Transportation modeling: Lessons from the past and tasks for the future

Abstract

In spite of the recent progress made in household activity analysis and travel budget studies, urban transportation modeling still remains a "not,too-well developed" research field. There are conflicting theories, analysis units are not uniform, terms are not precisely defined, basic studies of sub-systems involved are not yet completed, and many models lack behavioral background as well as basic attributes such as simplicity, sensitivity, compatibility, transferability and forecasting ability.Gaps in methodology may be partially responsible for this situation. There is an urgent need for simple, yet not primitive, easily applicable urban transportation models which can respond to the technical needs of planners and engineers. Lessons from the past, as well as experiences from other disciplines, suggest that future research should concentrate on: (1) new, "unconventional" approaches based on systematic, basic studies of all sub-systems involved; (2) proper definition and stratification of an analysis unit; (3) revision and unification of definitions, classifications, etc., in order to improve the behavioral background of the models; (4) dynamic rather than static approaches, able to describe feedbacks between transportation and land-use as well as between transportation demand and supply; (5) interrelations between subsequent sub-models, particularly between car availability, trip generation and modal split; (6) developing models which are not only sensitive to transportation policies but also to other local policies (e.g. land use, city development, social, etc.).

Counting Urban Carbon: Effective Modeling of Resource-Efficient Urban Design Decisions Under Uncertain Conditions

“…Earlier transportation models treated the actions of individual vehicles as simple and predictable elements that seek only to continue on their current path at the maximum possible rate. The errors of these models, and the failures of the systems constructed in response -particular the failure to alleviate traffic congestion for any but a short period -are now well documented (Supernak, 1983). Of course, human beings are decision-making agents in their own right, and they are able to decide to take alternative routes based upon dynamic conditions -or not to travel at all.…”

SMART: simulation model for activities, resources and travel

“…Reviews and criticism Of the UTPS process are widely available, and generally make points not in substantial disagreement. For instance, in reviewing the system, Supernak (1983) notes that with the exception of disaggregate mode-choice models, very little progress has been made in travel-demand modeling in the past thirty years. Specifically, UTPS continues to exhibit a lack of behavioral content that prevents the analyst from evaluating alternative policies that are unrelated to investment proposals for major facilities.…”

“…His explanation is that the range of adaptations consumers have available is far greater than what can be modeled easily, and therefore the analyst errs by concluding that, if major transportation investments are not made, congestion will overwhelm the city. Supernak (1983) calls for the development of "... simple, yet not primitive/ easily applicable models..." for urbanized areas. Among the characteristics such models should contain are the following:…”

A comment on Supernak's critique of transport modelling

“…In a recent article in this journal Supernak (1983a) presented a forceful critique of contemporary urban transport modelling and prescribed a number of reforms which in his view, would improve the scientific status of the field and produce models of greater practical use to transport planners. In contrast to the many well known technical criticisms of transport models that have appeared over the years (e.g.…”