The breakout session #7 at the 3rd VREF Conference led by Edoardo Marcucci and Gitakrishnan Ramadurai discusses the role of models in urban freight as instruments to search for solutions to problems in a complex and heterogeneous context where different stakeholders have different preferences, perspectives, and interests. Three specific issues are considered:
- What are the most appropriate models for the different issues the various stakeholders are faced with and what are the challenges that needs to be addressed first?
- What are the most appropriate models in a context where there is a structural lack of necessary and much needed data?
- How can we interlink various complementary modelling efforts and profit from cooperation and interdisciplinary approach?
People are splitted in 3 groups discussing the three issues mentioned above. Each group chooses a member (rapporteur) who keeps track of the considerations raised and summarises the main results to all participants. A final discussion concludes the session.
Group 1 (8 people) – models and challenges
The group identifies 6 main stakeholders (i.e. shippers, carriers, receivers, developers, policy-makers, citizens) and 3 main issues (i.e. behaviour, operations, policy-making). As it is for behaviour, random utility models and stated preference methods are considered the most suitable techniques. Private stakeholders can be used those models for marketing purposes, while the public sector is also interested in behaviour change where gamification techniques can be usefully adopted. As it is for operations, several models can be used especially for the private sector (vehicle-routing, demand forecasting, simulation, optimization models). Citizens can be very interested in the outcomes of those models, based on ICT, in terms of emissions, noise and routes. As it is for policy making, the public sector needs to have a clear picture of the actual situation and the likely changes due to possible intervention measures so to balance the various stakeholders’ interests.
Group 2 (7 people) – Data for models
The group acknowledge that there are several types of data that are not accessible. Most data are private thus not available due to markets and competition reasons. This issue should be addressed so to favour the open-access process. Surveys are often very costly. Technology can help acquiring more data (routes, parking, types of goods, types of vehicles, time of delivery, etc…) in a less expensive way. There is an issue of seasonality (i.e. a potential variation of data in given periods of time) that should be taken into account. Integration of various sources of data (e.g. insurance companies, private sectors, etc…) is very important and should be fostered, as well as sharing data (open data from policy makers is aggregate and not disaggregate, thus less useful for modelling purposes).
Group 3 (6 people) – Complementary models
Modelling is usually economic-centred (e.g. random utility maximisation, random regret minimisation). It should include also theories from sociology, anthropology, psychology, etc. Models dealing with longitudinal data can be very useful to dynamically understand changes instead of just having a static picture. Participatory design workshops and role-playing games can be adopted to better investigate choice motivations. Moreover, spreading opinions might have effects on policy effectiveness, dynamics and status quo bias, thus network and social analyses can help examining it. There is an issue of model transferability especially when considering developing and developed countries.
Final discussion
The main message that emerges from group 1 is the need of tailoring models jointly taking into account stakeholders and issues, while for group 2 is to start integrating the already-available data, and for group 3 is the need for holistic approaches especially for modelling behaviour change.
