Sunday, 14 July 2013

Uttarakhand floods: When accounting for the missing is not such a disaster

The 15 officials at the missing persons cell, set up on June 27 in the wake of the disaster that swept through Uttarakhand, knew that their work was akin to looking for a needle in a haystack.

With just five phone lines, they were entrusted with the task of taking down complaints 24X7 of those missing in the rain-ravaged state and feeding them into the National Informatics Centre’s data bank. To ensure a watertight system, the cell, set up under the charge of Ajay Kumar Pradyot, secretary of youth affairs ministry, roped in several agencies, like the Integrated Child Development Services that has presence across the country.

It even referred to data from FIRs, Google Person Finder — an open source web application that provides a registry and message board for survivors, family, and loved ones affected by a natural disaster to post and search for information about each other’s status and whereabouts — Facebook, Twitter, television channels, the army as well as the Indo-Tibetan Police Force.

Soon enough, though, the team hit a wall. With information coming in by the hour from various sources, duplications began creeping in: the figure of those reported missing crossed 25,000 by July 5. This made it nearly impossible to cross-check names manually or even through established computing procedures.

That’s when software giant IBM lent a helping hand. It helped the cell programme a spreadsheet application that collated data under several parameters, like name, father’s name, spouse’s name, address, photograph and cellphone number. All agencies involved in the work were then asked to fill as much data as possible under each category.

“The idea was to quickly get figures from all places. The basic strategy was to come up with a list of persons actually missing, untraceable or dead so that beneficiaries could get their compensation. But with tonnes of data pouring in, this became difficult. Thankfully, IBM volunteered to help deduplicate,” says Pradyot.

IBM sent a team of six professionals to the cell from its various centres in India. To make the task easier, the cell gave each reported missing person a unique identity (eg UP/MC/0001) and then fed in all possible parameters in each.

The data compression technique has been a huge success. Together, the teams have helped narrow down the figure of missing persons to 5,000-6,000. The cell also opened a call centre and roped in the 108 (emergency) service to cross-check identities. So far, it has made over 10,000 calls.

Leaving nothing to chance, it simultaneously asked all mobile service providers to provide call data records of people who had made calls or sent SMSes from Uttarakhand from June 14 to June 21. Besides, it asked them to provide information of cellphones in Uttarakhand that were disconnected between June 16 and June 23.

“This mammoth exercise was, perhaps, the first in the world. The idea was to get numbers of phones that had died, establish identities of their subscribers and then connect with their families or friends to check if they had reached home.

That was difficult too since in Uttarakhand, the telecom service provider changes as one moves from one circle to another even in a day. We hope to finish the task by July 15,” assures Pradyot.

The Uttarakhand government has asked all states to furnish a list of people who went missing in the state during the deluge. Following this, the list of those presumed to be dead will be handed over to a special committee looking into compensation, which is expected to be over Rs300 crore.

Sunday 14 July 2013


Post a Comment