Project: City Rhythm


City Rhythm a longterm research programs carried out by Delft University, AMS Institute and 6 cities in the Netherlands: Amsterdam, Rotterdam, Zaanstad, Den Haag, Helmond en Zoetermeer. The research is support by the "Digitale Steden Agenda", a network organisation around smart cities.

This research aims to enhance the sense of security in a specific neighbourhood by stimulating social cohesion through the creation of shared rhythms. Rhythms can be stimulating or rhythms can be detrimental. Rhythms can discipline people and rhythms can facilitate self-organisation. Rhythms can nurture well being and rhythms can indicate when something is wrong.

Delph is a stakeholder of this project and is responsible for developing  a Machine Learning model that detects rhythm structures in multidimensional longitudinal data.

The challenge

By looking at the development over time of city blocks as organic rhythms will allow policy-making to go beyond the static approaches that have driven most of its discussion. The City Rhythm research aims to identify methods of rhythm analysis, intervention in rhythms and it wants to formulate how rhythm analysis can be a new resource for urban policymaking. 

Important is that the model distills rhythm in a purely data-driven approach. This enables policy-makers to evaluate strictly objective indicators of what is happening within a block, as opposed to obscure indicators such as 'de veiligheidsmonitor' that are implicitly driven by assumptions driving the weighting function underneath them.

Project Objective:

Develop a data-driven model that identifies rhythms and interventions in rhythms in order to be used as a new resource for urban policymaking. 


Delph applied a Machine Learning model to a dataset containing extensive demographic information on a total of seven cities in the Netherlands. This model identifies rhythms in neighbourhoods and the development of these rhythms over time. The model leads to a characterization of the various areas as breathing organisms through space and time and will allow neighbourhoods to be compared not only on the presence but also based on what has been and what's next.