Changing campus safety with AI-driven video analytics

This article was originally posted on eCampusNews

By Liam Galin

Campus safety is rapidly evolving and a highly promising addition to the suite of security options for institutions is video analytics.

Roughly 19 million students in the United States are enrolled in university undergraduate or graduate studies, and most of them attend classes onsite and keep college campuses across the country bustling with activity. As in any area where large groups of people gather, safety is a primary public concern in college and university settings.

How these institutions maintain – and, ideally, improve upon – campus safety is rapidly evolving. A somewhat recent, and highly promising, addition to the suite of security options for college and university systems has been video analytics. Of course, campus police and local businesses have long employed video cameras for the purpose of surveillance, protection, and investigation. But increasingly, the latest artificial intelligence technology is being paired with existing tools to enhance situational awareness and real-time security monitoring on college campuses.

Managing traffic. Big cities are characterized by their unique movement signatures, with distinct ebbs and flows leading to expressway bottlenecks, train station crowds and other congestion, delays, and even public safety risks. Most college campuses are characterized by similar, if smaller, ecosystems in which both auto and foot traffic require management solutions to optimize the conditions of campus commuting. AI-based surveillance systems can detect trouble spots and help administrators alleviate complex, interconnected issues.

Monitoring occupancy. University campuses are home to lecture halls and libraries, stadiums and arenas, and sororities and fraternities. They are hotbeds for activism and revelry: environments that often draw many bodies into close quarters–and occasionally, too many of them. Video analytics can monitor building occupancy numbers, track crowd patterns, and help proactively alert security personnel of the need to divert or disperse individuals to mitigate health and safety violations.

Locating missing persons. Law enforcement and security officers are only human, and even a team of trained professionals is limited in its ability to parse through extensive video footage to locate a lost child or abduction victim. AI-powered surveillance coupled with advanced analytics, on the other hand, drives faster pinpointing of relevant data in video and identifies individuals of interest based on recognizable feature classifications (such as the color hat they were wearing or whether or not they carried a handbag).

Real-time response and investigation. Whether it’s in response to a crowd or a crime, artificial intelligence enables in-the-moment monitoring and virtually instant reaction times that can prevent property loss or damage, and even save lives. Additionally, video analytics empower officials to accelerate post-incident investigations and evidence review, saving time and resources while helping bring resolution to incidents faster.

What is immediately clear is the reach and scalability of artificial intelligence-powered video surveillance and analytics when compared to past security solutions. AI is already improving situational awareness and real-time monitoring on campuses across the country. Yet the technology holds enormous untapped potential for also helping colleges and universities recognize and mitigate future risks, informing event planning and infrastructure investments with unbiased, accurate, and quantifiable data.

AI-based video analytics is the future of campus security. As nationwide adoption grows, we will  see variations  from campus to campus, based on the unique conditions at a given college or university. But, these solutions will continue to evolve over time, and campus officials and community members will learn which specific video analytics solutions best enhance capabilities for  their distinct setting, while continuing to strike an acceptable balance between sensitivity and security.