Tihar jail is installing artificial intelligence (AI)-powered CCTV cameras to monitor inmates and fight crime, officials told The Hindu. The premises will also have a real-time grievance redressal system and optical fibre network, they added.
While the AI-powered CCTV cameras will help detect movements in “dark spots”, the grievance redressal system will operate like police control rooms (PCR) to address the problems of inmates in real time.
Tihar jail – the largest prison complex of South Asia – is home to some of the most high-profile criminals. Delhi has three prison complexes – Tihar, Rohini and Mandoli – and all of them are packed beyond capacity. The maximum-security Tihar jail, which has a capacity to house 5,200 inmates, has 12,762 prisoners at present across its nine central prisons.
The overcrowding has made monitoring inmates difficult. Despite advanced security surveillance, the killing of Punjabi singer-turned-politician Sidhu Moosewala last year was reportedly plotted inside Tihar jail. Jailed conman Sukesh Chandrasekhar, too, allegedly ran a ₹200-crore extortion racket from Tihar using cell phones. On Thursday, jail authorities said 348 mobile phones were recovered from different jails of Tihar, Mandoli and Rohini during drives against smuggled objects in the past three months.
Pilot project
In the pilot project, 20 AI-powered cameras with built-in facial recognition will be installed. As many as 1,248 additional CCTV cameras will also be put in place, said a prison source.
“We will roll out the technology in a phased manner and ramp it up later,” said the source.
At present, Tihar has 7,500 IP CCTV cameras (a type of digital video camera that receives control data and sends image data via an internet protocol network). “It’s difficult to monitor all cameras. That also requires more human deployment. AI cameras would help us detect suspicious movement,” another source added.
Explaining how the technology works, a senior officer said AI CCTV cameras cover repetitive hand movements such as while assaulting another person, or hand movements suggestive of a phone call, and automatically trigger an alert to the officers. The cameras will also capture instances of inmates leaving the designated perimeter.
However, the AI-powered cameras will not be installed in Tihar’s sub-jail No. 6, which houses only women prisoners, because of “privacy concerns”.
Senior jail officials at Tihar have identified the “dark spots” at different jails where even jammers that help block 3G and 4G mobile phone signals cannot penetrate. “Since these areas have become functional for mobile phones, most prisoners have discovered them. A prison official said most incidents of assaults, ill-treatment, bodily injuries take place in these corners. “Inmates also store contrabands in these spots and use their mobile phones,” the source said.
“While a grievance redressal mechanism has been in place for years, the existing system is time-consuming and completely manual. Now, we are planning to build a system with real-time responses,” the source added.
He said the new mechanism will address physical and mental distress calls in real time. A PCR call-like feature set up in each ward will help respond to distress calls. “Staffers will also be provided with walkie-talkies to communicate with each other,” one of the sources said.
In terms of the canteen management system, authorities are planning to replace the manual billing system with an automatic one. A computerised system will be installed at the canteen area and fibre-optic cables will be installed in each jail to increase system efficiency. A biometric attendance system will also be put in place to help manage records. The process to streamline and upgrade the prison management system is also on the cards.
