AI in ESS – How Artificial Intelligence is Enhancing Physical Security

For the past couple of years, artificial intelligence (AI) has been a major disruptor in many parts of the economy. The built environment is no exception. As it relates to electronic safety and security (ESS), there are numerous examples and ideas of how AI can and will modify the norm of planning, designing, and implementing ESS throughout most vertical industries. To gain an understanding of AI, it is important to examine real examples of how enterprises are using algorithms to assist in keeping the world safer, while limiting the resources needed when incorporating best security practices. Artificial intelligence is not going to replace the need for human intelligence to observe and take specific action for life safety and security. However, it will play a major role in mining big data to provide the best possible solutions.

Many professionals like the idea of AI helping them to streamline their processes and procedures but are unaware of what that means in real day to day applications. There are a lot of misconceptions of what AI can do (e.g., AI works like the human brain or AI can be 100 percent objective). Thus, there is a need for ICT designers, installers, and consultants to help guide end-user clients about currently available technology and what may be available in the future. Capital investment is being scrutinized, specifically on the accountability front. However, enterprises are not going to sacrifice physical security because it is of the utmost importance to their viability. Knowing this, ICT professionals have a tremendous opportunity to bring their knowledge base to the table and to assist with the plan, design, and installation of AI in ESS systems.


Conceptual Planning

At the earliest possible onset, it is best to begin evaluating what policies and procedures may already be required for a new project. These could come in the form of a standard (e.g., United Facilities Criteria (UFC) 4-021-02, Electronic Security Systems), campus security guidelines or corporate policies. If none of these apply, an ICT Professional should use industry standards as a starting point and continue from there. BICSI’s Telecommunications Distribution Methods Manual (TDMM), 14th edition, covers ESS extensively.

Do not put the cart before the horse. It is important to look at the project holistically and not get caught in the weeds. The specifics of which doors will need card readers is not important at this stage. Some professionals may consider early budgeting numbers, but typically it is not necessary to work with granular levels of detail this early in the project. Understanding the site, overall building program, physical barriers that may exist or be needed, landscaping, and the end user’s requirements are important in conceptual planning. Ultimately, this information helps to plan further and budget accurately the ESS systems that will be needed on the project.

Electronic Access Control

Inevitably, when someone thinks of electronic access control (EAC), the first thing that comes to mind is a card reader. Card readers have been around for decades, granting access to individuals with proper credentials. They are typically stored in a database either onsite or in the cloud depending on the preferences of the enterprise.

How can AI assist in EAC? Consider, for instance, that an employee needs to access a building but does not have the fob or credential card. With the ability of newer AI integrated IP-based CCTV cameras to provide facial recognition capabilities, the employee can be admitted with only the face being processed. Two-factor authentication can be incorporated with a pin pad or mobile application if necessary.

Wearables are also becoming more commonplace for EAC. Biometrics is used to determine whether some- one should have access to a space. Biometric, AI, and wearable technologies can allow someone not only access to a facility, but they also provide the ability to skip manual login procedures for workstations, disarm intrusion detection systems, and manipulate the controls of the building automation system (BAS).

Furthermore, AI can assist in determining access to different spaces and who should be where and when. The ongoing problem of understanding proper conference room etiquette is a prime example. Conference rooms now can be booked online via several different platforms (e.g., Office 365, Exchange, G Suite). However, the problem still exists of individuals incorrectly occupying spaces. Artificial intelligence can take this to the next level with automatic shutdown of non-life safety systems, including audio and visual indicators, if incorrect or unauthorized individuals are occupying the space. A little extreme, some may say, but the technology can be used to detect dangerous intruders as well.

Panic Duress and Location Systems

Panic buttons have long been incorporated into buildings to alert safety and security personnel that assistance is needed. These systems are somewhat “static” because the button is hard coded to a specific room. If someone initially presses the button and has to escape to another room for safety, the new location would not be evident. Individuals that work in hospitality and healthcare have concerns about personal safety while on the job, such as dealing with difficult patients, guests, and people who are not supposed to be in the facility. Radio-frequency identification (RFID) and the sophisticated algorithms of triangulation can help. While RFID is not new to the built environment, the use of panic buttons alongside this technology allows safety and security personnel to respond in a precise fashion to exact locations. Cities have passed ordinances in hospitality environments that require this type of technology to protect staff, which is a new technological challenge for many facilities.

As an ICT professional, this type of technology presents many new opportunities, not only as a mandated requirement in some localities, but also in the ability to interpret these requirements for clients and incorporate them as needed to satisfy any required laws or ordinances.

Closed Circuit Television Systems (CCTV)

Perhaps the most notable use of AI in ESS at this point is in the world of CCTV. Over the past decade, there has been an increase of relatively affordable high-resolution cameras, as well as their growing adoption by many verticals (e.g., health care, hospitality, retail, detention centers). This has led to further innovation in the use of AI and machine learning to mine billions of images for patterns, features, and likeness. These systems have the ability to predict certain events with some statistical certainty in order to help prevent security breaches and criminal events.

The use of drone technology with camera integration is a perfect complement to AI in assisting safety and security personnel with keeping an area or campus secure. For large venues with thousands of people, drones can survey large numbers of individuals and general patterns for suspicious activity or the identification of wanted individuals. Drones are also a perfect complement to a traditional IP CCTV camera system. CCTV cameras are static and permanently mounted in place. A CCTV camera, for example, may alert a bank of drones that attention is needed in sector A or another specific location. A drone deploys immediately to cover that specific area for responders, capturing video information that may be needed later to analyze the security situation and identify the person(s) involved.