Mobility changes during the COVID-19 pandemic in Italy

Aggregated mobility metrics derived from de-identified location data are made available to assess the impact of mobility restrictions and social distancing in Italy during the lockdown.

Like Comment
Read more

On February 21, 2020, a cluster of SARS-CoV-2 cases was officially reported in the province of Lodi, marking the beginning of the COVID-19 pandemic in Italy. Immediately thereafter, national authorities started issuing a number of orders aimed at containing the spread of the virus through mobility restrictions and social distancing. On March 12, the Italian government imposed a national lockdown, becoming the first country in Europe to quarantine all its citizens in response to the pandemic.

At that time, the need for evaluating the effects of public health policies on the mobility of Italians became immediately clear. Cuebiq was one of the leading companies to recognize that big data can provide critical insights into the impact of the pandemic. 

As we were already collaborating through Cuebiq’s Data for Good Program, we rapidly embarked on setting a real-time analysis of anonymous location data to monitor the impact of public health interventions. Since early March, we have been providing periodic reports through a dedicated website, describing the evolution of mobility in Italy through the different phases of the emergency: before the outbreak, during the lockdown, and after restrictions were lifted.

Our Data Descriptor published in Scientific Data details the data processing and curation pipeline that we designed to generate three different metrics that are relevant for epidemiological analysis. First, an origin-destination matrix describing the daily flows of users traveling between the Italian provinces. Second, the weekly median radius of gyration of all users living in a given province. Third, the average degree of a proximity network built through the spatial overlaps of groups of users in public places.

Compared to other similar data sources that are publicly available, our dataset is one of the most granular in time and space, while preserving privacy and capturing different notions of mobility, from long-range travel to short-range movements. 

We hope that our dataset will be useful to researchers and policymakers who will seek to understand the health, economic, and social effects of non-pharmaceutical interventions adopted in response to the pandemic. 


Michele Tizzoni

Senior Research Scientist, ISI Foundation