The views expressed by guest bloggers and contributors are those of the authors and do not necessarily represent the views of, and should not be attributed to, Campus Safety.
The Department of Education discloses that in 2020 alone, over 28,000 criminal offenses were reported on college campuses. We can safely assume that with many going unreported, the actual number is much higher. Some of the most serious crimes, including violence and sexual assault, can take place in on-campus student housing facilities or dorms.
Ask any college student and they’ll tell you — it’s not difficult to get into a dorm that isn’t yours. Thanks to the proliferation of keycards as the access security method of choice, coming across a card left in a classroom, dropped at a party, loaned out by a friend, or even stolen, can with one fell swipe grant you entrance into the living quarters of hundreds of students.
Exacerbating the concern, students living in dorms frequently leave individual room doors open and unlocked as friends migrate the halls and stop in to say hello. Taking this into account, it’s surprising dorm security hasn’t become more advanced in recent decades where solutions more secure than keycards have been available.
Coming out of the pandemic, many institutions have decided to finally address the issue as part of wider facilities upgrades. Artificial intelligence (AI)-powered computer vision technology has become a new access control method of choice for student housing for its high security and convenience, paired with reasonably low cost and flexible software and hardware options.
The Keycard Era
Keycards were for a long time the most convenient and cost-effective security tool, so it’s understandable that they became the most prolific access control method for many years on campuses. Cheap cards are easily purchased, programmed, and handed out, and students and staff gain a sense of security knowing they must swipe to enter a building.
In reality, keycards offer the illusion of security rather than security itself. With absolutely no guarantee that the user of the keycard is its intended owner, and no verifiable record of who entered the building when, they become frustratingly opaque when campus security needs to understand who gained entrance or you wonder how a stranger has ended up in your dorm hall. With thousands of students on campus, the chances of coming across a lost ID badge on any given day are quite high.
While keycards have been long known to suffer from these security weaknesses, the fear of losing their convenience and low cost has for a time held back progress. Finally, a stronger focus on security upgrades and touchless solutions has sprung forward the adoption of more modern access control technology, especially to safeguard student spaces that are the most vulnerable, such as dorm halls.
Enter: AI-based Security
Biometric AI vision technology is seeing some of the most rapid adoptions on college campuses worldwide. Schools like Beijing Normal University were early adopters, first using AI to secure their women’s dorms over five years ago. Many schools in the U.S. have integrated this technology for the first time this school year.
Biometric AI vision for access control works by comparing live video of an individual to a known facial template. Students opt-in to the program and provide consent as part of their dorm agreement, and from there benefit from extremely high security in their living spaces, plus unparalleled convenience when it comes to having 24/7 access to their dorms, whether or not they have a card on them. Students at Beijing Normal University specifically called out these aspects when describing their high satisfaction with the new security system following its implementation in 2017.
Since then, AI vision solutions have only advanced further, with industry-leading providers reaching near-perfect accuracy and offering extremely adaptable options. Thanks to highly flexible software development kits and low hardware requirements, it’s often a matter of adjusting or complementing existing visitor management systems (VMS) or security systems, rather than starting fresh. Software can even be programmed to alert specific dorm security officers if an unrecognized or banned individual enters a certain hall.
Another reason AI vision has become one of the top new security methods of choice on campuses is because of its ability, unlike keycards, to both guarantee the intended user is the one who gains entrance and to provide verifiable records of entry, a critical forensic tool for safety administrators and police.
Some schools, such as UCLA, have inquired about the use of the technology and ultimately have not moved forward with integration based on privacy concerns. Today, many leading facial recognition providers follow highly stringent privacy guidelines and the software is entirely compliant with biometric privacy laws across states. Leading providers offer software that does not store any facial photos at any time, instead leveraging secure, encrypted facial vector templates that are unintelligible if breached. Student security and privacy have never been so top of mind and AI vision security has offered a leap forward on both fronts.
After two years of hybrid and remote learning environments, more students than ever have headed into their first full on-campus school year this fall. Their safety must be prioritized, and to do so, we must move beyond the outdated keycard era.
Terry Schulenburg is the Vice President of Business Development at CyberLink. Terry’s 35+ years of experience in the technology space include roles at Blackboard, Genetec, Apple and more, and specializations in campus safety and AI security.