At the beginning of the 21st century, a growing number of agencies across the U.S. are abandoning established traditional modeling techniques to explore advanced practices in travel forecasting. This paper summarizes (1) The benefits reported by agencies implementing advanced models, (2) Implementation and institutional issues that may become a barrier to change, and (3) Lessons learned from those agencies that did explore advanced modeling practices. The work builds on research conducted for a National Cooperative Highway Research Program (NCHRP) Synthesis of Practice.

Findings are mainly based on interviews with agencies that work with advanced travel models. A total of 23 agencies were interviewed for this study (Figure 1). Though a questionnaire was provided, the interviews were narrative in format to account for the various stages from early consideration to full implementation of advanced models in different agencies. Literature reviews and practical experience of one decade of advanced travel forecasting completed the research.

This paper defines advanced transportation modeling as practices that go beyond the traditional four-step travel demand modeling approach. Specifically, this includes five areas of modeling: tour- and activity-based models, land use models, freight and commercial movement models, statewide models, and dynamic network models. Today, with the possible exception of dynamic network models, all types of advanced modeling have been deployed successfully to analyze policy questions that could not have been properly analyzed with traditional four-step models.