How information gleaned from mobile networks could help developing countries meet sustainability goals

CAN THE SAME TECHNOLOGY that’s used to pinpoint mobile phone users’ locations to target them with localized advertising also help alleviate poverty, stifle disease and create better, more equitable infrastructure in the developing world?

It’s an intriguing, if unusual, question. It’s also one that’s top of mind these days for Sriganesh Lokanathan. He’s the leader of the South Asian component of a two-year IDRC project to help countries in South Asia, Africa and Latin America use “big data” — the mountains of digital information collected in administrative records, commercial transactions, social media and through sensors and tracking devices, including mobile phone networks — to meet the UN Sustainable Development Goals, or SDGs.

Above image: Sriganesh Lokanathan (right) and Yashothata Shanmugarajah discuss insights on economic activity as inferred from the analyses of big data from mobile networks in Colombo, Sri Lanka (Photo: LIRNEasia)

The project’s inspiration was the United Nations’ call for a “data revolution for sustainable development” in 2017, following the body’s adoption of the 17 SDGs in 2015. But the subject has been a focus for Lokanathan, head of big data research at LIRNEasia, an information and communication technology policy think-tank based in Colombo, Sri Lanka’s capital, since 2012.

At first, Lokanathan, a veteran data scientist and policy expert who has studied in the United States, Singapore and South Africa, was a little ahead of the curve. “It’s like I was going door to door saying, ‘Have you heard the good word of big data?” he says.

But what he saw then — and what the UN and many more policy makers and government officials have now formally recognized — is that big data offers a trove of timely demographic and socioeconomic information to help data-starved countries enact policies to meet the SDGs and measure progress toward them.

“A census is only done every 10 years. And surveys are expensive — you cannot do them frequently,” Lokanathan explains. “So how do you get insights otherwise?”

Most of Lokanathan’s work in the IDRC project’s first year has focused on the policy front, building coalitions with the private sector and government in Sri Lanka and across the region. But his earlier research with mobile phone usage data highlights its potential.

“Every time you make or receive a call or send an SMS, the system records the phone number calling, who it is calling and the base station they connect through,” he explains. His team worked with several mobile operators to obtain the call data records — with all identifying data stripped out to ensure anonymity — from two provinces in Sri Lanka, including the area in and around Colombo.

He and his team ran their analyses — massive number-crunching efforts done on their own custom-built server farm — to extract observations based on the call data itself or in conjunction with other data from traffic cameras, satellites and social media. The results shed light on mobility patterns, population density (hourly, daily, weekly), traffic flows and transportation systems use, land-use patterns, informal economic activity and more. They were also able to show mobile data’s potential in forecasting potential disease vectors.

“One of the things we wanted to understand, for example, was how people come in and go out of cities, where they come from and where they go,” says Lokanathan. “We can also start to understand a little bit about who’s rich and who’s poor based on the areas where they’re coming from.”

The latter, he says, helps address questions like: Do poor people have to travel more for work? “If you can understand this better, then it can translate into transportation policy decisions.”

A key advantage with mobile big data is that it provides widespread coverage of all segments of the population. In Sri Lanka in 2016, there were 118.5 mobile subscriptions for every 100 people compared to just 32 internet users per 100 — with comparable ratios found across much of the developing world.

But Lokanathan stresses that the idea is not to draw on one data source in isolation for decision-making, but to combine it with other information. “You need to take that big data and throw in some official statistics, figure out ways of putting them together. That’s what really gives you the insights.”

In a similar vein, a key goal of Lokanathan’s work within the IDRC project is to help create networks of expertise in the use of big data across the developing world. “My ideal scenario,” he says, “is that in two years I won’t being doing half the things I’m doing today in other countries because someone else will be doing them.”

This article was originally published in January 2019 on the Charting Change website, published jointly by Canadian Geographic and Canada’s International Development Research Centre. Image courtesy of LIRNEasia, IDRC and Canadian Geographic.