will be cleaned and completed with suitable small area statistics, models, and network data. The lessons learned from applying the methodology will be collected and the methodology will be corrected accordingly.
5. Compilation of the passenger transport demand data set. The compiled data will be inserted into the ETIS reference database.

Supply

In addition to the demand information, modal attributes have to be identified and sampled to generate the competing alternatives which influence consumer or shipper behavior. The attractiveness of the different modes leads to a traveller's decision. Evaluating the direct or indirect influence of transport policies requires analysis of the interdependence and structure of the level of service, cost, and demand indicators.

Generating a data set is difficult, as we are dealing with information of a highly competitive market where prices and services are fluctuating (e.g. on hourly base in aviation). As good as no homogenous data source exists, either across modes or across the member states. Often different national offices with different guidelines deal with the same problem on different levels of quality and technical depth. Also, the shippers and the operators are not keen to deliver any commercially sensitive information.

After data collection, the data model has to take into account that the LOS and the costs are subject to the competitive situation between modes. Not only the air mode, where such approach has been applied for a long time, but more and more profit-maximizing tariff structures are also seen in the long-distance rail services . Services for air and long-distance rail have been becoming more and more demand-driven; thus the link to the demand model. The link to the demand model is also essential in order to ensure that the ETIS-BASE database as a whole becomes a highly consolidated and consistent information system.

A pilot data set will be generated and will be subject to analyses and checks in terms of consistency and plausibility. Data gaps will be identified and methodologies of how to fill these gaps will be elaborated.

Finally, the reference data set for passenger LOS and cost will be generated by carefully taking into account the lessons learned from the tasks carried out earlier.

Links

ETIS-BASE
ETIS-LINK