A Database is Born: NADRA in the Early Days

Azhar Nawaz walks us through challenges overcome and the impact NADRA has had!

Technology is only as good as the people it empowers and the benefit it brings. The success of an application, software, project or initiative cannot be measured in how much money was spent or how many companies it awarded contracts to, rather it has to be a long term assessment of the gaps it filled and the needs it fulfilled. If it turns into a platform which would impact the lives of millions, then the value it brings is definitely greater than the worth of the project itself.

Perhaps the inherent challenge that technology faces around the world is when it is viewed in isolation. A box or a system that has a specific purpose. When it comes to viewing e-Governance, if e-enabling projects were viewed in the same manner, they would fail quite miserably. But if you learned to look long term, and saw the kind of impact that the project had on transparency and accountability, not only is the project deemed a success, but the initiatives that it will further enable, will also be a probable and significant contribution to the uplifting of an economy. And yes, government projects which are deployed at the mass scale take a tremendous amount of time because of the sheer numbers they need to reach out to.

The National Database and Registration Authority or NADRA was created to address a need -- the need to know what the composition of the citizen is. Who is out there, and how they are legally registered to use their right to interact with official authorities of a country.

National Database and Registration Authority (NADRA):The National Database Organization (NDO) was established under the Ministry of Interior as a sister organization to undertake the task of handling the data being collected through National Data Forms during the Population Census in 1998. The National Data Forms (NDFs) were designed by a committee of experts encompassing maximum social and fiscal indicators as implements for future planning, documentation of the economy, and for creation of a comprehensive Citizens' Database. To cater to a cohesive and unified approach in both registration and social fields, NADRA was formed on 10th March 2000 by merging the National Database Organization (NDO) and the Directorate General of Registration (DGR). It is an independent corporate body with requisite autonomy.

NADRA was formed to aid the government in developing and implementing a fact-based system of good governance in the country to encounter undocumented population growth and register items belonging to its citizens and organizations. This data would prove to be a critical factor in everything from city planning and addressing the population explosion, to better and more effective allocation of resources for densely populated geographical areas.

It was established with the objective of evolving a new, improved, and modernized system of registration for the citizens of Pakistan, to undertake the collection of demographic and geographic data. Initially, the main objective was the issuance of state-of-the-art National Identity Cards (NICs) to all eligible citizens of Pakistan. These NICs were backed by a computerized database and a data warehouse respectively called the Citizens' Database and National Data Warehouse (NDW). This comprehensive Citizens' Database was completed at the dawn of the new millennium and is regularly updated based on the National Data Forms.

Azhar Nawaz was already interacting closely with NADRA while working at Jaffar Brothers Limited, on two milestone projects: One, Computerized Electoral Rolls and the National Database Forms Project.

The Computerized Electoral Rolls, covered in the CIO Pakistan August 2008 issue, was the complete solution for Electoral Roll, which are the lists of voters printed onto sheets and used during elections. The project included everything including Data acquisition, printing/update of electoral roll, and consolidation of data into national data-warehouse. Close to 65 million voters were registered in the process. This electoral roll was used in the first local bodies (Nazim) election in 2002. His second project was the National Data Form Project which was the computerization and data acquisition of the 10-year census, executed in 1998. The paper-based census forms were computerized and 72 million registrations were completed in 8 months during 2000-2001.

A New BeginningAzhar was already working closely with NADRA on key projects and the GM HR of Jaffer Business Limited was contacted by the Chairman NADRA to transfer his services to them. "After a 45-minute interview at NADRA," says Azhar, "I was hired at the position of a Senior Analyst for Project Integrations." Azhar Nawaz was the lead on projects such as CNIC, NSRC, Mobile Registration Centers, and others where he manned everything from project management, integration research, execution and delivery.

"The Data Entry Computerized National Identification Card (CNIC) was already in progress. I took the lead role and improved the data acquisition issues. You see," explains Azhar, "the NADRA offices in the Tehsils and Districts distributed the forms. People would collect these forms, fill them out and return them to the office. Even with this just much information, there are already a lot of obvious problems with this methodology." The original method was to enter the data from paper-based forms into an electronic format. The scanned versions of forms were available on each terminal and the data entry operator would read the forms and fill them in electronic format.

"There were lots of mistakes in the process, where bad data would get added in the database. There were several groups of 21 members each in a hierarchal structure, where the senior would randomly check the quality of data entered into the forms. Lots of complaints were coming from customers related to wrong information in their ID cards." Azhar observed the ongoings and introduced several changes. "There needed to be a double data entry system where every scanned form was sent in queues to 2 different operators. Any mismatch in data entry between 2 operators was sent to an examiner, who would manually correct the data. Also, a point-based ranking system was introduced to identify operators with poor data entry skills."

Though the implementation of these minor changes helped to improve the quality of data, there were still several issues in the overall process and data quality. "The fingerprint and photograph qualities, for example, were not good. The scanned photographs and fingerprints could not be properly used for automated recognition. There was also no Optical Character Recognition (OCR) for Urdu and each data entry operator interpreted the handwritten forms differently, which led to inconsistent data quality. To overcome these issues, I led a project to implement online registration centers all over Pakistan capable of advanced biometric capturing. The project was called NADRA Swift Registration Centre (NSRC)."

The first NSRC was launched in November of 2001 in Rawalpindi as a pilot project for the duration of 3 months. The center was launched with 8 data entry stations and it began registering approximately 300 customers per day and quickly become very popular.

"Despite reservations and resistance from some members, we implemented something unique in Pakistan that was focused primarily on customer convenience and gathering quality data. Each center had a number of data entry terminals that would take information directly from customers and feed it directly into computers without using any paper forms. The center also had biometrics equipment, a server with RAID storage and some networking equipment that would connect all the terminals through a LAN and provide WAN connectivity to the head office."

In this scenario, the customer would verbally give his information, which would be directly fed into the system.

After this data entry, the biometrics information would be collected digitally where he would neither have to ink his fingers nor bring any photographs from outside. "We were interested in maintaining the integrity of the data that was being fed into the system." Processes such as a Computerized Token and Queue Management System was introduced for customer care where large LED displays would show the queue position. "After the initial data entry, the customer would wait in a second queue for photographs and fingerprinting. Once completed, a printed copy of his computerized data would be provided to the customer so that he could verify it and take it back for attestation from a gazetted officer or his locally elected representative such as Nazim or Councilor.

At end of every business day, the data stored on the local server would be uploaded to the Head Office through WAN for processing." The implementation wasn't an easy task. The effort that the project was putting in was monumental. In order to get the requisite resultant success, it was critical that the data being input was clean. "In addition to our resources, the project was integrating millions of people. We had to take every precaution possible in order to secure the quality of the data. There was just too much at stake," explains Azhar.

To implement and optimize the first centre, in addition leading the project and being the project manager, Azhar had to invest a lot of time in research. "I had to understand how photographers used lighting to get best picture results and replicated these concepts across the NSRCs."

In addition to the technology, there was a need to observe the improvements that were to be made. Observing human behavior and understanding how their psychology works, what would make a line move more efficiently or how a question should be best phrased in a form -- all these things were part of the learning process. Each process needed to be experienced from a usability perspective, and then implemented within the IT infrastructure we had available to us. "The alignment had to be something that felt natural and comfortable to do."

Due to popularity and success of the NSRC at Rawalpindi, the Interior Ministry gave NADRA a target to establish 120 similar centers throughout Pakistan by June 2003. "I led the project and we had 120 operational centers connected to NADRA HQ in time."

Over the next 3 years, Azhar and his team established more than 400 NSRCs around Pakistan and abroad. "It was a huge success and we were registering up to 50,000 citizens in day during peak registration period." There was no previous history of this kind of project anywhere in this world. What makes you appreciate this fact to an even greater extent, is the fact that NADRA was working within the confines of a political bureaucracy and red tape that was simply not used to such rapid escalation of any project. The kind of technology that had been deployed may have been extremely complex in its time, but it served a basic purpose -- to build a database and archive the records for every registered citizen of Pakistan.

Customizing Technology To Be Intelligent2003 gave way to Mobile Registration Centers and the concept was fairly similar to that of the NSRC. "We implemented the center in a high tech van that had all similar equipment. Complete interior of mobile vans was custom designed to provide similar workflow as the NSRC, with customer convenience as a first priority. 3 different sizes of mobiles vans were chosen (Hino, Mazda, Shehzore) to accommodate different number of data entry terminals." NADRA coordinated the registration requirements with local elected officials in different rural districts of Pakistan and scheduled the visits to these towns. The local officials would announce the schedule well in advance to residents of his Tehsil/District and people would make themselves available for registration at the mobile NSRC at their town.

"At the end of the day, the mobile vans would arrive back at their Base Centers and connect themselves to NADRA HQ through connectivity at Base Center to dump the registration data. By end of project, we had 150 mobile vans operational."

Managing DataBy the time the team was in the middle of the first few projects, it was becoming painfully apparent that they needed powerful tools to mine the data that was being archived. Azhar explains, "there were now so many different sources of data collection which was all being aggregated in one place, we needed a way to be able to retrieve the data efficiently." The team sat down and identified the solution to yet another pain area.

There was a huge need of managing every process going at NADRA for customer services and integration. "We developed an in-house CRM (Customer Relationship Management) system at NADRA and linked it to all critical systems. The CRM system, called the Customer Relationship Management Project, enabled NADRA to track every NIC application from start to finish with Call Center support. The system was also used for MIS decision-making purpose for things like progress, data quality, performance and census information for government planning."

1 2 Page 1
Page 1 of 2
Now read: Getting grounded in IoT