Press release | 26 September 2022 | pp

Using Artificial Intelligence to Plan Climate-Friendly Cities

Study conducted by the Climate Change Center Berlin Brandenburg and TU Berlin shows that Berlin’s neighborhood structure is particularly climate-friendly

The CO2 emissions of motor vehicles must be greatly reduced if Berlin’s climate goals are to be met. Urban form interventions have major potential for reducing emissions in the transport sector. This is one of the findings of the latest Assessment Report of the Intergovernmental Panel on Climate Change. However, it is not yet clear exactly how the infrastructure of the capital influences the driving behavior of the city’s inhabitants and the associated CO2 emissions.

Researchers at TU Berlin are developing an AI-supported method to assess the influence of the built environment on motorized urban traffic and thus create a basis for climate-friendly urban planning. A study of their first findings has now been published in the leading academic journal “Transportation Research.”

The study conducted by the newly founded Climate Change Center Berlin Brandenburg (CCC) and supported by the Mercator Research Institute on Global Commons and Climate Change (MCC Berlin) shows that the size of a city in particular but also the distances to local neighborhood centers can greatly influence the distances covered by cars and by extension CO2 emissions. The report makes clear that shorter distances to neighborhood centers are associated with shorter car trips and any increase in the distance from the city center results in exponential increases in commutes as well as CO2emissions. This is particularly true of the boroughs in the southeast (Marzahn-Hellersdorf, Treptow-Köpenick) and southwest (Steglitz-Zehlendorf and Spandau) of Berlin. The report also shows that mainly cars from areas associated with lower incomes have to make longer journeys. According to the researchers, this makes a strong case for urban planning strategies aimed at creating a denser inner city while also seeking to liberate suburbs from their dependence on cars.

3.5 million car commutes and high-resolution urban form data were evaluated

The authors of the study used an AI method recently developed for urban planning for their research together with a sample of 3.5 million car trips over the course of a year and high-resolution urban form data. The results demonstrate how AI can be used for applications relating to climate protection. The lead author of the study is Felix Wagner, doctoral researcher at the MCC, an affiliated institute of TU Berlin: “The potential of AI in the area of sustainable urban planning lies in the ability to incorporate city-wide dynamics as well as local details in one single model.”

Local subcenters are important for sustainable mobility in Berlin

Professor Dr. Felix Creutzig, scientific coordinator at the Climate Change Center Berlin Brandenburg (CCC) and co-author of the study stresses the significance of the new method for climate policy: “Urban planners can use the analysis and application of urban big data components to quickly and flexibly assess desirable goals such as climate-friendliness. This approach can prove particularly effective in achieving the city’s climate goals by 2045 during times where too few qualified staff are available.” Accordingly, the CCC, the Berlin-Brandenburg climate alliance led by TU Berlin, is continuing to drive forward the development of AI applications relevant to climate protection with the support of the MCC Berlin.

The results also show how important local subcenters are for sustainable mobility in Berlin. While Felix Wagner’s study represents a first approach, future research projects will be more focused on predicting the influence of future planning strategies on sustainability. According to Felix Creutzer, publicly accessible data on portals such as daten.berlin.de can help drive climate-relevant urban planning strategies. The various departments of the Berlin Senate and the city’s urban planners can use this data to make decisions taking greater account of climate protection.

Publication in “Transcription Research”

Wagner, F., Milojevic-Dupont, N., Franken, L., Zekar, A., Thies, B., Koch, N., & Creutzig, F. (2022). Using explainable machine learning to understand how urban form shapes sustainable mobility. In: Transportation Research Part D: Transport and Environment, 111, 103442: https://doi.org/10.1016/j.trd.2022.103442

Contact

Prof. Dr.

Felix Creutzig

creutzig@tu-berlin.de