The $1,000 Iris Reader (SRI Tablet) Examined

Avatar
Brian Rhodes
•Published Apr 19, 2016 12:21 PM

Is this the breakthrough needed for iris?

The accuracy and resistance to change from age and wear are common claimed advantages of iris over other biometrics.

However, using irises as an access credential has been expensive typically costing $3,000+ per reader. Now, SRI's new IOM Access tablet aims to disrupt the market, costing just $1,000.

However, aside from low price, what other considerations like ADA compliance, access control system integration, and mounting location have an impact on unit use? We take a look at those factors, and how SRI's price compares to exiting iris readers inside.

Iris ********** *** *************

**** ******** ** * ********* *** some ********** **** ***** *******. *** biggest *** ** **** ****** ******** in ****, * ******'* **** **** not ****** **** ****. *** ******* of * ******'* **** ** *** 80+ ** ********* ********* **** *** one ** *** *.

***** ************ ****** ** ******* ***** 3 - * *****, *** **** common ******** ** ******* **** ****** that **** ** ******* ****** *******, irises *** ****** ********** ********, *** many **** ******** *** **** ******* corrective ****** ** **********.

****,**** ****** ****** *********** *****. ******** ******** ** *** iris ****** ** *** ****** **** long ****** ***** **** ****** **** than * ****** *********, ** *** 'urban ****' ** ******* **** ***** used ** ***** ******* ** ******* not ********.

*******, ****** *** *** ****** **** to **** ** *** ****** ******. For *******, ***** **** ******* ** ******** eyes ***** **** ****** ******** ****** **** for **** ******* ** **** *******. This ********* ***** *** ******* * read, *** *** ************* **** **** the ******* ******* *** *** ***** compared ** * *********** ** ****** vein.

SRI **** ****** ********

***'* **** ** *** **** (***) Access ******* ****** ******** ***'* **** ****** and ********* ********** ******** ***** **** a ******* ******.  ***** *** **** can ** ******* **** * ****** tablet, *** ******** *** ****** ** their *** **** **** ***** ** a ********** ****** **** ** * door.

*** ***** ***** ******** *** ******* demo ***** *****:

***** *** '****** *******' ****** ** new, ***'* *** **** ****** ********** is ***, ** ** ***** ******* in ****'* ******* ****** **** **** ****. **** **** ****** ***, *** added *****-****** ************** *** ******* *** two-way ***** ************* ******* *** ********.  The **** ** ******** ** * customized Android **, *********** ******* ******* ****, measuring ~*"**".

** ***** ** *** **** *****, the **** *** ** ******* ** a **** ** *** ****/*****, ** can ** ******** *** ****** ** the *** ****** ******* *********** ******** mobile ********** *******.

*** ******** **** *********** ******* *** tablet *** ** ****** ******* ****** is **** *** "* ******* ******* */* **** **** can ** ******** ** ***** ******* interface ********* **** ** ******* ** RS485", *** *** ****** ** *** equipped **** * **** ** *******. SRI ******** **** ***** *********** ******* may **** ******* ********* *** **** this ****** ** *** **** *** be ********* ** *********.

Key ******

*** ****** *** ** **** ******* ways **** ******* **************:

  • **** ***** ** * ******: ** ****, ***** **** ** be **** ***** ** *** ****** (as ***** ** *** ***** *****)
  • *****-****** **************:*** ****** ******** ******** ******* ** PIN ****** *** ** **** ****** in ******** ** **** ********.  ***, up ** *** ***** ******* *** be **** ** ****** *****.
  • **** **********: ***** **** ****** *** ** enrolled, **** *** ***** *** **** can ** **** ** *******. **** irises *** *** **** ** ****.
  • ********** *******: *** ****** ****** **** ** hardwired ******** ************ ******-*** *********** ****. 
  • ****** ****: *** ****** ****** ** *** rated *** ******* *** ** ******** to ****. *******, **** ******* ************ might ** ******** **** * ****** cover ** ****.
  • ******* *******: **** **** ********* *** ****** on *** ****** ******** **** ** GB ** *******, ****** *** ********* of *********.

*****

**** ****** ***** ** * **** MSRP ** $**** *** ****, ******** adding ******** ******** ** ******** ********* cards (******* ** *****) **** ******** the **** *****.

ADA ********** ********

******** *** ****** ** ** *** be ******* *** **** ******** ***** be * ******* *** **** *****. We ***** *** ** ******* *** the ****** ****** ** ******** ** ensure*** **********, *** *** ****** *** **** more **** *** ****** ***** ** installed - ** '****' *** '***' positions, ** ** **** *** ****** was **** * '******* *****' ****** be *********.  

*******, ***** *** **** ** ** actual ******* ******* *** *** ******* (bombastic?) ********** **** *** ******* *** opening ** * ********, ********** ******* should **** *****, ********* ******* *** accessibility ********** ****** ******.

*********

** ***** ** *******, *** **** has ******** **** ******* **** ******* and **** **** ***** ********* ******* do ***.  *** ********* **** ** us:

"*** ****** **** *** ****** *** ******* digital *******, **** *************, ************ ******** and **** *** ********** ***********, ***** other ************. *** ********** ***-**** ****** *** provide * ***** **** ** * VMS ******** *** **** ********** *** recording."

**** ** ***** ******** *** '********' with ********** *********** *** *** ** default, ******* ******** ** * **** reader *** ****** ****** ****** *** potential ** ***** * ******* ******* sensor.

***** ** **** * ********. ***** ~$1,000 ** ** ** *** **** than * ******* **** ** **** fingerprint ******, ***'* *** ****** ****** is ** *** *** *** ** iris ******* **** ******* ***** ** price **** ** ** ** ****.

**********

*******, *** **** *** ******* **** points **** ***** ****** *** ********** success. ** *** **** *****, *** time ** ***** *** * **** to **** *** **** ** *** read ** **** ** ***** **** compared ** * ******* **** *** a ********** **** ** **** *********** reader. ***** * ***** ** *** or *** ******* *** *** ** an ***** *** ***, ** ***** become * ****** ** *********** *** others *** *********** ** * *** enough ***** ** ******* ***** *******.

*** ****** **** ******** ******* ********* problems, ** ***** **** ***** ***** eyes ****** *" **** *** ******. The ***** ***** ** *** ****** can ** **** ************ ** ***** users ** *** **** ***** *****. Even *** ****** ********** ** **** height, **** ***** **** **** ** hunch **** *** ******* ***** *** struggle ** ***** *** ********** *** unit ** ** ****. ** ***** of **** *****, *** ******* ** the **** ********* ********** *** *** reader *** *** *** *** *********** delay ** ***** *** ********** *********** beyond **** *********** **** ** *********** sensors **** *** ****** *** ****** to ***.

** ******** ** *****, *** ****'* fuzzy *********** ** *** *********** *** indoor-only *********** ***** ** ** **** specific ************ ** * ********, *** it ****** ** ******* ** '******* purpose' ****** ******* ****.

***********

***'* *** ****** ****** ** ****** at *** *** *** ** *** market ** ******* **** *******, *********:

  • *******'* **** ******, **** "***" ***** ******* ** $**** **** from *******
  • ******'***************, ******* ~$**** **** *******
  • *********'***-*****, ******* ~$****, ***** *** ** Panasonic ** ***** ****** ********* *** resold ** **** ****** ******* [**** no ****** *********].

*** $**** **** ***** *** ***'* IOM ****** ****** **** ** * pricing ********* **** ***** ** *********** enough ** **** *** ***** ************ of *** ****.

Comments (14)
Avatar
Brian Karas
•Apr 19, 2016
IPVM

There seems to always be talk about biometric access control systems, and I see the theoretical benefits, but these all seem like a tough sell over cheap cards/badges.

I get the "no lost cards" benefit, but it seems like as your probability for lost cards scales up (more employees, more doors), the costs of biometric systems would scale up even faster.

I'm not sure if these biometric systems are solving real problems, or just giving manufacturers a method to convince users to upgrade/replace existing readers, which have a long life-cycle and probably don't get touched for 20 years after install.

(1)
KA
Konstantin Avramenko
•Apr 19, 2016
In reply to Brian Karas

Independently of biometric modality, the main benefit of biometrics for the access control is that this credential is "locked" to the user. It makes easier to control attendance. It is more convenient in case of temporary workers. It is more secure in case of large organisations or premises where employees do not not know each other personally. It is more secure to build the unified access control system (physical and logical) on the biometrics. But, of course, everything depends on customers' needs.

(2)
DK
Dale Kougel
•Apr 20, 2016

Full disclosure, I work for EyeLock. I just wanted to clarify two points as we were mentioned in the article.

#1 - Although at one time Stanley Security had exclusive distribution rights for EyeLock product in certain regions, that is no longer the case. EyeLock has setup their own sales channels around the world through distribution, oem and integrator partners.

#2 – MSRP for our nano NXT is $1650 USD and includes all hardware and software for interfacing to any access control panel via standard protocols (Wiegand, F2F, etc).

(1)
Avatar
Brian Rhodes
•Apr 20, 2016
IPVMU Certified
In reply to Dale Kougel

Thanks Dale! I updated the post.

U
Undisclosed #1
•Apr 20, 2016
IPVMU Certified

While both irises can be enrolled, only the first one that can be read is matched.

What do you mean by read? Do you mean matched or just scanned regardless of whether it matches?

Avatar
Brian Rhodes
•Apr 20, 2016
IPVMU Certified
In reply to Undisclosed #1

Authentication is performed on one iris, not two.

SRI tells us that users can enroll both eyes and present both eyes for a potential read, but the decision to grant/deny is based on whether the first recognized iris is a match.

U
Undisclosed #1
•Apr 21, 2016
IPVMU Certified
In reply to Brian Rhodes

In that case it sounds like you would pretty much need to enroll both eyes, otherwise you would risk getting denied half the time, no?

Avatar
Brian Rhodes
•Apr 21, 2016
IPVMU Certified
In reply to Undisclosed #1

No, if you enroll your left eye and only present your left eye, you aren't going to be denied half the time because of it.

Enrolling two eyes may help matches be made sooner, but it doesn't increase accuracy rates.

(1)
U
Undisclosed #1
•Apr 21, 2016
IPVMU Certified
In reply to Brian Rhodes

Got it. I didn't realize the FOV of the reader was just one eye.

Avatar
Brian Rhodes
•Apr 21, 2016
IPVMU Certified
In reply to Undisclosed #1

It isn't necessarily. If a user (think child or someone with close eyes) is able to present both within the 8" range, it only matches based on the first iris it can read.

I'll ask SRI for the FoV spec on the iris sensor and update here.

U
Undisclosed #1
•Apr 21, 2016
IPVMU Certified
In reply to Brian Rhodes

Ok, now I feel stupid again. :)

So if someone presents both eyes, the system might choose the left or the right, since it doesn't know which eye is enrolled. If it chose one that was not enrolled, I'm understanding you to say it would deny. But you're also saying it wouldn't help increase accuracy to enroll both eyes, which it would seem to in this case.

Avatar
Brian Rhodes
•Apr 21, 2016
IPVMU Certified
In reply to Undisclosed #1

Accuracy has no bearing on whether an iris is enrolled or not. It affects how well/to what level of quality an iris is read. Of course - if an iris is not enrolled it will not be valid - but that isn't a function of reader performance.

More opportunities for a match is not the same as increasing accuracy of the read.

If a user only enrolls one eye, they should only attempt to verify on that one eye. The demo video in the post may help clear up how the user does this.

U
Undisclosed #1
•Apr 21, 2016
IPVMU Certified
In reply to Brian Rhodes

Ok, I think what you are saying is that if two eyes are presented it will read one and try to authenticate on the one eye. If it fails then it tries on the second eye.

It only uses one eye to "match" but it still may "read" both eyes if it needs to and both are in view. Is that right?

U
Undisclosed #2
•Apr 21, 2016

We've been using Eyelock for quite some time, still have some test units on our walls as well. I can't speak for the SRI solution but the biggest challenge with the iris scanners is securing them - no one really makes a good vandal resistant unit and most of your access control systems start at the perimeter.

In regard to false reads, we get about 1-2% false reads (no entry) but the bigger issue of positive false reads we are seeing less than 1 in 1000. If you enforce dual iris the chances of a false positive are.... I dunno, we have not had a false positive with dual iris yet.

(2)