Using telecom data for development
Where to eradicate malaria first? How to increase access to banks and loans? How and where to improve air quality? How to design better public transport? These intractable public problems have been troubling policymakers and city leaders for decades, but a new ally is emerging: telecom data.
Every time you make a call, send an SMS or check your social media on the go, you generate data. This data can be anonymized, aggregated and used to reveal the movements of populations and economic and psychological profiles—all without putting individual privacy at risk. By merging telecom data with other relevant data—such as the number of new cases of a disease—and using visualization tools, governments and development organizations can make better-informed public-policy decisions. New projects, across Uganda, Zambia, Brazil, Haiti and many other countries, are using this method of data analysis.
Malaria kills around half a million people every year worldwide. We know that the disease itself can be transported over long distances within human bodies, before being transmitted to other human beings through mosquitoes. Eradicating malaria requires us to understand where people live, how and where they travel, who they meet, and where their acquaintances then go.
Mapping the movements of a population across an entire country, matched with malaria-infection data, can help zero in on problem areas. By following this process in Zambia, we identified one crossroad which was facilitating the spread of malaria across the country. With the location pinpointed, public authorities and medical staff were able to effectively prioritize interventions in this area. This approach is being replicated to fight diseases in other contexts as well, such as the Ebola virus epidemic in West Africa, Zika virus epidemic in Brazil and dengue in Latin America and Asia, and standardized approaches are being prepared.
We are all victims of traffic snarls, and it only seems to get worse with time. The time spent by a person in New Delhi on roads has doubled over the last six years—and recent studies show that driving 40km in peak hours can take almost 4 hours. In the US, average commuting time has increased by 20% since 1980. Local governments have been unable to solve this problem for many reasons. One important reason is that most municipalities do not have up-to-date granular data on the movements of populations on and off the roads.
Telecom data can be used to map and test the impact of changing or optimizing road infrastructure on actual traffic. For instance, the municipality of Kampala, Uganda, redesigned a key junction in the city. By using telecom data to compare the traffic situation before and after, it was clear that changing the junction had mixed effects on overall traffic. Time spent in traffic jams reduced by 19% for people driving from the Kibuli neighbourhood. However, people living in Nakasero spent 43% more time on the road. This type of insight has helped municipality staff to take informed decisions on adding traffic lights or changing road directions, and eventually change and optimize actions over time.
Using data, especially telecom data, for social impact is still a relatively new concept. Trust is the cornerstone of such an approach. It requires securing the trust of the two main contributors: consumers, by protecting their privacy, and telecom companies, by addressing their concerns that competitors may benefit from the strategic insights emanating from their data. This requires four levels of trust-building security measures. First, analysing the data within the databases of telecom operators, rather than extracting it so that it is kept secure within their own networks. Second, anonymizing the data to avoid any possibility of individual tracking. Third, aggregating the data across dimensions like geographies or communities, to prevent any possibility of identifying individual patterns. Finally, ensuring that telecom operators have final approval before any insights or visualizations are released in the public domain.
A billion mobile phone users in India and 427 million mobile subscribers across sub-Saharan Africa could shape the future of their countries without even looking up from their phones. Solutions could be tailored to the changing needs of communities through real-time data analysis in a way that has never been possible before. This is the next stage towards increasing automation and introduction of Artificial Intelligence based on digital data. It is now important for the public authorities to help the private sector to accelerate such developments.
India is ripe for innovation in using telecom data for social impact. With the world’s second largest telecom subscriber base, professionally run telecom operators with large customer databases, and several social problems waiting to be solved, telecom data in India could unleash a new transformation. Issues ranging from preventing the spread of dengue and chikungunya, to prioritizing the next batch of rural roads, to placing bank mitras to drive financial inclusion, are all waiting to be solved using the power of telecom data.
It is time for India to join the group of countries who are already harnessing the benefits of this approach and show strong leadership. This requires an effort that brings together telecom companies, researchers, regulators and other development actors to come together to begin exploring what is possible.
Frederic Pivetta and Varad Pande are with Dalberg, a strategic advisory firm.