Multi-modal modelling – A new look

The project produced a classification of differing modelling approaches and subsequently commissioned assessments of these approaches by transport research practitioners and other interested parties. This canvassing was undertaken in order to understand how the practitioners viewed the state of play of their art of modelling, and these views together with other material was used to produce a report outlining modelling possibilities. Focusing on the topic of travel, the many variations and techniques of the four stage approach to modelling, originating in the late 1950's, is still viewed as the mainstream transport research modelling tool covering the stages of generation, distribution, mode choice and assignment.

The main characteristic of this modelling approach is that it implies:

• a zonal basis (spatial aggregation);
• demand is measured by trips or tours;
• a static or cross sectional structure (all input fixed in time);
• a structured set of choices;
• and a requirement in principle to iterate to equilibrium.

Even discrete choice or disaggregate models are seen as an enhancement of this model and not a departure from the four stage approach. Choice of time (potentially seen as fifth stage), or trip linking again is seen as enhancement to this mainstream approach. The alternative approaches identified, developed enough to be considered as mainstream, include the 'activity' model, 'dynamic' model', and 'land use/ transport' . As well as these, mainstream type approaches, other approaches were also considered in this study. No new approach has been identified, but significant advancements in the techniques (especially micro simulation) of implementing the approaches have taken place, especially with the advancing power of computing which has enabled earlier theory to be realised. The power of the latter is thereby enabling more holistic approaches to be developed which place the transport system within wider socio economic contexts, for example the inclusion alongside the topic of travel, the topics of car ownership, and population and employment characteristics.

The objective of this integration theme project is to review multi modal modelling, taking into account the development of modelling techniques, the supply of data, the possibilities of computation and the availability of personnel able to develop and use models.  The project produce

Policy implications

The convergence of differing approaches is deemed worthy of encouragement, with a word of caution, that the step change in complexity in leading edge projects might be resulting in due diligence not being paid to individual components. The project recommends that an appropriate balance is struck between over simplicity and excessive complexity, bearing in mind the requirements for the state of the practice (as opposed to the state of the art). The complexity of models is an important issue in relation to considering how they can be brought into general use. This caution is expressed in light of the concern that a gap already exists between state of art and practice. In addition, the project recommends that consideration be paid to whether ever increasing complexity is justified in terms of either a comprehensible model structure or some demonstration that the model's predictions accord with actual outturns.

Thus the project especially contributes to knowledge in the integration sub theme of policy features, as the multimodal model serves primarily as a tool by policymakers to better understand the conditions for interoperability.