AI and machine learning adoption is the future of security and law enforcement

Video analytics based on AI/ML algorithms are gaining popularity with time and are soon expected to become a requisite for all law enforcement departments and public places to optimise safety

By Lopamudra Roy and S. Arunachalam

Published: Jan 7, 2022 03:40:51 PM IST

Updated: Jan 10, 2022 04:54:23 PM IST

Image: Shutterstock

Rising crimes and safety concerns worldwide have made video surveillance an utmost necessity across all industries for physical security. With inflating global threats, governments are becoming increasingly proactive and investing heavily in deploying video surveillance infrastructure to manage crowds or detect threats across sectors like defence, railways, traffic management, to name a few among many.

However, the massive amounts of video data come with challenges—from storage to real-time surveillance to data security. Emerging tech, such as cloud storage, video encryptions, and AI-enabled tools, can bring a paradigm shift in surveillance and security. AI algorithms can analyze each frame in real-time, and its self-learning capabilities enable identifying, tracking, and controlling any object in the recording.

Indian emerging tech startups have built solutions based on video analytics technology and computer vision to step up safety and security, law enforcement, and traffic management while assuring protection even in high traffic zones or crowded locations. Forensic investigations of CCTV footages using high-speed AI-powered video summaries and precise object detection technologies can potentially aid state police departments and law enforcement officers in preventing or significantly controlling crimes, such as attempted robbery, mob shootouts, kidnapping, theft, and so on; even in high-traffic or busy areas, or large government campuses.

India-based tech startup Dragonfruit AI works with the law enforcement departments to provide them with necessary and quick information in a crisis. The company’s Video-search and Summaries solution can be overlaid on existing CCTV infrastructure with no additional installation. In case of a crisis, Dragonfruit can pass on necessary information to the law enforcement departments with a click of a button to help them expedite coordination with their relief and rescue teams and address the issue at the earliest.

AI fast-forward techniques enable significantly long hours of video data sets to be scanned through and reviewed to identify and gather evidence of a crime or potential crime in just a few minutes. Accurate filter and search features can help the police identify suspects and objects, such as people or vehicles, precisely in no time. While existing technologies take a lot of time to gather evidence from lengthy CCTV footage making the investigation process lengthy and tedious, the new AI-powered video analytic solution reduces video reviewing time considerably and enables swift investigation and action. On-demand “pay-as-you-go” solutions enable the installation and subscription pricing for only those functionalities that the customer/client requires. Also, the new technologies can be overlaid on existing cameras or infrastructure without incurring expenses for additional installations. These features make the solutions affordable and cost-effective for public law enforcement departments. Furthermore, online reports can be generated using evidence captured and seamlessly and swiftly shared with relevant stakeholders through secure URLs instead of time-consuming emails or physical USB drives.

Public transportation or traffic police departments can also deploy these solutions to search through crowds, locate accident locations in high-traffic zones, and ensure safety protocols in crowded areas. Besides producing quick summaries of lengthy videos, augmented video search capabilities based on a wide range of filter features such as object colour and type can help locate objects even in crowded places. This can help public transportation services like railways and metros track people and objects even during peak traffic, thereby preventing theft or life-threatening incidents and ensuring safety and security to daily commuters. Traffic police officials can also use these technologies to be overlaid on their existing cameras to locate road accidents and ensure immediate care to these locations. During the pandemic, the technology can also be deployed to identify people violating Covid-19 protocols and ensure adherence to public safety policies.

AI-powered video analytics technology can also help various public sector departments monitor the effectiveness of an already installed network of cameras—from tracking infrastructure health, camera tampering or failures to seamlessly managing a large network of equipment.

However, though state-of-the-art technologies can resolve operational issues of public sectors with advanced and cost-effective solutions, it requires considerable tech-based education and training for a ready and seamless adoption. Video analytics based on AI/ML algorithms are gaining popularity with time and are soon expected to become a requisite for all law enforcement departments and public places to optimise safety.

Lopamudra Roy is a Research Associate-Content Development at ISB’s Centre for Innovation and Entrepreneurship.

S. Arunachalam, an Assistant Professor in Marketing and Academic Director at ISB’s Centre for Innovation and Entrepreneurship.