Sun 20 April, 2014  Drop a Visiting Card
"We'll use best biometric, storage & search solns"

Unique Identification Authority of India Chairman Nandan Nilekani admits that technology is a key challenge in rolling out the world's biggest UID project

Infosys' success story has become a legend in India's corporate history. And the key writer of the story Nandan Nilekani is now busy writing another story—of India's Unique ID project. Chosen by the Government of India to lead the project, Nandan has chalked out a clear roadmap to reach the goal. In an interview with Pravin Prashant and Shubhendu Parth he shares his priorities and plans of the project.

You have moved from writing about India's urgent need for an ICT enabled national ID card in your book Imagining India to now being at the helm of affairs driving the project. How do you plan to translate your imagination into a reality?
I have planned a very systematic hassle-free approach to complete the marathon task. Next 12-18 months are crucial for us to make administrative structure and central IT infrastructure working to its fullest capacity. During this period we will partner with a few registrars across the country to begin initial enrolment—for which standard guidelines for the technologies and infrastructure to be adopted by them to ensure error –free enrolment avoiding duplication and verification of residents, have been spelt out.
We will expand our partnerships with registrars as we build scale. We plan to issue the first UID number in next 12-18 months and cover 600 million in the next five and half years.

The unique ID project has been there for long, but failed to take off due to some reason or the other. What are the top three challenges that you see in rolling out the mega project?
The key challenge we face in implementing the project is not political or social, but technological. The largest biometric project prior to the UID initiative built a database of 100 million people. Our scale will be ten times that, and this creates new challenges in speed, de-duplication and information recovery.
We will also face some challenges in reaching the UID number to every resident across the country. We are shaping our approach so that we can bring the number even to people in marginalised communities. Special focus will be given to reach them.

Unique Identity Number (UIN) will provide significant savings to national exchequer. What’s the total saving India can make once UIN is allocated to 600 million people?
Though no exact estimation of the savings can be made at this juncture, I believe that the money saved by using UID would be substantial to plug the leakages in the system. It would lead to better targeting of the individual benefits and will ensure that those entitled to get the benefit actually get it.
Talking about the technology intervention in the government sector, you had said in your book that the new systems cannot be built over a 'creaky base'. How do you define this 'creaky base' and how bad do you think is the situation from the end result or good governance perspective? What do you think needs to be done to streamline the processes?
I think that technology should indeed be implemented with caution, especially when we are building it over existing systems. The challenge for country's governance system has been that we have long lacked the tools to target public funds more effectively, and prevent their leakage.

For example, when it comes to welfare, direct benefits—where subsidy benefits are given directly to an individual's bank account—are far less prone to leakage compared to indirect subsidies. But India lacks the infrastructure in financial access and services to implement such a direct benefit approach. This has begun to change only recently, with mobile and online banking, and with the growth of microfinance institutions.

Now that we have the technology infrastructure available to implement better and more direct welfare programs, we can't build such technology over the existing models—the 'creaky base' that we had to contend with when we lacked better tools.

 For example, with the UID number, we decided that we cannot use the existing government databases for collecting resident information, because they were collected without the de-duplication tools and biometric technology we now have.

What will be the administrative structure of the Unique Identification Authority of India (UIDAI)?
The UIDAI would have its headquarters in Delhi, Technology Centre in Bangalore and eight regional offices in the country. Regional offices at Delhi, Chandigarh, Lucknow, Mumbai, Ranchi, Guwahati, Bangalore and Hyderabad would have Assistant Director Generals. Besides, Deputy Director Generals would be posted at Lucknow, Mumbai, Ranchi, Guwahati and Bangalore. The UIDAI would recruit  people both from the public service as well as the private sector. The teams would be in place in the next six months.

To manage Central ID Data Repository (CIDR)—which is one of the complex tasks performed by any computing system on the earth—is the most challenging task for the UIDAI. CIDR will also perform online authentication of identity on key demographic fields and biometrics for each new enrollment, to ensure that no duplicates exist. Any thoughts on how you plan to move on the CIDR implementation and online authentication and the technology best suited for this purpose?
The technology design is at a very early stage. We plan to use the best biometrics, storage and de-duplication technologies available, but it is too early to specify those.

Have you finalised on the type of biometric information you plan to collect from individuals? What would be modus operandi for feeding them into the CIDR systems? What are you planning to do in cases where a person does not possess such biometric features?
We are considering a range of biometric options at this point, including fingerprint and iris scan. We will also build enough redundancy into the system in case biometric features such as fingerprints are completely unavailable. The UIDAI has set up a Biometric Committee to recommend the standards for the use of biometrics. Exceptional cases will require special handling and care will be taken to handle such cases.

What is the age group applicability for UIN? Have you finalised on number of digits for UIN and what does each digit or group of digit signify?
The UID number will be accessible to anyone who is a resident of India, including infants. In the case of infants and very small children whose biometrics—fingerprints —are not fully formed, we would take their parent's fingerprint and UID number as part of the record.

The UID will most likely be a 16 digit number (12 + 4), with check code digits at the end. The number will be a random number, so the digits will not signify anything. This is to ensure that the number cannot be deduced from the individual’s personal information. This will be finalised soon.

Do you foresee entrepreneurs designing specific applications using UIN for different departments and also for private companies on a transaction based model?
Yes, we will not make the information on the database publicly available for security and privacy reasons, but entrepreneurs will be free to build third-party applications which use the individual's UID number with his/her consent.

India will be the first country to implement biometric based UIN for its residents on such a large scale. How is UIDAI planning to move with respect to project implementation?    
The scale of the project is indeed unprecedented. To implement it, we'll have access to substantial computing resources across our data centers. We will employ distributed computing and search and indexing mechanisms to retrieve data quickly and respond to identity confirmation and authentication requests. We intend to use the best-in-class technologies when it comes to biometric, storage and search solutions for this project.

UIN will be a massive exercise for the UIDAI. What are the plans on the security aspect to make the system secure proof?  
We intend to take specific measures to ensure that the entire UID system is secure. We will have very clear registrar agreements in place on sharing and storage of information, and any agency that fails to follow our security standards while issuing the UID will no longer be a part of the system. All data transferred will be over an encrypted network. Access to the database will be highly restricted, with clear audit trails on information records accessed. We will also have automatic alerts in case of patterns of suspicious access to the database.

Additionally, we will not share or distribute any information we collect on individual residents. If you wish to authenticate someone, and send the Authority the person's name and UID number, our system will only respond with a 'Yes' or a 'No'.  

India has a huge number of e-Governance projects to boast of in terms of ICT interventions in the government sector. Do you think these projects are headed in the right direction or are there things that we have missed out in our zeal to look at 'e' as the panacea of all ills?
There have been many ICT interventions that the government has experimented with—some failures, and some admirable successes. A few of these projects did give us a sense of what is possible when it comes to reforming government institutions: the digitisation of the Railway ticketing reservation system is a very good example of how ICT can make a change in governance.
I think the problem the government faced here with the e-Governance projects that didn’t work is what I mentioned earlier—the creaky base problem. When implementing e-Governance projects, we must first address the weaknesses of existing systems.

You talked about setting up of a national grid that could enable better coordination and communication between governments and departments. The concept is akin that of the EU, whose Vice-President Siim Kallas and the former President of Slovenia had recently said that "European Union should ensure that all its e-Government services are turned into iGovernment ones". Do you think the time is ripe for India to create a policy framework for 'I' or Integrated Governance through standardisation of processes, platforms and technology?

I think that we are on a path towards better integrated systems and governance, and I believe that the UID number will be important in enabling this. The Authority intends to partner with public agencies across the country. These registrars will all be using the UID number in their databases, and will be adopting our security, enrolment and authentication guidelines. In time as the UID proves effective, I believe that agencies will migrate more closely towards the number and our standards. This will create significant opportunities for governments to streamline processes across agencies, share information, and integrate public programs.  

It is strange that most of the e-Gov initiatives in the country revolve around IT and not around communication technologies that certainly have better reach. Why do you think has communication-enabled services lagged behind in the overall e-Gov scheme of things?
I think delivering e-Governance services on mobile phones touches up against issues that do not exist when you deliver such services over the Internet. For example, there are interoperability concerns between different mobile service providers, if applications and services are not universally available. We still lack regulation and clear standards when it comes to financial transactions made over the mobile phone, which complicate e-Payments for government services. Third, there is the usability issue. Complex commands remain a challenge for mobile phones, especially for basic models, and voice technologies are still imperfect.

There seems to be a huge disparity in India—in terms of the growing aspirations of people and the ability of the country to even meet the Millennium Development Goals (MDG). Do you think technology can be leveraged to mitigate these challenges?  And how?
Technology can be a valuable tool to build better solutions. It can be a particularly important enabler in accessing scarce resources. For example, in a country where banks still have to reach large parts of rural India, we can use mobile phones and online banking to broaden financial access to the poor. Once that is done, governments have many more options to deliver benefits and social schemes to poor residents.

Technology also has value in building more open institutions, and giving us access to more information. The Internet and the mobile phone have already transformed the lives of India's middle and upper classes, in the access to information.

For poor income groups with scarce resources, such access to information becomes even more valuable. For instance,  farmers had information on crop prices across the country and India's National Commodities Exchange enabled them to negotiate prices and sell produce at markets of their choice.

Such access to information has had other effects — it has strengthened overall price setting by markets, and reduced the dominance of a few farming families when it came to setting crop prices.
I think this is where technology can play a critical role in tackling poverty — in the cascading impact it has in opening up and giving the poor entry into our markets and institutions.

Post new comment

Content limited to 50 characters, remaining: 50
Content limited to 1500 characters, remaining: 1500
This question is for testing whether you are a human visitor and to prevent automated spam submissions.
Enter the characters shown in the image.
Todays Main News »

Click below for your free email iGovernment newsletter