Govt may turn to supercomputing for better use of Aadhaar database

The article by Moulishree Srivastava was published in Livemint on February 10, 2015. Sunil Abraham is quoted.

When India launched a scheme in 2009 to give every one of its residents a unique identity number called Aadhaar, there were two main objectives: an efficient delivery of welfare services, and a tool for monitoring government schemes.

Six years on, as the government moves a step closer to covering all 1.25 billion citizens with Aadhaar, technology experts working with the project have spotted other potential uses, all to be powered by a string of supercomputers. With Aadhaar having crossed the 700 million milestone in 2014, the government is preparing to launch supercomputing applications to make more sense of the scheme’s massive database. These will help link the database to public services by running data analytics, which in turn will throw up trends that can help promote better-informed policy-making.

“We are talking about the Aadhaar database and the huge population data associated with it. Right now it is only a collection of data, but once we start rolling out applications, the amount of information that will be generated and the amount of usage that will happen will be enormous,” said Rajat Moona, director general of the Pune-based Centre for Development of Advanced Computing (CDAC), the research and development arm of the department of electronics and information technology responsible for the National Supercomputing Mission.

“Big data analysis and big data visualization, these are supercomputing problems. It (handling Aadhaar database) is actually visualization of huge data that can help in the planning,” he added. “When you have that much amount of information, you can do a lot of data analysis and figure out trends. We can thus see what are the policies needed to be defined. Therefore proactive policymaking based on previous trends is possible provided we have data,” he said.

Aadhaar is being used by the government for transferring cash subsidies directly into the bank accounts of beneficiaries in order to plug leakages. It is used to identify beneficiaries for transferring funds under scholarship schemes, pension money, cooking gas subsidy as well as seeding of bank accounts to weed out multiple beneficiaries under the government’s financial inclusion programme.

There are also plans to use Aadhaar to curb black money in real estate transactions. Developing these applications further, says Moona, needs supercomputers, which will give far better results. “In order to develop these applications we need to understand supercomputing and use of supercomputing,” he said. “We don’t have such applications developed…we usually take larger applications developed elsewhere and customize it according to our needs. So applications that are developed for solving India-centric problems will actually give a better output,” he said.

“We are talking about creating related applications that can run on supercomputing cloud of million cores (cloud supported by million-core supercomputers),” he added. “It will be a platform, where we can solve a large number of problems using shared model.” “It (supercomputing cloud) will not be a public cloud. There will be research institutes, research organizations, security agencies, government departments and each of them will have its own data and access pattern.

This is going to be a part of our application-centric usage of (the proposed) supercomputing (grid).” The million-core supercomputing cloud is a part of the government’s Rs.4,500 crore project aimed at setting up a grid of more than 70 supercomputers to be hooked up across these institutions and organizations over the next five to seven years. This will enable Indian researchers to build applications on the network, as well as to harness the power of supercomputers for research and development. India has 12 of the world’s 500 most powerful supercomputers, the largest currently deployed being a 500 teraflops (equivalent to half a petaflop) computer, administered by the CDAC. To work on supercomputing applications in areas such as the Aadhaar database, terrorism, advanced weather monitoring— including cyclone and flood prediction—and geo exploration, the government plans to train 22,000 high-quality professionals over 5-7 years.

Some experts point out that a centralized Aadhaar database faces cybersecurity risks that can threaten people’s privacy. Centralization is a terrible idea from the perspective of cybersecurity, said Sunil Abraham, executive director of the Bengaluru-based research organization Centre for Internet and Society. “This is a honeypot (a computer system that is set up to attract and trap people who attempt to penetrate other people’s computer systems) and variety of bad actors, i.e. criminals, terrorists, as well as states and corporations, will target this database hoping to compromise it.”

A petition filed by former Justice K.S. Puttaswamy in 2012 said that Aadhaar scheme infringes citizens’ privacy as applicants need to provide personal information on biometrics, iris and fingerprints, which infringes their right to privacy that is a part of the fundamental right to life under Article 21 of the Constitution. Security experts have been raising concerns about lack of secure and foolproof system to ensure that all the data will be safe and will not be misused. “How do we address the security and innovation imperative without compromising on the right to privacy? We must build individual transaction databases for each government department/ministry and (have) decentralised authentication. Only anonymized dataset should be made available to the supercomputing applications that are being built,” Abraham said. “For innovation and improved e-governance, anonymized data is sufficient.”