Using Video Analytics For Anti-Tailgating At Access Controlled Doors/Portals

I need to find an alternate to using full height turnstiles for 100% access control.

My client needs to verify that all persons entering a door/portal has badged in to the door's access control reader.

My idea was to have an access control reader door followed by a short vestibule and a second access control door (similar to a mantrap/interlock, but with free egress out). Each person must present their access badge to the reader. The analytics would count the number of people that entered the first door, and compare that to the number of authorized access badge reads.

Example - If 1 person badges in, and 2 people enter through the door, the 2nd set of doors would remain locked unless the 2nd person goes back and presents their valid badge to the reader (the number of badge reads have to match the number of people that enter through the door before the second door will unlock.

We have tried IR and Photoelectric beams, but these have been troublesome, particularly if someone is carrying something large or pushing a trolley or are in a wheel chair.

My client is using Lenel OnGuard and Axis IP cameras. They are currently using Lenel VMS, but are searching for a new platform.

Your suggestions and SME knowledge would be appreciated.

I did something similar using a programmable relay. using the ACS door replay output and having the camera momentarily pulse a built-in relay upon activation of a tripwire analytic, in this case you might need people counting. The timing and throughput is where it gets tricky. What happens when someone is escorting a guest?

A different option may be to use half height optical turnstiles.

Hello Jeff:

What aspect of using full-height turnstiles do you need to find an alternative for? Cost? Wheelchair access? Aesthetics? Constructability?

What you describe (mantrap/free egress) can be done, but it could run into some of the problems you're having with turnstiles.

Also, from reading your post, it sound like you are interested in 'positive stop' options that prevent tailgating, not just sounding a response alarm if/when tailgating occurs.


Glad to see you are on IPVM.

Have you looked at the Newton Security Mantrap Product?

These use machine vision cameras and are commonly used at mantraps in data centers. My clients who use them seem to like them.

The reason I am looking for alternative options to turnstiles are costs and space restrictions.

I can see where there can be potential issue with free egress while someone is entering. Hadn't considered that.

I am looking for "positive stop" options to prevent tailgating. Currently they have full height turnstiles and security guards at secure portals, but this isn't cost effective or practical for most locations.

The client's policy is 100% access control into controlled spaces. Currently they are using Boon Edam Tourlock turnstiles or barrier turnstiles with security guards to verify one badge read per entry.

I am looking for more cost effective and practical means of accomplishing the same end results as turnstiles are not always practical. I was hoping someone out there may have found a solution using camera analytics interfaced with card reader access control. I am open to any practical and reliable suggestions. Thanks for your feed back so far.

Hello Jeff.

First, I assume 'tailgating' means 'letting more than one person through an open door'. The fundamental issue to weigh lays somewhere between 'tailgating detection' and 'tailgating prevention'.

You indicated that 'prevention' is the ultimate goal, but unfortunately sensors and analytics do not prevent anything. Best case, they may squawk and alarm if tailgating is detected, but the incident still requires human intervention to address.

"Positive Stops" include the methods you are already employing - guards, turnstiles, revolving doors, mantraps - but those solutions are not inexpensive nor small.


Thank you for the suggestions of Newton Security Mantrap Solutions. I looked at there website and the current advertised products don't meet the criteria, but the are not far off. I will reach out to Newton to see if they have any suggestions or products that more closely match what we are looking for. Thanks