A drone hovers silently above a congested arterial road, scanning traffic movement and feeding live images to a command centre where algorithms map vehicle density, detect bottlenecks and compare faces against suspect databases within seconds. Across the city, officers track repeat offenders and vulnerable hotspots remotely through live analytics instead of relying solely on physical patrols.

This is the policing architecture that the newly formed Malkajgiri commissionerate now wants to build - a technology-driven “plug-and-play” model powered by drones, artificial intelligence, predictive analytics and integrated surveillance systems aimed at transforming traffic management, crime detection and public safety response.

Speaking to The Hindu, Malkajgiri Police Commissioner B. Sumathi said the vision stems from the mounting pressure on policing in one of Telangana’s most densely populated commissionerates, housing upscale residential neighbourhoods, the Secunderabad Cantonment zone, major industrial clusters and some of the city’s busiest traffic corridors, where conventional policing methods are increasingly inadequate for handling complex urban movement and rapidly evolving security challenges.

“Unless technology is used in a proper and uniform manner, results cannot be achieved. We already have extensive suspect databases. That is a wealth of intelligence which has to be utilised for effective policing and reducing dependence on manpower,” she said.

Officials said the idea is not merely to introduce gadgets into policing, but to create a live intelligence network where drones, CCTV feed and police databases continuously interact to generate actionable inputs in real time.

Under the proposed model, five to six AI-enabled drones, equipped with facial recognition and AI analytics, will routinely monitor isolated stretches, vulnerable crime zones and identified hotspots during late-night hours or between patrol cycles when manpower deployment is thinner. Live drone feeds and CCTV footage will be analysed against existing suspect databases to identify anomalies, detect repeat offenders and flag suspicious movement patterns.

For traffic policing, the system will analyse vehicle density, flow rates and congestion patterns across key traffic corridors during peak hours, allowing officers to intervene before bottlenecks spiral into gridlocks.

Officials also plan to use the system to secure VIP movement corridors, monitor suicide-prone locations and automate evidence-based processing of certain petty offences through digital documentation and surveillance-backed enforcement.

For example, officials said drones could be used to monitor establishments repeatedly operating beyond permitted hours. Instead of officers visiting the same location every night, live images could be compared with existing databases to automatically flag repeat violations, generate e-petty cases and issue non-contact challans or fines depending on the offence.

Each drone can potentially replace three to four physical patrols and give police live situational awareness, enabling quicker decision-making and response.

The broader framework was first tested during the Medaram Jatara in Mulugu district through an AI-driven drone policing project named TG-QUEST, developed to manage massive crowds and traffic movement in a forest region with almost no internet connectivity.

Police created a temporary command-and-control grid by pushing internet access across nearly 15 to 17 kilometres using antenna-based systems. AI-enabled drones, flying at intervals of roughly every five kilometres, streamed live wide-area aerial footage to dashboards at the control room, allowing officers to monitor crowd and vehicle movement in real time, manage parking flow for nearly two lakh vehicles daily without major congestion and trace missing persons.

The commissioner acknowledged that the initiative would require substantial investment in drones, analytics infrastructure and technical training. A dedicated lab is also being developed to repair and maintain drones internally, while specialised analytics teams are being trained in-house.

As part of that effort, the commissionerate recently mapped the educational background of personnel recruited between 2020 and 2024 and found that nearly 620 among roughly 2,400 constables and assistant sub-inspectors were engineering graduates.

These personnel are now being trained to lead technology-oriented policing units focusing on drone operations, surveillance analysis, backend analytics, real-time operational monitoring and various security verticals at police station level.

Once implemented, the project will mark a major shift in policing, from reactive field response to predictive surveillance driven by live analytics, aerial monitoring and integrated intelligence systems.