Workplace security with AI
Tailgating continues to be the largest physical security risk affecting workplace access, this type of social engineering can affect anyone from the CEO to the outsourced cleaning firm.
What is tailgating?
Tailgating is when someone accesses a building or restricted area with their access card, key, bio-metric identifier opening the door or barrier and someone follows in behind therefore not requiring them to identify themselves and thus gaining unauthorised access.
People tailgate all the time, and most of the time it is innocently and very British to hold the door for someone. Adding pod entry systems isn’t a cost-effective answer, but what if an existing infrastructure could be used to highlight this issue.
Your workplace has a CCTV system that may not be high quality enough to identify the person entering the building or area is capable of identifying if more than one person is entering a door when only a single access card is scanned.
By analysing the CCTV footage around these entry doors checking for people tailgating, correlating this act with the swiping trigger of an access card you can see if this is an area of security to enhance in your business. This process can also be done passively without the need of affecting employees experience.
How does it work
Your CCTV system is already recording entry doors within your workplace, therefore, providing a history of training data to generate a data model. The video data is manually tagged providing a source of video that is and is not tailgating.
Supplying a new video stream with the trained data model can provide an identification of tailgating or not with an accuracy score, events can be manually inspected and validated (Were three people entering the door) providing constant training of the model to increase accuracy over time.
This vision detection trigger of a tailgating happening is then linked to the swipe of an access card, providing the business with the data to act and with discretion notifying the access card holder they have been tailgated. This information can be collated to the business, providing an understanding of the level of risk tailgating is exposing.
What else could you detect
Imagine Jim leaves for lunch swiping out of the building; his car licence plate is detected leaving the car park, but then his swipe is detected back into the building without his car return. Jim may have parked elsewhere, but Jim could have also dropped his card giving someone unauthorised access to the building.
Having this data available enables your business to decide how to act with discretion, but it provides much greater insight and control to what is happening.