Project
Research teams led by Prof. Waller have developed novel methodologies that utilize machine learning (e.g., 20+ years of research on applied evolutionary algorithms) and data science to increasingly automated the process of building and calibrating models suitable for a range of transport planning questions. The core techniques leverage established principles of transport network supply and travel demand equilibration while relying on broadly available, pervasive, transport data. Early applications include:
Analysis of traffic behavior during the Ukrainian conflict (Waller et al., 2023) has been made possible via the automated approach since models can be established automatically within hours rather than months. Specific research applications include the rapid assessment of network loss to assist reconstruction planning as well as the development of scenarios to assist in the design of cities more resilient to natural disaster and human conflict.
NOTE: A more complete set of event data from the Ukrainian analysis can be downloaded from here.
