To better understand how practitioners are using formal tools and where they need to be improved, a targeted survey was conducted online with 16 professionals working in transport from a variety of regions within the European Union.
The four step model is the one most widely used. The professionals surveyed were satisfied overall with four step models as forecasts are sufficiently reliable, taking into account the data that are available. That said, these models do not have the ability to fully understand chain trips. To clarify, chain trips are trips that, on the surface, may seem to be a simple home-work or work-home commute, but which actually include multiple stops, such as the kids’ school, the gym, the supermarket, and so forth. The reason why these extra stops are important is because they reduce people’s flexibility, which in turn affects how they will respond to incentives. Four step models that reduce the complexity of daily trips will therefore tend to overestimate the impact of certain measures, such as road charging, on people’s behaviour.
Practitioners are aware of the potential of activity based models, which are capable of modelling the emergence of new social trends and technologies and new types of transport policies. However, activity based models are not always practical as they require more resources for data collection and more computing power, and as a result are costlier.
A major issue is that there are still a lot of gaps in the data, particularly where the cities and regions in question do not control all aspects of mobility and therefore might not have access to information that is collected by transport operators. Surveys tend to focus on peak travel, leaving travel during non-peak times or by non-commuters and tourists (useful for estimating environmental impact) underrepresented. In addition, travel surveys often do not include parameters related to personal values and attitudes that may underlie the fundamental motives for modal choices. Another challenge involves how quickly technologies, and thus behaviour, are evolving. The impact of emerging technologies, such as autonomous vehicles, and “sharing” services is not yet fully understood. Major societal and technological shifts, along with changing lifestyles and values, still need to be studied in depth. Better data and a better grasp of mobility behaviour could go a long way towards improving the accuracy of these models.
Mobility lifestyles are becoming increasingly diverse, and simple variables are not enough to accurately forecast behaviour. This kind of forecasting is particularly important if policy makers wish to induce behavioural change. But new data sources and technologies, such as GPS information from mobile phones, public transport smart cards, data from sensors in the vehicles and technologies used to enforce traffic schemes, could fill some existing data gaps and provide information on behavioural responses to policy or infrastructural changes. The question of models and tools, and how these can be improved, will be looked at more closely in future MIND-SETS activities.
For more information, contact Laurent Franckx of VITO