New Delhi: After the Tirupati temple, authorities at the Mahakal Lok Corridor in Ujjain, Madhya Pradesh, are using facial recognition technology (FRT) along with artificial intelligence (AI) analytics for crowd management.
The Mahakal Lok Corridor — a plaza area spread over 2.5 hectares for devotees visiting the Mahakaleshwar temple — gets a daily footfall of 1.5-2 lakh. During festivals and days of religious significance, for instance every Monday of the Shravan month, this number goes up to 4.5-5 lakh.
Phase-I of the corridor was inaugurated by Prime Minister Narendra Modi in October last year.
There are 700 high-resolution CCTV cameras installed in the newly developed corridor, temple and adjoining streets, all of which are connected to a control room located within the temple complex. Here, a dedicated team of 50 people monitors the crowds and directs field staff to take measures when needed to ensure smooth movement of visitors.
“We officially started using the AI-enabled facial recognition technology for crowd management at the Mahakal Lok corridor from 1 July,” said Roshan Kumar Singh, commissioner of Ujjain Municipal Corporation and executive director of Ujjain Smart City Limited, adding that the temple had been testing the tech since February.
“Close to 5 lakh people visited the complex last Monday and the use of this technology has helped us in crowd management,” he added.
Explaining the use of FRT, Sandeep Soni, administrator of Mahakal temple and chief executive officer of Ujjain Development Authority, said, “When a person enters the temple complex, a profile ID is generated using their face and appearance. The facial recognition technology helps get the exact head count of people visiting the complex and we can track the person’s movement inside the Mahakal Lok.”
“As CCTVs are installed at strategic locations, it is easy to track the movement of the person within the complex,” he added. “The data is stored for a period of 15 days.”
The use of FRT has also enabled the temple administration to have a dedicated entry gate for Ujjain locals to give them quick access to the temple complex. This ‘Avantika Dwar’ was opened on 11 July.
“Earlier, we used to get 30,000-40,000 people daily, but now the numbers have increased manifold. Due to this, locals find it difficult to enter the complex and have been complaining about this for some time,” said Ujjain Mayor Mukesh Tatwal.
For this, residents have to register with the temple administration once using their Aadhaar or any valid identity card and get their face scanned for the records. “After that, though, they will be allowed entry without an ID card and will require to simply have their face scanned at the Avantika Dwar to enter,” said Soni.
According to temple administration officials, the Mahakaleshwar temple complex is among the few in the country which sees such a large number of visitors — close to 2 lakh daily — and has adopted AI and FRT for crowd management and security.
In March this year, Tirumala Tirupati Devasthanams (TTD), the trust that manages the Tirupati temple in Andhra Pradesh, started using facial recognition technology on an experimental basis in accommodation booking and to ensure hassle-free entry for devotees. And though the Kashi Vishwanath Dham corridor is yet to dabble in FRT, it is using tech to get real-time data about the number of people entering the complex.
The authorities of the under-construction Ayodhya temple, too, are planning to use hi-tech solutions for crowd management once the Ram temple is inaugurated next year.
The use of AI and FRT, however, has raised concerns.
Speaking to ThePrint, Prateek Waghre, Policy Director at the Internet Freedom Foundation, said, “We don’t have data protection legislation in effect right now. There are multiple issues with the use of facial recognition technology for mass surveillance. There is an element of inaccuracy of the technology and it can have serious ramifications if someone is misidentified. It should be used with the consent of people. But, in most cases, people don’t have a say.”
Also read: Why India can’t let private sector develop facial recognition tech for law enforcement
Using tech for security, streamlining crowds
The Mahakal Lok corridor administration is using 10 AI-enabled analytics to check the density of people at a particular location and protect the infrastructure against damage.
In the phase-I of the Rs 850-crore corridor, the plaza area surrounds a lotus pond that holds a statue of lord Shiva along with water fountains. A 900-metre-long Mahakal Path within the corridor contains 108 stambhs (pillars) and has murals on the wall running along it.
To protect this corridor from damage, the temple authority has geo-fenced certain areas.
“The statues, mural walls, green spaces and other areas where entry of people is restricted have been geo-fenced. If anyone tries to touch the statues or the mural walls or enters prohibited areas, an alert is generated in the control room and the guards deployed at the site are informed about it on their phones. The NCRB data is also proposed to be integrated with the system to help in alerting the police if a wanted person or criminal enters the complex,” said Roshan Kumar Singh.
The TTD, which is using face recognition and AI, said in a statement to the media in February, that the “idea is to enhance transparency in tokenless darshan and accommodation allotment systems, and provide more effective services to the multitude of visiting pilgrims.”
Explaining the process used at the TTD, L.M. Sandeep, general manager IT, said, “The use of facial recognition technology is limited to just two important facilities — booking accommodation and visiting the Vaikuntam complex for darshan.”
For smooth movement of devotees at Tirupati temple’s Vaikuntam complex for darshan, the temple authorities are using facial recognition to ensure devotees can’t enter the complex multiple times. “One person is allowed entry into the Vaikuntam complex only once in a day. We have installed a camera at the entry point, where the person’s photo is captured. In case they try to enter again, they will not be allowed as the system will generate an alert. The person entering the complex is issued a token for the prasad. People often try to enter again and again for the prasad. This new system has helped in crowd management,” said Sandeep.
Now, Ayodhya temple, too, is planning to use hi-tech solutions. Vishal Singh, Vice-Chairman of Ayodhya Development Authority, told ThePrint, “Close Rs 80 crore is being spent on installing latest gadgets, including CCTVs, to provide security and manage the crowd. We are also planning to use facial recognition technology.”
At the Kashi Vishwanath Dham corridor, too, there was a massive turnout of pilgrims and tourists after it was inaugurated in December 2021 by PM Modi. Here, the authorities are not using facial recognition technology, but are using headcount cameras to get real-time data about the number of people entering the complex. Around 400 CCTV cameras have been installed at Kashi Vishwanath Dham corridor to help with this.
“Last year, the corridor was visited by close to 7 crore people,” Sunil Kumar Verma, chief executive officer of Kashi Vishwanath Temple Trust. “We get around 2 lakh visitors daily and the number sometimes goes up to 5 lakh. We now more or less have exact data about the number of people visiting the complex. This helps us in better planning, especially during important days.”
Concerns over privacy
Kamesh Shekar, Programme Manager, The Dialogue, a public policy think tank, has however said that there is no legal framework that specifically governs the public and private sectors’ deployment and usage of facial recognition technologies.
“Although technologies like FRT and data collected through it are used for security purposes and preserving state order, they are developed without a data protection regime and robust surveillance reform, which is concerning. Moving beyond the legality, this also brings to the fore the question of what are the legitimate use cases for deploying FRT systems, where suggested objectives by Ujjain temple, like crowd management, bring the question of proportionality of using such systems at the cost of privacy.”
When asked about the privacy issue concerning use of facial recognition, Soni said, “We are having video recordings of people entering the complex. These days you have CCTVs everywhere. We are just using this to ensure safety of people and the complex. It is helping people have smooth access to the complex.”
(Edited by Zinnia Ray Chaudhuri)
Also read: ‘Whose history?’: In Telangana and Andhra Pradesh, the archives are fighting