Sensing Danger in 2020

AI-powered detection, prevention and privacy

• By Clayton Brown
• December 01, 2019

When bullet-proof backpacks are sold out at your local Wal-Mart, you know something must change. Yet many school leaders are left wondering how to begin taking steps to properly ensure greater school safety. While politicians debate gun policy in Washington, innovative school leaders can be proactive and effective today using AI-powered physical security technologies. AI-assisted security can have a real impact on student life safety through better prevention, detection and management – all while balancing privacy concerns.

Before we dive into how AI can help improve safety on our campuses, let’s examine how it works. Essentially, AI simulates the human intelligence process in that it acquires information and identifies patterns to then appropriately respond. One of the core components of this intelligence is an artificial neural network (ANN). Similar to the human brain’s nervous system, an ANN works with countless, various types of sensors.

AI and Situational Awareness

Human sensors are eyes, ears, nose, skin and tongue. They identify features of the environment – people, things, temperature, light or darkness, moisture, sounds and many other things. As those sensors activate, humans respond. Internal variables adjust, patterns are identified and reactions are formulated. For example, the smell of smoke indicates fire, and the brain sends signals to the body to take appropriate action to escape danger.

A more sophisticated example would be how human behavior and, ultimately, a person’s character is assessed. Over time, personality traits are identified and changes in behavior are observed. All of this information leads to the development of a trust or no-trust relationship.

When it comes to physical security technology today, the artificial neural network sensors can be surveillance cameras, sensors, access control systems, IoT devices, big data, social media plug-ins and human input. As those sensors feed internal variables, the ANN learns and identifies patterns of risk, providing real-time situational awareness to human personnel.

AI may seem like magic, but it is real and it’s available today within innovative security solutions. It can evaluate large amounts of data to identify threats that might otherwise go unnoticed. This fundamentally changes the game in physical security and will affect all traditional components of a layered solution.

When AI is integrated into physical security infrastructures, it can help to address the number one threat concerning parents and school districts - and likely number two and three as well: active shooters. This phenomenon is now so prevalent that it can be considered an epidemic. Last year was the deadliest on record for school shootings – at least 83 died or were injured in active-shooter incidents. Since Columbine, at least 228,000 students have been exposed to gun violence at school, according to a study by The Washington Post.

Thankfully, school violence is receiving more of the attention it deserves from innovators and policymakers alike. An estimated 67 percent of districts are currently conducting active shooter exercises; others are designing schools that are built with safety in mind. Recently, lawmakers also proposed a bill to establish a national database as a resource for districts to assess, identify, and share information on school security technology and innovation.

Let’s take a quick look at some of the ways AI can enhance the security technologies and efforts that every school district, parent and teacher can begin advocating for and implementing going into 2020.

AI-Powered Weapon Detection

High security is the objective, but there is a dual-mandate to make sure it doesn’t feel like high security. New AI-powered software and sensor systems are making this dual-mandate possible in a discreet way.

Previous generations of school security have relied on metal detectors, forcing students to queue up like they are entering a courtroom. New AI-powered weapon detection is changing this with discrete and passive detection methods, such as:

• AI-powered microwave systems that can detect concealed weapons such as guns or knives
• AI-powered video analytics that can detect handguns, rifles, objects left behind or a fight breaking out
• AI-powered chemical detection that can detect gunpowder, explosives and narcotics
• AI-powered gunshot detection that can determine the location of a gunshot and the type of firearm used

AI-Powered Security Management

While improved detection systems are important, AI-powered security management is what transforms these analytics into action. When any or all of these detection systems are used in conjunction with AI-powered security platforms, mustering is improved by helping to:

• Notify police instantly and stream surveillance to first responders for improved mitigation
• Communicate with students and parents to share the safety status of the student and risk status of the threat
• Prescribe appropriate procedures and remediation actions to resource officers and security operators, based on the location, the criticality and the stage of the situation underway
• Automatically adjust access permissions and initiate emergency lockdown procedures for certain zones based on the specific location of the incident

AI-Powered Privacy

There are new technologies that carry certain elements of controversy, particularly regarding privacy. Facial recognition has been in the news, which we will discuss, but I will also introduce a concept farther on the cutting-edge - monitoring student behavior.

At the core of both technologies is the privacy of a student’s identity - and this isn’t trivial. The good news is that new AI-powered security platforms allow school districts to have their cake and eat it too - meaning privacy is the norm - but when safety is at risk, safety is the highest priority.

Video analytics and facial recognition are popular ways to learn more about any environment, but student privacy is also key. AIpowered security management can configure video analytics to only identify “unknown” faces, while allowing known faces to go detected, but without reporting that a student’s specific identity was detected.

When these platforms detect such a situation from video analytics or weapon detection systems, “safety first” mode can be activated, allowing the facial recognition to start identifying the student faces to better know their location and status, without relying on mobile apps or muster procedures - both of which are easily forgotten in an emergency.

This improves situational awareness and drastically reduces the chances and potential harm by an active shooter event, while still keeping student privacy central to everyday school operations. This is called “conditional facial recognition” or “safety-first privacy.”

AI-Powered Perimeter Defense

With the advent of AI-based approaches, perimeter defense solutions are the initial line of defense of a layered security system and help to detect a threat from a distance. Specialized sensors can be used in conjunction with other elements such as cameras and monitored with AI-based security software to identify threats just outside the campus walls.

For example, license plate recognition can be integrated to detect, “vehicles of interest” or “persons of interest,” and AI-assisted drones can be used to help detect threats from above before they become an issue, or follow the threat in case an incident occurs.

AI-Powered Entrance Control

While turnstiles may not have changed dramatically in form, their use within a layered security solution can provide invaluable data about traffic patterns to an AI-based solution , as well as become the first barrier in an active shooter scenario.

Randomized spot checks, controlled throughput and directed traffic flows can be initiated and managed via intelligent, risk-based solutions as threats are detected.

AI-Powered Prevention

The ultimate goal is to stop an individual before they bring a weapon to school in the first place or to stop a student on their path to harm, before any force is needed to keep a community safe.

To continue pushing the envelope, there are systems that use community observations to better monitor dangerous behaviors. The FBI has concluded that there are 17 symptoms of a student’s path to becoming a potential active shooter.

Many times, these behaviors are noticed by various members of a community, but are not communicated to an appropriate central authority. This is most likely because there has not been a technology capable of tracking this until now. However, it is also likely to stay this way, as parents want to ensure their student is not falsely accused or permanently labeled by any human or technology. Predicting behavior before it happens may seem a little like the “Precognitives” in The Minority Report, and simply too far for what our society is ready to handle today.

This is where an important conversation is required by all community stakeholders. Privacy and security are two-sides of the same informational coin, so to speak. As with many concerns and questions, our society is looking for a mutually exclusive answer, rather than a balanced hybrid solution. AI-powered security management allows for such a balanced solution that keeps privacy in mind while keeping safety a top priority.

There is no right answer, or one-size-fits-all policy for such a multi-faceted issue. But the good news is that AI-powered security technology is here to help you build a proactive security posture that protects your students’ identity as well as life safety.

This article originally appeared in the November/December 2019 issue of Campus Security Today.