Many of the bodies recovered from the Balasore train crash site were in a poor state. (Image: News18/File)
Cyber experts said the railways used the recently launched Sanchar Saathi AI tool to identify 64 bodies and was successful in 45 cases. From photographs of bodies, it traced mobile numbers as well as Aadhar card
The biggest challenge that has surfaced after the recovery of bodies from the train accident site in Odisha’s Balasore, is their identification. It was important for the bodies to be claimed and handed over to families as soon as possible. But it was not easy.
Many of these bodies were in a poor state – disfigured, parts missing while some others were only remains. But the ministries of telecommunication as well as electronics and information technology, senior railway officials, cyber cell, government officials, UIDAI officials among others met and found a solution: artificial intelligence. Experts in the field were called in to take stock of the situation.
First, Aadhaar card experts were taken to the site where the bodies were being kept; they took fingerprints from the left thumb of the bodies. In addition, a SIM card triangulation system was used. As a result, the identification process was linked with the ‘Sanchar Saathi’ system. At least 65 bodies were fingerprinted.
But since the skin on the fingers was destroyed in many cases, identity could not be ascertained through Aadhaar card. Later, this process was added to Sanchar Saathi, which is an AI-based system, through which your identity is known on the basis of mobile connection.
Railway minister Ashwini Vaishnaw launched this system a few days ago, where the phone number of the accident victim was known through the photograph. In addition, complete information related to the Aadhaar card was also revealed. The identity of 44 bodies was ascertained through this process. Even their home addresses and family’s mobile numbers were found.
Cyber expert Prasant Sahoo said the railways used the recently launched Sanchar Saathi AI tool to identify 64 bodies and was successful in 45 cases. From photographs of bodies, this AI tool traced victims’ mobile numbers as well as Aadhar card linked with that SIM. After that, their family members were contacted. The same method was used earlier to identify fake pre-registered SIM cards. Almost 52,000 fake SIM cards in Odisha were blocked in May.
The next method used was SIM card triangulation. In this method, officials traced phones active just before the accident and were shut down or inactive soon after, an expert said.
