Speaker(s)

Dr. David Boyce

Abstract

Travel forecasting in general, and the traffic assignment problem in particular, have continuously posed computing challenges. In the 1950s, for example, travel forecasting was one of the first civilian applications of the emerging technology of mainframe computers. As computers evolved, larger networks and more complex models were developed, fully utilizing available computing capacities up to the present day.

In recent years, the need to identify the impact of modest local changes on the performance of large scale networks led to an increasing need for higher convergence levels. This need has been addressed by algorithmic changes now being implemented by leading software vendors. The future challenge in static traffic assignment appears to relate to the identification of routes and plausible route flows.

Recent findings from practical solutions for this emerging need for route flow forecasts will conclude the seminar.