How did lockdowns affect urban mobility? Research tracks Rio de Janeiro cell data

Insights help health officials make more informed decisions on future pandemic response

Researchers at Binghamton University and Brazil’s Fluminense Federal University collaborated on research tracking cellular data in Rio de Janeiro during the COVID-19 there last year.
Researchers at Binghamton University and Brazil’s Fluminense Federal University collaborated on research tracking cellular data in Rio de Janeiro during the COVID-19 there last year.
Researchers at Binghamton University and Brazil’s Fluminense Federal University collaborated on research tracking cellular data in Rio de Janeiro during the COVID-19 there last year.

Lockdowns due to the COVID-19 pandemic were a fact of life in 2020, especially during the first few months as the virus spread from China to the rest of the world.

But how did they affect people’s everyday movements, especially in urban areas? A new study by researchers from Binghamton University and Brazil’s Fluminense Federal University analyzed cellular data to see how the coronavirus threat affected travel in Rio de Janeiro.

The anonymized data from Brazil’s second-largest urban area, collected between March and July, included 2 million users per day and 120 million connections to antennas around the Rio metro area. By pinpointing the locations of those connections, researchers could create a day-by-day look at how many still trekked around the city and where they went, as well as how many residents stayed home as advised.

Binghamton Assistant Professors Arti Ramesh and Anand Seetharam, PhD student Necati Ayan and master’s student Sushil Chaskar.— all in the Thomas J. Watson Collegel of Engineering and Applied Science’s Department of Computer Science — collaborated on the research with Professor Antonio Rocha from Fluminense Federal University and his students.

Previously, Ramesh and Seetharam have studied the early U.S. response to COVID-19 on Twitter and utilized data collected from around the world by Johns Hopkins University to build coronavirus prediction models using artificial intelligence.

The Rio de Janeiro research — which included data from before, during and after the city’s strictest lockdown protocols — showed that while around 85% of residents showed low mobility, 15% still ventured significantly outside of their neighborhoods.

“Despite the lockdown period, some people were venturing out,” Seetharam said. “They could be essential workers or Uber drivers — people like that. They could be workers who don’t have the luxury of working remotely. There are also people who just do not follow the rules.

“There was a significant number of cell-phone users — about 4% — who were traveling enough to be tracked by more than 10 antennas. Each antenna can be 2 kilometers in range, so that covers a large amount of the region.”

As part of the study, researchers designed an interactive tool that shows mobility patterns, which can potentially help health officials understand the mobility of individuals and the number of COVID cases in particular neighborhoods.

“A couple of things we’ve seen will be beneficial to study further,” Ramesh said. “We have a socioeconomic index for different sections of Rio, and some of the data suggests that those from a lower socioeconomic index tend to travel more and have more mobility. In other regions that have a better socioeconomic index, the mobility is not as pronounced.”

After the lockdown eased, residents’ mobility did not return to pre-COVID levels — and because the cellular data collection in Rio is ongoing, longer-term trends could be assessed.

“If the numbers do bounce back, maybe they are not going to the same areas or maybe they are not going to the beach,” Ramesh said. “What other things could have changed? We need more data to come in so we can start studying it.”

And even though this research is specific to Rio de Janeiro, Seetharam believes that it could guide public-health strategies during COVID-19 and any future pandemics.

“We could take the broad conclusions and apply them to other cities,” he said. “We can assume that people’s behavior is generally similar and does not change that much from country to country and city to city. The long-term planning strategies could be applied to other countries.”

One way it might be used is to plan better ways to keep residents safe without closing everything down — perhaps by reducing the stress on essential services such as groceries.

“Blanket lockdowns have not been as welcome across many, many countries, and going forward that would not be a good approach to take if you want people to comply with the rules and support them,” Ramesh said. “How do we work out plans that get rid of overcrowding but also make people more compliant to the rules?”