It’s no secret that artificial intelligence (AI) is exerting its influence on society in profound ways. AI can be found pretty much everywhere – in applications ranging from self-driving cars to online assistants to game-playing computers to the prediction of judicial decisions. It’s solving problems, moving markets and changing lives.
AI is also weaving its way into the realm of physical security. While its uses are still evolving, AI is positioned to play a key role in the new-world security paradigm, where terrorist attacks and mass shootings have broadened the threat landscape and made it more unpredictable. Security responses are shifting from reactive to proactive, and vendors are integrating AI technologies into some of today’s security solutions. Here are three AI applications that can help organizations take a proactive security approach to prepare for future threats.
Determining what’s “normal”
Before you can declare that a situation is abnormal and worth attention, you need to be able to define “normal.” Using machine learning, an advanced form of AI, computers can be taught how to identify an object once certain characteristics are specified. Once the computer learns what a normal environment is, it can monitor for anomalies and alert security personnel when it identifies something out of the ordinary.
More specifically, computers can be taught what is allowed in a certain area at a certain time. For example, a system can be taught that figures moving around outside a public building are normal during the day but abnormal at night. When an environment changes from normal to abnormal, an alert can be automatically triggered.
Using AI, computers can do more of the work of monitoring environments, giving guards and operations personnel more bandwidth to focus on higher priority tasks, such as quickly responding to an actual threat. Automating monitoring of these environments also reduces human error.
So, now that AI is helping to define what’s “normal,” it can go to work determining what constitutes a “threat.” Machine learning can be taught to identify an object as something specific based on certain characteristics. This is referred to as “object recognition.” It can be used to identify outliers to the norm – which, in a security context, can be defined as threats.
For instance, in the situation above, security personnel want to be alerted at night if a specific type of figure enters the scene. People obviously qualify as potential threats. Perhaps some large animals or vehicles would be worthy of an alert. Small animals? Birds? Windblown trash? These would be picked up by some sensing systems, but they’re not actual threats. If a computer can be taught to recognize certain objects by their size, shape or specific actions, it can flag threats and filter out benign activity.
With the right sensors, when surveying a large crowd of people, guards can determine if a visitor’s bag might contain a threat object and then track that visitor or object. Should that person appear back on the screen without the bag, the computer can search the environment for the item, quickly sending security guards to that specific area and clearing the crowd.
In a security checkpoint scenario, this application drastically reduces the need for hand wanding or physical full-body pat downs as the technology itself can alert guards if someone is carrying an item of interest. Guards can then focus on a subset of people as opposed to screening thousands of visitors or travelers with the same level of rigor – which would improves the visitor experience for all involved.
Tracking a figure based on characteristics is advanced; re-identifying the figure without using facial recognition takes security to a whole new level. This technique, referred to as “object re-identification,” is used to track an individual through multiple fields of view from different cameras. With law enforcement teams often working from blurred and obscured images when identifying a suspect, this emerging capability is extremely promising despite its current limitations.
Being able to track someone or something by its shape, clothing or gait could best be described as a game-changer for security operators. Using AI, law enforcement could, in theory, track a known threat and detain a suspect before he commits an attack. However, performance improvements in faceless detection are still needed as false match rates continue to run high.
AI + IQ
While AI’s adoption is helping organizations and security professionals proactively prepare for future attacks, it can’t do this alone. It has the power to analyze data quickly and identify patterns, but it can’t necessarily determine if these patterns are relevant. AI needs help from the kind of real, grounded intelligence that only humans can provide. Using the combined power of AI and human intelligence, security teams can truly arm themselves to fend off emerging threats.
Read more here about how AI will impact the airport.