Exclusive Interview: Nandan Nilekani, Chairman of the new Unique Identification Authority of India (UIDAI)
05/07/10
1. In a nutshell, what is the scope of this project by the Unique Identification Authority of India (UIDAI) and why is it being undertaken?
The goal of the project is to issue a unique identification number (now known as ‘Aadhaar’
[1]) to every resident in the country. The scope of the project is ambitious – enrolment into the number hopes to eventually cover all 1.2 billion people in the country, and 600 million people over the next four years. The number will be linked to the individual’s basic demographic and biometric details, and will be stored in a central database.
Post-enrolment, every time the resident wishes to confirm his or her identity to an agency or service provider, the agency can contact the central database with the individual’s unique number, along with the demographic or biometric information that the person has provided. The database would then check if there is a match in its records, and respond in real-time with a ‘yes’ to confirm the identity, or a ‘no’.
2. Why is biometrics central to the entire approach?
A challenge for residents in India, particularly the underprivileged, is the lack of documentation they have available to establish their identity. Many of India’s poorest residents don’t have basic identity documentation such as a birth certificate, or a proof of address such as an electricity bill. These residents are consequently unable to access services and resources – such as opening a bank account, or renting a house – since they are unable to fulfill the KYC (Know Your Customer) requirements these agencies have. The challenges increase when people migrate, since most identification documentation in India is provided by local administration, and invalid when the person crosses state lines.
Biometrics simplifies such identity verification, especially for individuals who are in the poorest groups. Once these residents enroll for an identity number linked to their biometrics, they can subsequently prove their identity multiple times, anywhere in the country and to any agency, by providing their demographic details or biometric scan.
3. Does it not worry you that a system on this scale has not been rolled out before?
India has excelled at taking projects to an unprecedented scale and making them work. Due to its size, India has seen skeptics virtually from the beginning of its existence as an independent nation – if you look back at the press coverage surrounding India’s independence in 1947, much of it predicted that India would not survive as an independent democratic country due to its size and diversity – democracy in India would ‘fall back’, ‘retreat’ and ‘disintegrate’.
India has since managed the demands of its size and population quite effectively while rolling out nationwide implementations. For instance, every five years India conducts national elections where record numbers of people vote – the logistics and infrastructure rollout that is required is incredible. In the last elections, over 714 million people were eligible to vote; approximately 58% of the electorate cast their ballots.
The UIDAI is taking steps to ensure that the scale of the initiative is managed effectively.
a) The UIDAI has built a highly consultative, inclusive process in selecting the technologies that are most effective for a project of this size. For instance, we involved the world’s top biometric experts as consultants on the biometric systems for the initiative, and are collaborating with world class institutes such as NIST.
b) We are employing open processes to design and review the UID architecture. We have partnered with dozens of subject matter experts from across the world, who are involved in helping us design our systems.
c) We are committed to open standards wherever possible, and the use of open source in the initiative has allowed us to tap into a large community of experts that supports us and gives us feedback.
d) Our design architecture is completely vendor neutral and scalable. For example, we intend to deploy three separate biometric solutions for de-duplication. In our design we can route requests to any, or all three biometric solutions. We’ve been complimented for creating a revolutionary architecture that could change the way biometric systems are being designed around the world.
4. What modalities were chosen, and why - in particular, iris?
The modalities we have selected for the project are face, finger and iris capture. The idea behind using multiple modalities is that it would ensure effective de-duplication – we are keeping in mind that we have a population of 1.2 billion to de-duplicate, so uniqueness is a significant concern.
While capturing face and fingerprints is expected, the use of iris biometrics for such a large project may have attracted the biggest ‘why’. The reason we included iris was twofold. One was the concern surrounding uniqueness.
The government had constituted a Biometric Standards Committee, headed by Dr. B. K Gairola, to advise us on choices of modalities
[2]. The Committee recommended the inclusion of iris to the biometric modalities, would help guarantee the uniqueness of the number, by increasing the estimated accuracy in de-duplicating the individual from 95%, and ensuring a very low FAR (False Acceptance Rate). The decrease in error from including iris was clearly substantial, an important factor to consider for a population of 1.2 billion, and if the unique number is to be usable in high-security applications.
The second reason for adding in iris was inclusion. The challenge with both fingerprint and face biometrics is that these are limited when it comes to providing a unique number to children – these biometrics do not stabilize before the age of 16. Iris biometrics however, are stable from infancy, and can be collected from children approximately from five years of age. This is an important factor considering multiple programs aimed at child welfare. The use of iris will also enable us to ensure the inclusion of the very poor, many of who work in physically intensive jobs. People working in such jobs – for example, in fireworks factories in arecanut plantations – often have their fingerprints rubbed off, which makes iris useful in ensuring a unique biometric.
5. How will you be handling inevitable biometric exceptions?
The focus of the UIDAI is in enrolling the poor, and the goal of the Aadhaar is in ensuring better delivery of social services. Consequently, we are emphasizing ‘zero failure to enroll’ – we must enroll everyone who wants an Aadhaar. We believe that limiting exceptions is possible through careful enrolment process design and by using multiple biometrics. We’ve also designed appropriate enrollment processes, by defining exceptions (people without hands/fingers, people who are blind and whose iris cannot be collected, etc) and how to handle them.
6. What will be the future use of biometrics in commerce and social benefit delivery (entitlements)?
Biometrics can play a very important role in both our social programs, as well as in the working of our markets. Identification is often a key barrier in accessing private and public services in India, resulting in denial of service to the poor. The lack of effective, non-repudiable identity verification has prevents government and private agencies from providing accessible financial services, as well as remote payment, retail and information services to the poor – services that would be very useful to BOP (Bottom of Pyramid) markets as well as for residents in hard-to-access rural areas.
7. How is UID different from the UK’s cancelled national identity project? What lessons do you take from the UK?
The focus of recent identity projects in the UK, and in the US with RealID, has been security and regulating immigration. The principle behind the UID initiative is to ensure better and more effective welfare services, and give individuals greater capacity to access the resources and services they need. In this sense, this effort is more like the social security initiatives of the early and mid-20th century, and the Aadhaar number will be similar to the national insurance number in the UK and the social security number in the US. It is a universal number which will ease problems in identification, and simplify procedures for a large number of agencies that people interact with.
The more recent efforts in UK and US however, did carry some lessons for us – and an important one was that a nationwide identity initiative needs popular support to work. We’ve taken steps in the last few months to popularize the goals of the project, and the benefits people will gain from it. Our information and education efforts will intensify in the coming few months, and will adopt multiple strategies to communicate the value of the number to people in rural and urban India.
8. What was your strategy for building the system? And to such tight deadlines…
Our enrolment strategy is an open, collaborative one. We encourage a variety of agencies to partner with us – state governments, banks, post offices, telecom companies – to issue Aadhaar numbers to the residents they interact with, and their customer base.
A wide range of such enrolling agencies for Aadhaar will ensure that we meet our enrolment deadlines, and cover a large segment of the population in a short period of time. It will also give individuals choice in the service provider they apply to for their identity number.
9. How has the project gone to-date? What should we expect in the next twelve months?
The UIDAI has completed some important milestones – we’ve carried out multiple proof of concepts (PoCs), where we tested our enrolment processes in rural areas, as well as our biometric and de-duplication technologies. Our goal is to start enrolling residents in August this year, and rapidly scale enrolment over twelve months from that period. We will also roll out our communication efforts in the coming months, so that we can spread awareness of the number across the country.
We also hope to deliver some benefit to Aadhaar holders in this first phase of the project, so that the benefit of the number is clear. One way of doing this is to link Aadhaar’s strengths to India’s financial inclusion effort. The Reserve Bank of India (RBI) has set the objective that by March 2012, all villages with a population of over 2,000 have access to a bank branch. To enable this, the RBI is likely to mandate that banks can accept the Aadhaar as sufficient KYC to open a basic bank account, and can authenticate bank account holders using their Aadhaar-linked demographic or biometric information.
10. Do you have a message for biometric vendors?
The Unique Identification Project is creating new opportunities for biometric technology – these solutions have never been employed on the scale of this project before. The use of biometrics in the UID initiative will begin with enrolling residents for the number, but such use will expand quickly as various agencies and applications use the number and biometric-linked information to verify the identity of individuals across the country.
Our success can therefore, determine the course the industry will take, since these technologies will be tested in India on an unprecedented scale. We are working with biometric device vendors to build the next generation of devices for large scale enrolment and authentication applications. We are also working with software companies to create the most innovative multi-modal multi-vendor architecture, and engaging hardware manufacturers to utilize the most cost effective hardware for large scale biometric identification/de-duplication. Finally, we are building the entire application on an e-governance cloud platform, using open source products.
We need your active collaboration and participation here, to help us forge new directions for this industry. We are implementing this process in the most inclusive, open manner that is possible, and we hope that the best of the industry will join us in shaping a great future for biometrics, and in making this initiative a success.
[1] The name Aadhaar translates into ‘foundation’ or ‘base’ in several Indian languages.
[2] The committee submitted their report on December 2009, and is available on the UIDAI website.
This article appears in :-
Article Comments
No reviews have been submitted
Please add a comment on this article from the Add Comment below.