Using artificial intelligence to improve site access security

AI has great potential to improve safety across the building and construction industry, by keeping unauthorised personnel off worksites. However, because many security entrances do not have AI built into their technology, integrating intelligence into secured entrances requires a collaborative effort with a third-party solutions provider, writes MICHAEL FISHER, Managing Director, Boon Edam Australia.

Many building automation systems, fire systems, intrusion detection technologies, and physical and network access control systems already have AI built into many of their core competencies.

However, despite the fact that AI will be able to help in many security-related tasks — such as discerning people from objects at a facility’s perimeter and interior entrances, detecting attempted piggybacking, spotting and analysing potentially lethal objects and dangerous people — AI does have its limits. It is not capable of taking action to prevent unauthorised human entry or deny the entry of dangerous objects. So what is the answer?

Third-party solution providers and AI integration

Integrating AI at site entrances requires collaboration with a third-party provider. Video analytics are increasingly deployed to address use cases such as people detection, piggybacking, dangerous object detection and facial recognition among other issues relevant to secured entrances.

The increased integration of AI providers with traditional security entrance partners has resulted in improvements in speed, price, ease of use and usability. Not only that, but it has improved the ability of machine learning to improve algorithms over traditional modelling and correlation approaches, and integration with other systems and sensors.

AI works to replace human efforts at the entry

So, what can AI do alone at a secured entrance? Basically, it can replace human effort at tasks that would be tough for people themselves to undertake reliably: learn behaviours of staff, employees and contractors; and identify people and monitor them more consistently than operators.

Currently, many security entrances detect tailgaters by using near-infrared sensors — if it appears that two separate objects are breaking through the sensor beams, an alarm is generated. In security revolving doors and portals, near-infrared, time of flight technology is paired with optics to create a 3D image of the objects inside the door; algorithms and sampling data are used to determine whether there is one person or two. When these technologies reject a user ‘for no reason’ (for example, a person enters a door with a pizza box and wears a backpack) that’s known as a ‘false rejection’.

Advanced AI can fill the gap by recognising people (through learned movement patterns and spacing of features) and objects the way humans can, which can bring the false rejection rate to near zero. For example, it could register the difference between any inanimate objects being worn or carried through the entrance, as opposed to living users. It can intimately know the identity of authorised users, regardless of clothing, current weight, hair colour or facial hair, and the process of aging.

AI is a double-edged sword

Machine learning and deep learning have been used for many years in the big data world to identify trends and produce metrics regarding human intent. Within the security industry, a trend has been witnessed whereby companies leverage these specific engines to gain greater benefits from access control, video surveillance management, intrusion detection and tracking systems.

Machine learning can be seen as a double-edged sword, because tracking learned behaviour can help define potential vulnerabilities and unmask potential threats. However, it can also lead to privacy and discrimination concerns, especially when intent and analytic detail are not clear.

A layered approach to security entrances

The current best practice for security entrances, particularly in high security applications, is to employ a layered approach.

A layered approach could involve a combination of:

• Full height turnstiles, which are useful at the outer perimeter, because they provide both a visual and physical deterrent against unauthorised access.
• Speedlanes to prevent tailgating — unauthorised people tailing authorised personnel through security gates — through to the use of alerts and visual recognition features that alert security staff to a potential breach.
• High security portals, which use biometric scanning and overhead sensors to ensure the credentials of each user. It guarantees each user is alone and is exactly who they say they are. This is the ultimate security frontline.
   

AI can provide enhancements to much of this existing technology. For example, facial recognition software can be integrated into speedlanes, so that instead of each person needing an access card, they can smoothly walk through and be authorised using their face.

The future of AI and security entrances

AI in the form of machine learning and deep learning has become a disruptive technological force in the physical security industry. The push for touchless and frictionless access options will expand the integration of secure entrances with building control systems to help provide additional insight to potential threats and help mitigate them.

Image credit: iStock.com/Marco VDM