Understanding how people use the spaces they inhabit—where they live, work, and gather—is key to effective urban planning that meets their needs . For example, knowing which routes are most commonly used to travel from residential neighborhoods to workplaces, and during which time periods, makes it possible to adjust bus routes and frequencies, and to optimize traffic control resources. Identifying meeting, leisure, and event spaces, as well as the time periods when crowds are largest, is useful not only for keeping the public informed, but also for strengthening cleaning and transportation services and developing promotional strategies. The more accurate this information is, the more it will help public administrations to plan urban space and services in effective and realistic ways.
With this goal in mind, a team from the University of Cordoba has developed a tool that makes it possible to map the structure and pace of cities based on how their residents use their cell phones in different locations. This software tool, called MAPLID (Multi-label Approach for Place Identification), is unique in that it provides a multi-label classification, which means that a space can be classified into more than one category : for example, a university campus can be a workplace, but also a residence, and even a leisure space, at specific times.
By recording the calls made on a given day and tracking their recurrence throughout the week, the tool can identify which places are visited and how often, thereby tracing the population's routines and deciphering the functionality of different areas of the city. It provides information on when population density changes in a specific area; for example, on a downtown street that transitions from being exclusively a residential area to also becoming a work area as businesses open. The tool also reveals at what times the highest volume of road traffic occurs en route to industrial parks, and how the holding of a one-off, large-scale event alters the local population's pace and habits.
A pilot test in two Italian cities
The development of the algorithm on which MAPLID is based is part of the doctoral thesis by researcher Manuel Mendoza Hurtado, who, along with his colleagues at the Department of Computer Science and Artificial Intelligence, Juan A. Romero del Castillo and Domingo Ortiz Boyer, has published the first results of this tool in the International Journal of Geographical Information Science . The model has not yet been tested using data from Spain. As the researchers explain, " access to large-scale mobile phone data is severely limited , due to privacy concerns." Thus, for the initial project they used a database of phone calls that Telecom Italia released for scientific use.
This dataset was used to map and analyze the population's activity, first in Milan and then in Trento. As these are two Italian cities with different layouts and characteristics, it was possible to validate the tool's reliability in different settings. To do this, Manuel Mendoza explained, they used geolocated data that was generated each time a mobile device connected to one of the network's antennas to make a call or send a message. Every day, millions of data points were generated, which, once overlaid on the city map, were supplemented with contextual information taken from another open-access tool: OpenStreetMap. The results of this work showed that a multi-label tool like MAPLID is useful for understanding how the inhabitants of a place move , yielding a map that can become richer the more complementary information is added to it.
The researchers unanimously highlighted the potential of this tool for urban planning, as it offers "the ability to capture multiple urban functionalities and uses, unlike other models that tend to lose information upon oversimplifying it." The next step is to contact local governments to make the tool available to them . They point out that the tool's most important application would be in traffic management, as the data "would make it possible to adjust public transportation schedules and frequencies, plan routes, and even optimize fuel costs for private vehicles."
International Journal of Geographical Information Science
Computational simulation/modeling
People
MAPLID: a new multi-label approach for place identification using data supplied by mobile network operators
29-Jan-2026