IPVMU Certified | 06/23/15 11:56pm
It's a good question. In my experience, anti-tailgating effectiveness looks like this in terms of strength:
- Turnstiles/revolving doors/mantraps
- Optical turnstiles/beam type frame mounted sensors
- VA detection
I do not think analyic tailgating detection is especially difficult for a well lit, interior single door, movement all the same direction, with pedestrian traffic. When VA gets tough is when the camera covers multiple doors, outside, mixed lighting/environment, with mixed traffic (delivery people or carts).
It may look ugly, but dropping a turnstile and funneling traffic to a single point seems most cost and operationally effective.
I am curious to follow feedback on this one too.
IPVMU Certified | 06/24/15 02:28am
Also, on this question:
Or are the Video Analytics operating largely 'standalone' from the Access Control System?
I am not aware of any integrated or native access control based anti-tailgating video analytics. The analytics need to be incorporated into the platform through alarm configuration or similar.
In the example I am familiar with, OnSSI used AgentiVI. This was a totally seperate analytics server.
A few years ago I spent 3 months immersed in the anti-tailgating industry, on the manufacturing side. I was surprised by how easy it is to defeat many of the systems that are currently sold as reliable anti-tailgating solutions, whether they are a physical man-trap, turnstile, photo beams or video analytics. Assuming these are automated systems without a human observer, some of the tactics that defeat these systems include 2 people hugging each other (1 in front of the other) as they pass through the device, crawling on the floor, one person pushing a cart that another person is riding on or in, one person wearing a very large loose fitting garmet the covers another person, etc. Mantrap doors or other access controlled doors may be tricked by putting a magnet on the contact after the door opens, the door can be left propped open, yet the system sees the door as closed. In the case of data centers/server farms around the world, which all the major players have, the level of anti-tailgating security measures to prevent unauthorized access can be ridiculously low. I am very skeptical of anti-tailgating systems as a 100% reliable method of stopping tailgating. And human observers can also be tricked, confused, asleep or bribed. Considering how important it can be to prevent unauthorized access to what should be a secure facility, the anti-tailgating industry has a long ways to go.
I might take Jeff's comment as a little skeptical if I didn't have the same experience! I would add that there were effective measures with all technologies and the best start being physical in design. Just not 100%, but then what is 100% and reasonable in cost?
If you don't mention tailgating does it really exist ;)
What about car tailgaiting? Anyone have experiences with that?