Simulating travel duration data for Flanders

Abstract

The growing need for timely, high quality and large amounts of data and information from national statistical agencies has increased continuously over the years. These data are fundamental in enhancing modern transportation planning and policy development endeavors. Provision of large quality data on travel demand related to the socio-demographic and travel characteristics of individuals and households, largely relies on household travel surveys (HTS). Nevertheless, it goes without mention that conducting these surveys is continuously characterized with numerous problems. The surveys involve high expenses, necessitate a lot of time to plan and implement and further impose a high burden on the respondents, subsequently bringing about low response rates. This often has severe consequences on the quality and representativeness of the resultant collected data. It is fair to believe that even when more advanced and recent technologies such as the global positioning system and personal digital assistant (Murakami et al., 2000) are used, the final total cost will only become higher. In search for plausible approaches to solve these problems, combining travel data with data from other sources becomes an attractive option. Furthermore, the success in research directed towards simulating travel data is of absolute interest in this respect. Simulating household travel survey (HTS) data (Greaves and Stopher, 2000; Stopher et al., 2003; Pointer, et al., 2004) is a relatively fresh field of research with many potential benefits. The benefits of this approach would be enormous to all actors who utilize HTS data and could potentially provide a very low cost technique for generating a local sample of many additional households in comparison to collecting household travel data. Simulated data would also conceivably form a component of the HTS currently being undertaken, and facilitate these surveys to reduce sample size requirements without necessarily compromising the quality of the planning activities supported by these data. However, this area of research is still in its infancy and significant work still has to be done. Such work should be geared towards developing a state-of-the-art technique, testing and also establishing the clear role for household travel data simulation. In past research (Nakamya et al., 2007b), promising results with regards to trip rates data simulation were observed. In general, the procedure provided results that were comparable with results from an actual travel survey.