#Berlin #Mobility
10/22
AI Analysis of Urban Mobility
Published in:
Transportation Research Part D:
Transport and Environment
(Volume 111, October 2022), Elsevier
Using Explainable Machine Learning to Understand How Urban Form Shapes Sustainable Mobility
Felix Wagner, Nikola Milojevic-Dupont, Aicha Zekar, Lukas Franken, Ben Thies, Nikolas Koch, Felix Creutzig
Technische Universität Berlin, Mercator Research Institute on Global Commons and Climate Change et. al.
Description
An international research team with scientists from TU Berlin and the MCC developed an AI-supported method to determine the influence of urban structures on motorized city traffic and used it to create the basis for climate-friendly urban planning. The authors of the study used high-resolution open-source data on urban development in Berlin for their investigations and analyzed a sample of 3.5 million car journeys over the course of a year using AI tools.
Results
The study shows that it is primarily the distance from someone’s home to the city center that influences carbon emissions from transportation. It also demonstrates that residents of districts with local amenities travel less by private car and thus mitigate carbon emissions. In addition, longer car journeys are made in districts with a low-income population structure because they are less well connected, and these districts tend to be located on the outskirts of the city, such as Marzahn-Hellersdorf.
Background
The study was funded by the Climate Change Center Berlin Brandenburg as part of the Center’s research interest in „Climate Change and Artificial Intelligence“.
Picture: Birgit Holthaus
Contact
Prof. Dr. Felix Creutzig
Mercator Research Institute on Global Commons and Climate Change
creutzig(at)mcc-berlin.net
