Quantifying traffic dynamics to better estimate and reduce air pollution exposure in London
Fellows: Sam Blakeman, Jack Hensley, Oluwafunmilola Kesa, Caroline Wang
Technical Mentor: Maren Eckhoff
Project Manager: Sam Short
Project Partners: Imperial College London, Transport for London, City of London, King’s College London
Although London has made significant efforts towards tackling pollution, there is an acute need to improve the capability to understand how traffic disruptions and policies affect traffic congestion and, in turn, how this affects vehicle emissions and air quality.
Currently, traffic statistics are obtained through high-cost manual labor (i.e. individuals counting vehicles by the road) and extrapolated to annual averages; they are not detailed enough to evaluate traffic or air pollution initiatives, and routinely underestimate emissions from vehicles.
The purpose of the project is to create an open-source library that processes junction-level traffic video data and extracts descriptive statistics, such as the number of counts and type of vehicle, and the number of times each vehicle stops and starts.
The library quantifying traffic dynamics would allow researchers to improve emissions model predictions, and therefore enable more accurate evaluation of the impact of future transport initiatives. Such a library would also allow policymakers to better understand the impact of road works and road closures, as well as unlock the possibility of optimizing traffic flow to reduce congestion.