Advancing the Science of Travel Demand Forecasting

The travel demand modeling field is ripe for re-invention as we are on the cusp of the next generation of models. We find ourselves in arguably the most dynamic transport environment in the history of the field. Massive data are becoming available and new developments in data analysis methods are being developed. Researchers from disciplines such as computer science and physics are entering the domain, along with high-tech, entrepreneurial firms. There is also a change in the nature of data. Whereas historically the most important data sets in transportation—Census data, household travel surveys, origin-destination surveys—were collected by public agencies, the new generation of “big data” is most often held by private entities (Shuldiner and Shuldiner 2013). There are important questions around the continued availability of such data to modelers, as well as what model developments need to take place to fully exploit the potential of such data. Finally, not only are new data and methods changing the modeling, but transport technology and patterns are themselves rapidly changing as new modes such as app-based ride sharing of cars, bikes, and scooters are making their presence felt in large US cities. Looking to the future, partly or fully autonomous passenger, freight, and aerial vehicles are rapidly advancing.