Physical Security & AI Now and in the Future

Jason Wright
Director of Sales, Hakimo
Northland Controls Guest Blog
‍

Imagine if your physical security operations team knew that every alarm was genuine and could risk-rank incidents in seconds. Or, if your staff could detect tailgating on a global scale by utilizing the already installed access control platform and cameras. This capability is no longer science fiction.

Today, security systems produce extensive amounts of data, monitor countless risks, and create a flood of information that can quickly overwhelm security operations. Finding the real threats within the noise takes skill, training, and often time a large, dedicated team to respond and maintain systems at optimum performance.

Up until recently, AI has been used at its basic level and traditionally reactive in nature. But thanks to the benefits of machine learning, it has now crossed the next threshold. Designed to automate complex decision-making, AI can complete tasks like identifying a person by the color of the shirt or detecting when a person crosses an imaginary line around a perimeter. For security operations teams, this type of information can be incredibly useful so they can receive notifications of activities without having to exert the manpower to monitor every camera.

Today's AI Superpowers

Machine Learning AI in physical security has been designed to make human-level decisions. For example, it’s easy for a human to watch a video and determine if a person went through a door and which direction they were traveling, or if everyone entering the facility swiped their badge. But where AI becomes even more powerful is when it can manage these basic tasks, and more, at hundreds of locations simultaneously without break.

Physical security solutions can generate terabytes of underutilized data from badge records and surveillance systems. But translating that massive amount of data into actionable insights and finding the important data or video is the AI superpower. It can analyze trends over several months or years and identify unusual behaviors, such as denied access, changes in access patterns, or other trends that can point to an insider threat.

Future Possibilities

In the future, we can expect to see advanced AI manage data generated from platforms beyond just physical security, resulting in the convergence of many corporate systems. AI will soon be able to support faster response and greater event intelligence. This type of intelligence can enable key shifts within GSOCs, including moving event response to the edge and enabling GSOCs to handle investigations, inside threat management, and travel security support, instead of clearing alarms.

‍