It’s nearly impossible to talk about technology these days without mentioning artificial intelligence (AI). Most people are familiar with generative AI tools including ChatGPT, Google Bard, and Microsoft Copilot. These companies incorporate the power of AI into the software we use every day. But did you know AI is also hard at work behind the scenes protecting many facilities, keeping thieves and vandals at bay? It might not be as buzzworthy as generative AI, but it’s providing breakthrough capabilities for physical security. And, notably, it’s in use and providing benefits today.
When it comes to physical security, AI isn’t just the hot new tech trend on the block. It has transformed the way security companies help organizations detect and deter theft and vandalism. By pairing remote video monitoring (RVM) solutions with AI, facilities are now able to detect suspicious activity and respond faster than ever before.
On the other end, companies monitoring security video feeds can have trained professionals alerted when AI says there is something happening on video that is worth investigating, making people more effective at their jobs. This equates to better cost efficiency, greater return on investment (ROI), and better overall business outcomes. And because AI models only become more accurate over time, these solutions are constantly evolving for the better.
The AI Education Gap
This begs the question: If AI is such a powerful tool in the fight against crime, why aren’t more facilities leveraging this technology to prevent physical security incidents from happening?
Each year, Pro-Vigil conducts a survey across a range of vertical industries to understand the physical security threats businesses are facing today and what’s being done to combat them. This year’s report, “The State of Physical Security Entering 2024”, examined perceptions of AI in security strategies and found that a vast majority (71%) of businesses said they are not using AI for physical security while 23% of respondents weren’t sure if they were using it. Additionally, 29% believe AI can be a useful tool in stopping physical security incidents, and 57% don’t know.
These findings illustrate an education gap when it comes to the power of AI in physical security strategies. Simply put, businesses, including facilities, are not yet turning to AI as a dedicated security tool. And for some, they aren’t aware that AI is powering their crime deterrence solution of choice.
AI’s Emerging Role in Physical Security
Whether you’re looking to secure an office complex, hospital, warehouse, school, university, or beyond, these facilities are all looking to achieve the same goal—keeping the bad guys out. That’s no easy task, especially for facilities that can’t easily define their perimeter using fencing or afford to have multiple security guards patrolling several areas of a single site.
In recent years, we’ve seen the rise of RVM alleviate many of those challenges. With RVM solutions, virtual guards review video surveillance feeds remotely and deploy deterrence features like recorded announcements, flashing lights, and alarms when needed. As a last resort, the proper authorities can also be notified. (It is always the best outcome for all involved if the would-be criminal simply leaves the site without causing any harm.)
With AI, this concept is taken to the next level. Rather than having a virtual guard review hours of footage daily, AI models can be trained to detect specific behavior, like a person jumping a fence or walking across a parking lot towards a building after-hours. It can also identify events that aren’t worth a human’s time to investigate, such as a dog walking by or the presence of wildlife. The AI determines which events need further attention and escalates those video clips for a closer look, filtering out an enormous volume of false positives and allowing virtual guards to focus on the events that matter most. In physical security, seconds make a difference, and filtering out those false positives can have a profound impact.
Customizing AI for Facilities
For this security strategy to work, AI models are trained with enormous amounts of data to determine what should and should not be considered suspicious activity. Video captures an enormous amount of intelligence (e.g., color, shape, size, direction) and all that information helps train the AI to become more accurate over time.
Because of this, the AI can be fine-tuned for specific facilities and individual sites. If a facility is located near a forest, the AI can be trained to tell the difference between animals like a deer walking by vs. a human approaching a restricted area. Or a person riding by on a bicycle vs. a car turning into the parking lot. While these acts may seem innocent enough on the surface, they could indicate the first signs of theft or vandalism. The earlier they are detected and alerted for further review, the better chance of stopping crime from happening in the first place.
Physical Security’s Future
Because AI has transformed physical security, facilities of all types would be wise to further educate themselves on how AI can be used to protect against theft and vandalism. Early adopters will not only experience physical security benefits, they will also experience greater ROI as AI permeates other areas of business operations leading to better business outcomes throughout organizations. AI is not just the future—it’s making a difference now.
Satish Raj is the chief technology officer at Pro-Vigil, a provider of AI-enabled RVM solutions.