Editor, IFSEC Global

Author Bio ▼

James Moore is the Editor of IFSEC Global, the leading resource for security and fire news in the industry. James was previously Editor of Professional Heating & Plumbing Installer magazine.
February 18, 2020

Sign up to free email newsletters

Download

The State of Physical Access Control in EMEA Businesses – 2020 Report

The role of AI in physical security

Kevin Waterhouse, Managing Director at VCA Technology, details how AI can support companies and staff in the physical security sphere.

It’s clear by now that the excitement around artificial intelligence (AI) is accompanied by a good deal of confusion as to what this technology can actually achieve. In the security sector, this confusion is aggravated by the concept of facial recognition with which AI is often associated. In truth, there is a specific area of AI, namely machine learning, which can make a huge difference in terms of making environments more secure and, in particular, supporting dedicated staff in this effort.

For companies with complex protection requirements, a significant chunk of their security operations is represented by the human workforce. However, the high demand for monitoring room staff as well as guards, coupled with the labour shortage currently plaguing the UK, is causing these companies’ security resources to be dramatically overstretched. That’s what makes technology so indispensable.

Let’s take a look at the role AI can – and can’t – play in security, and how it can empower staff and businesses to better protect their premises.

Barriers to action for AI

What’s interesting about AI (and its perception among businesses and individuals) is the discrepancy between what the technology has the potential to achieve in theory and what companies can use it for in reality. For starters, some of the most ambitious uses of AI are only attainable with enormous amount of processing power, and extremely costly hardware – spending such exorbitant amounts to make these aspirations a reality would simply be impractical.

AI-PhysicalSecurity-20

Furthermore, the use of AI to monitor others’ activity is interlinked with complex ethical implications. The recent announcement that the Metropolitan Police would be rolling out facial recognition across London was met by outrage and distrust from many, while we are all familiar with how this technology is said to be used in China.

When you eliminate what’s practically impossible and what’s unethical – what uses of AI in security are you left with? Ultimately embracing AI in this field isn’t really about doing things we’ve never done before, it’s about doing what we already do – but better.

What can be done

This coming year will see businesses investigating and unlocking the real benefits of AI within security. Some of these are related to how machine learning, in particular, can enhance video surveillance. While a simple CCTV system that merely records events is just about as useful as a broken lock, analytics-based video surveillance can hugely improve the protection of a business environment. It offers invaluable support to security workers by flagging suspicious events, making these infinitely easier to manage. However, if there is one challenge that threatens to make analytics-powered surveillance alerts almost pointless, it is the occurrence of false alarms which simply waste monitoring staff’s time and disrupt business activities unnecessarily.

So, in order for a security system to be “trained” to recognise potential threats, an element of machine learning is required. This enables the pre-calibration of the surveillance system which can then accurately distinguish between humans and vehicles (which can represent a danger to the premises), from wildlife elements and other innocuous objects. The machine learning technology provides video surveillance solutions with added accuracy, which can make a huge difference to workers’ day-to-day activities. In a situation where the reduced workforce can’t afford to spend time on unimportant events, the value of machine learning as a tool that helps filter out unnecessary alerts is undeniable. For security teams, this represents an opportunity to optimise resources and safeguard their business more effectively by reacting to real incidents proactively rather than retrospectively.

Human interpretation of AI

Back when the hype around intelligent machines first started, we would marvel at how a computer was able to beat a chess champion at their own game. Chess being a matter of pure logic and strategy, a computer was perfectly capable of identifying all the right moves that would lead it to victory. It’s a mathematically calculable game that leaves nothing to interpretation. This superiority of machine over human, however, doesn’t translate to the world of security.

It’s understandable that some companies wish to entirely remove the human component from security processes – how wonderfully cost-effective would it be for businesses to eliminate their workforce altogether, and rely solely on artificial intelligence to identify and respond to threats?

However, while machine learning can help increase accuracy of detection and greatly reduce false alarms, systems are bound to flag oddities from time to time. That’s why a worker’s input is always required. Afterall, there’s always a human doing the initial “teaching” – setting the rules, calibrating the system. Of course, modern and sophisticated technology makes security staff’s life easier – promptly directing their attention to suspicious events in real time – but a human will always be involved in the process; interpreting the alarm and using their intuition to determine the appropriate reaction.

Think about it. As appealing as it may seem, eliminating human decision-making from physical security operations has the potential to cause catastrophic incidents, for individuals, businesses and the growth of AI itself. Things may still go wrong – if innocent people were hurt in the process, we’d only have technology left to blame, and this would cast a dark cloud over the development and use of artificial intelligence in security.

As new technologies emerge and the business landscape grows increasingly digital, companies have the privilege of choosing the most appropriate tools to achieve their aims, but are also tasked with distinguishing industry hype from true value-adds. Machine learning has the potential to streamline and enhance the analysis of data, including that collected by video surveillance systems, thus empowering businesses to take threat detection to the next level, and therefore represent some of the most useful instruments in a security manager’s arsenal.

Download the Intruder Alarm Report 2020

Download this report, produced in conjunction with Texecom, to discover how increasing processing power, accelerating broadband speeds, cloud-managed solutions and the internet of things and transforming the intruder alarm market, and whether firms are adopting these innovative new technologies.

AlarmReport-Main-19

Related Topics

Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments