We envision a routing system that leverages multiple data sources coming from the city, to provide personalized routing based both on a balance between personal preferences and common interest. People should be able to express their preferences about their desired destination, ideally through a mobile device, and be prompted with a routing solution:

  • > specifically tailored to their needs;
  • > that minimizes the negative effects for the community;
  • > that feeds the route taken by each user back to the system, so that other routes will be positively affected by the availability of this information.


What matters to each person
The city space is represented in grids. Each layer represent high or low levels of some particular aspect. These data can come from data sources of different nature (e.g. institutional, crowdsourced, ...) Each person can decide how much each level matters to him/her. A normalized weighted sum of all layers, tailored upon the weights decided by the users, gives a personalized layer that tells him which areas to prefer and which to avoid, solely based on his preferences and the city data sources. We call this the combined individual layer.

Visualization of the routing for simulated individuals moving in the city of Milan, where constraints due to crimes and air quality are considered.


Real-time Monitor of the "Pulse of the City"
Realtime monitoring of the permeability of the city, related to the average speed of individuals in the system. Larger the permeability, faster and easier the mobility through the city. The anomaly measures the number of standard deviations from the historical mean permeability and provides information about critical congested areas, triggering intervention.


Access to Big Data to Improve Mobility
Our simulations show that if only a small fraction of individuals in the city adopt the smart-routing system, the overall result is still far from optimality. Only when all individuals are part of the system the time required to reach the destination is reduced. Note that "Ideal" is the unrealistic situation where each individual travels in absence of any other individual and constraints of any type.


Synergy to Improve Mobility
Using approximations to model start and destination areas of individuals does not favor mobility improvement. Data-driven simulations show significant improvement in the time required to reach the destination. Here, we used the call exchanges between different areas as a proxy for correlated mobility patterns.

Antonio Lima

PhD student.
School of Computer Science,
University of Birmingham.

Website - E-mail

Manlio De Domenico

Postdoctoral Research Fellow.
Computer Science,
Universitat Rovira i Virgili.

Website - E-mail

Alex Arenas

Professor.
Computer Science,
Universitat Rovira i Virgili.

Website - E-mail