Meet architects of Aarogya Setu, the app meant to stem Covid-19 spread

Aarogya Setu, the contact tracing app which this high-powered team went on to build and is continuing to refine, is aimed at mapping the contacts of users by analysing GPS data
Sometime towards the third week of March, Lalitesh Katragadda, a former head of Google India engineering centre and the man behind Google Maps, received a call from Ajay Sawhney, secretary in the ministry of electronics and IT (meitY). Sawhney wanted Katragadda to lend his expertise to the government’s efforts to come up with a technology-led intervention to battle the coronavirus pandemic.

“I requested Lalitesh to come on board because, though we already had a team in place, we wanted someone very senior and respected to be at the helm of managing the product they were building,” says Sawhney.

 
An IIT-Bombay alumnus and a PhD in robotics from Carnegie Mellon University, Katragadda is widely credited with developing Google Mapmaker, a tool which eventually led to the creation of Google Maps. In fact, he joined Google in 2002 when the tech giant acquired a robotics startup he had founded in San Francisco. Subsequently, Katragadda relocated to India to set up Google’s India operations.
Given the magnitude of the problem and the enormity of the task at hand, Katragadda found Sawhney’s invitation too exciting to turn down. He decided to become a part of the Aarogya Setu project, which has turned into a one of a kind collaboration between government agencies, the academia, researchers and the private sector. 

Aarogya Setu, the contact tracing app which this high-powered team went on to build and is continuing to refine, is aimed at mapping the contacts of users by analysing GPS data, so contact tracing becomes easy if a user gets infected. Though some have raised concerns over privacy and the security of user data, the government claims that the app has already proved its worth. 

 

 
According to the government, Aarogya Setu has predicted close to 1,000 hotspots and recommended testing for 8,500 people based on their risk profile. Of them, 23 per cent have tested positive for Covid-19. The app  recommends testing users based on their own assessment of their health conditions.

“From the technology point of view, this is India’s finest ever,” says Katragadda, who is currently building the technology backbone of Avanti Finance, a financial inclusion platform built on an open-access digital model and promoted by Ratan Tata and Nandan Nilekani. 
Like Aadhaar, Aarogya Setu, too, has attracted significant participation from the private sector. MeitY secretary Sawhney says it is like a “Team India” effort with the contribution of around 80 of the best minds, and an almost equal participation by government agencies such as the National Informatics Centre (NIC) and NITI Aayog, and the private sector. 

The team has been working tirelessly, often putting in 20 hours work daily. Many of its members are based in different cities. “I have not even seen the faces of many of the people who are working on this project though I talk to them almost every day. I don’t even know which cities they are in,” says Sawhney. 

The main technology architecture and software (platform) development work of Aarogya Setu has been done by a team of around 15 top engineers from travel portal MakeMyTrip (MMT). They include Vikalp Sahni, co-founder and chief technology officer at Goibibo, which is now a part of MMT, and Rahul Goyal, a senior vice-president, engineering, at Goibibo and MMT. In fact, MakeMyTrip co-founder Deep Kalra was one of the first from the private sector to offer his services to this effort. 
As the lead volunteer, Katragadda is building the technology architecture and product-managing the solution. A team from online pharma company 1mg is also involved in the project. And everyone is working in close collaboration with the team at NIC headed by Neeta Verma (mission head) and NIC deputy director general R S Mani (project head). 

In addition, there is a team of around eight computer scientists and researchers led by V Kamakoti, professor of computer science at IIT-Madras. At least five among them have 25 years’ experience in the industry. Another team from the Indian Institute of Science (IISc), Bengaluru, led by Amrutur Bharadwaj of the department of electrical communication engineering, is helping with various aspects of the technology, including security testing  and big data and analytics to give actionable insights to the health ministry and the team on the ground. 

That’s not all. There is also a team of senior scientists from the Defence Research and Development Organisation (DRDO), which is helping with the advanced security analysis, and a team from Dailyhunt, an Indian news mobile app, is helping to internationalise the effort. Among the large companies, Tech Mahindra and Tata Sons are helping with testing and analytics, among others.  
“Though we have successfully worked on many research projects at IIT-Madras, this is more than research for us. There is much more accountability since you are dealing with the lives of millions of people,” says Kamakoti. The app, which has been downloaded by 98 million users, is expected to be used by150-200 million people as lockdown restrictions are eased. The Aarogya Setu mobile app works as an interface with users. 

What makes it so effective is the way it has been powered by technologies such as Bluetooth and GPS mapping, as also its robust analytics engines.

As far as the safety of user data is concerned, the architects of the app say that the only data stored in the server and shared with the health ministry are the name, age, gender, travel history and so on, that the users put in while registering themselves. The phone’s Bluetooth, which is always active once you install the app, records those whom the user comes in contact with, provided they, too, have the app on their phones. Hence, if anyone turns Covid positive later, it becomes easy to trace those he or she has come in contact with through syndromic mapping. 
“Syndromic mapping simply means syndromes or the symptoms that people have, and the places where they are located. This data, if taken in silos as an individual data point, does not mean anything. But once you anonymise it and create a map out of it,” says Katragadda. 

The team led by Kamakoti, on the other hand, is running an algorithm which is fusing data from the syndromic map with that from the trace history of all the people who have been detected Covid-positive. If during the analysis, several infected persons are found to have spent time in a particular area or pin code, that could mean that it is a Covid-19 hotspot.

“Aarogya Setu is not a copy of innovations done elsewhere. It was produced by some of the best minds of India who came together to figure out what we could do using computing and smartphones so that we could achieve, with limited  resources, the kind of containment that would require far more resources,” says Katragadda. 


Business Standard is now on Telegram.
For insightful reports and views on business, markets, politics and other issues, subscribe to our official Telegram channel