Anti-Tailgating Startup: Spyfloor

Avatar
Brian Rhodes
•Published Oct 03, 2018 12:50 PM

A Canadian startup, Spyfloor, is using a different approach to warn against tailgating, a common access control problem. By counting feet, Spyfloor's floor mat tracks how many people are trying to pass through a door.

The developer says their sensor approach is simpler to implement and more accurate than typical tailgating detectors, and offers a high-performing solution to one of access control's toughest problems. Inside this note, we examine the offering and contrast it to other tailgating detection options.

Tracks ***** ** ****

*** ****************** *** ******** ********* ** ****** and ****** **** ****-********** *********: ** counts ****.

*** ****** ****** ******* *********, *** pair ** **** **** ** ** and ******* ** ******* *** **** read ** *** *****. ** ******** pairs **** ******* *** ******* ** close ********* ** ******* ******** *********** being *******, ** ** ****** **** 'Tailgating', ** ************ *****, *** ***** place.

******** **** * *****-********* *** ** detect ***** ** ****, *** ****** if **** **** *** **** ** footprints *** ******** ** *** ******:

******** **** ***** ** ****** * characteristic ***** ** **********, *** *** company ****** **** '*** ***** ** shoes *** ********', **** ***** ** boots, *** *** ****** **** *** false ***** ** ***** **** ******* suitcases, *****, ** ***-******* ******* **** or **********.

Spyfloor ******** *****

*** *******'* ***** **** ***** ***** provides * ****** ** ******* *** cases *** ***** *** ****** ** use:

** ****, **** *** ****** *** direction ******** ** ********** *** *******, and *** ******* ****** ******** ********** of ***** **** **** *** ********** are ******* *** **** *** ********* of **** *** *******.

Layered ***** ***

*** *****-********* ****** ** *********** ** glass *** ******* * ******* ******* board **** ** ** */*" *****.

*** ****** *** **** ** *** of *** ******** ***** ** ** installed ** ******** ***** **** *** floor. ** **** *** ****, ******** *** * ****** ~* square **** **** ****** ** ***** booth:

*** ******* *** ******** ** ******* to * ************** **** **** **** reads ********** *** ******* ******* ** not ** ********* **** ********** ** incorrect ******* ***** **** ***** ***** and **** ***** ** ***** ****** to *** ****** ******* ********** ******.

Custom *********** ******

******** ******* *** **********, *** ** order ** ** **** ** ****-********** for ****** *******, ********** *********** **** is ******.  *** ******* **** **** any *********** ** ******** *********** **** can ** ******* ** ***** *** staff.

** '***-***-*****' ****** ******* ************ *** available, *** *** ******* ***** ** that ** *** ******* *******, ***** are ****** ** ** ********* *** offered.

Spyfloor *******

*** ******* ** ***** *** ******* are ~$*** *** ****** ****, ** for * ******* *' * *' mat *** ** ***** ** * door, *** * ****** **** **** run ~$*,***. ****** *** ****** *********** are ********** *** **** *** ******** based ** ******** *********** ********.

*** ******* ********* **** ******* *** decrease ** ***** ********** ******* **** up.

*********

*** **** ****** *********** ** ****** control ** ******, ****** **** **** and **** ** *** ******** ** not *** **** ********* *******. ***** issues:

  • ****** ****: *********, ********'* ******** ** *** rated *** ******* *** *** ** not ***** ********* ** ****** ******* environments **** *** *** *** *************** controlled. ******** ***** **** **** ** outdoor-rated ******* ** ****** *** ******* next ****.
  • **** ******* ***********: ** ********, *** ******* **** that ******* ***** **** *** ***** will *** ****** ****** ***, *** the ****** ** ******** ****** ** track ***** ******* ****** *******. *******, the ******* **** *** ****** ******* must ** **** ***** ** ***** mud, ****, ** ***** *** ** cleaning ** * ******* *****, ** often ** **** *** *** *** typical *****.
  • ********* ********** & ******** **** ********: ******** **** ********* **** *********** and ******* ********** *** ***** *********** problems. ***** *** ****** ********** *** rolling ****** *** **** *** ***** damage, *** ******* *********** **** **** of ******* ******** ** *** *** gates ** ********* *** ** *** irregular ***** ********** ********** **** *******.

Sensor *** *********

**** ******** *************: **** *** **** *** ******, the ****** *** ** ****** ** most ***** *** ******** ******* ********* building ************ ** ** ******* **** common ****-********** ********* **** **********, ********, *** ********* *****. *** */*" thickness **** *** ******* ******* ********** or ******** * ******* ** ****** architectural ***** **************.

*****-** *********: *** ***** ******** ******* ** a ********** *********, ********** ******** ** Time-of-Flight ******* **** **** ** ******* mounted ** * ******** ****** *** whose *****-**-**** *** ** *********** ********** by ********** ******** ** *** *****.

Less ********* **** ***** ********* 

******** ** ***** '****-**********' ********* **** ***** ********* ** ****-**-****** *******, ******** **** ** **** ****** for * ******* ******-**** ******** ****. A ****** *** **** ***** ** cost **** $*,*** - $*,*** ***** on ****, ***** * ****** ***** sensor ****** **** $*,*** - $*,*** based ** ******** ******. 

***** ********* **** ******* ********** ***** ***** in ***** ~$*,*** - $*,*** *** are ********* ****** ** ********, *** will ******* ******** ************ ** * doorway *** *********** *************** ****** ************* ******** **** ***.

*******

*** ******* *** ********* *** **** need ** ******* ** ************ *** pricing ** ****** ***** ********* ******.

Vote / ****

Comments (9)
Avatar
Skip Cusack
•Oct 03, 2018

Tailgating is the industry's biggest PACS problem, but that doesn't equate to a huge market opportunity. It seems like this abuse is just accepted, and many believe it can be mitigated through training. Which only serves to stop the honest abusers, not the people you really want to catch.

I'm glad to see new approaches to this problem. And this one may work well, but I would guess that the cost of an additional stand-alone technology will limit its adoption.

When anti-tailing comes for almost free, like part of the reader itself, then I think we'll see everyone claim how vital it is. As long as it costs extra money, it's a tough proposition.

(3)
(2)
Avatar
Iwan Debets
•Oct 04, 2018
ASB Security
In reply to Skip Cusack

Isn’t a 3D counting camera the better and cheaper solution to tailgating? I’ve successfully built a custom solution using a tof camera and some intelligent analytics. I do appreciate this innovation, but it feels like I’m missing something here?

(2)
Avatar
Dori Ribak
•Oct 04, 2018
RBtec Perimeter Security Systems

Seems like a great solution that will fail miserably, there are cheaper sensors that can do the exact same thing.

As mentioned above, a camera with AI analytics can easily count people from 10ft away.

(2)
(2)
JH
John Honovich
•Oct 05, 2018
IPVM
In reply to Dori Ribak

a camera with AI analytics can easily count people from 10ft away.

Accurately? 10 feet away? No.

I am not endorsing this floor option but dismissing it with your counter is silly. There is a reason that minimally high accuracy people counting analytics are mounted directly above because otherwise obscuring is a significant problem. You can't count (with video analytics) what you can't see.

(2)
(1)
Avatar
Dori Ribak
•Oct 05, 2018
RBtec Perimeter Security Systems
In reply to John Honovich

All of those systems are "pre-yolo" and all of those algorithms that have no problem counting people every when standing near each other especially if the camera is positioned properly. 

Avatar
Skip Cusack
•Oct 05, 2018
In reply to Dori Ribak

I shouldn't think you need to look out into a busy hallway and detect, track and count people. You only need to have a view over the portal's threshold and then counting blobs with enough resolution to accurately classify a blob as a person and not a filing cabinet on an upright dolly.

JH
John Honovich
•Oct 05, 2018
IPVM
In reply to Dori Ribak

especially if the camera is positioned properly

You said '10ft away'. It's simple physics / logistics that it is hard to properly position cameras that are 10ft away from an entrance. Unless the ceiling is 100' high, which it won't be, you will have obstructions due to the angle from the camera position to the people entering 10ft away.

Avatar
Skip Cusack
•Oct 05, 2018

A 3D camera with AI analytics is a fine technical solution. But I wouldn't expect it to be less expensive than a sensor mat or some other low tech solution. 

A solution that is packaged as a separate device from what is already installed at the door, namely the reader, will just be too expensive to take off. As readers continue to become commoditized, as surveillance cameras did before them, manufacturers will look for ways to add technology to sustain the price point. Many cameras now have embedded analytics from the factory. I predict readers will follow this trend.

A savvy reader manufacturer would do well to integrate an anti-tailgating option into a reader for an incremental additional cost. It makes good business sense, but also makes good technical sense as the reader already has comms to the controller, or even the door directly in edge applications.

 

U
Undisclosed #1
•Dec 05, 2018
IPVMU Certified

It might stop tailgating, but it won’t stop piggybacking...

(2)
(2)