Our sales team coerced our operations group into evaluating some of the ZK product line a few years ago to expand our lineup offerings. Of all the products we tested, the most disappointing was their biometric camera (not the same as reviewed in this article). It was a low quality build. After seeing it up close and examining the construction of the device, I cannot imagine their quality has improved since that time. The entire company focus is on low quality "security" products.
Thank you for sharing your ZKTeco experience from a few years ago. I ask you to consider all our panels are UL certified and our patent pending SpeedFace readers received SSI Magazine 2019 Top 30 Technology Innovation award and 2020 Govies for best biometric access control device. Our SF1008-WP is both IP68 and IK04 rated. We've also introduced our new Atlas Series of software-less access control panels designed and engineered in the U.S. They support up to 84 doors and support both PoE and Wi-Fi. Many changes have occurred these past few years including opening a 5,000 sq ft user experience center in Atlanta. Our company continues to grow by learning from candid customers like yourselves where we need to improve. Thanks for sharing.
our patent pending SpeedFace readers received SSI Magazine 2019 Top 30 Technology Innovation award and 2020 Govies for best biometric access control device.
Citing trade magazine and govies awards mean nothing, as lots of dubious products claim the same.
Our SF1008-WP is both IP68 and IK04 rated.
To be clear, the '-WP' does not include thermal. Is that correct? Can the body temperature sensor model be used outdoors?
The "WP" in our part number refers to "weatherproof" (IP68). While our SpeedFace model SF1008-WP IS weatherproof . . . we lost the IP68 rating when we attached the external thermal sensor. So SF1008+ is NOT weatherproof.
Great question about outdoor use. It's very challenging to have accurate body temp reads outdoors. "Best practices" for outdoor-use calls for using a tent to block the wind and attempt keeping the air temp inside the tent a constant. Even using a black body with thermal camera will be a challenge if you cannot control the ambient temperature and wind.
Please let me know why you don't value the SSI and Govies awards we've won. We're quite proud of them.
I have to say I was very surprised by the statement:
"… why visible light, most traditional facial recognition solutions, they pretty much rely on infrared technology because infrared gives you that accuracy."
While this certainly isn't the case. These days, I would say pretty much every facial recognition solution relies on artificial intelligence facial recognition but then, effectively, the measuring the various data points upon the face to actually produce a facial signature that will be encoded and stored in the database for matching. Nothing in there regarding infrared at all. Know what I particularly expected to be given the, typically, degraded resolution of an infrared image compared to visual light image.
Thanks for your comment, Malcom. From my perspective (as a 15-year biometric reader manufacturer) I think of competing biometric reader manufacturers (not face recognition software companies taking feeds from common surveillance cameras which use visible light). The most popular applications utilizing biometrics are 1) Time & Attendance and 2) Access Control. Since these applications require precise matching, the devices normally use an infrared light source. This is also the reason why you can find very cheap ordinary visible light camera modules online for less than $50 whereas infrared camera modules are much more expensive. As I'm sure you're aware, accurate recognition requires 1) Clear image and 2) Fast-accurate matching algorithm. My remark regarding "image quality" was contrasting infrared vs visible light cameras.
Interesting Larry, and very informative, I honestly had no idea that facial recognition style access control systems relied so heavily upon IR. Almost every facial recognition system that I have ever examined, going back to 2002, uses visible light cameras to collect the images. Now, I must confess, these systems have always been for deployment in specific [hotel and casino] environments that are probably closer to "public space facial recognition" and the systems you appear to describe.
Thanks Malcom. You're exactly right about public spaces. You typically won't see infrared cameras being used. The biggest obstacle I face as a relatively new vendor (only being in the USA security market for 5 years) is people understandably having a natural bias for the vendors and technology they already know. For instance, an IPVM member still thinks ZKTeco copied Uniview's body temp detection reader because (I assume) they like Uniview's cameras. Likewise, most people think "cameras" when they think "security". Therefore, a "camera centric" person will assume face recognition is best accomplished by combining an ordinary visible-light camera with any 3rd party face-matching algorithm running on a computer. And that's fine if you need to pick out "the most likely person" who matches the face in a big crowd. But for any application which requires matching a face precisely 1:1 . . . infrared provides the best accuracy. You can't unlock a door by an algorithm concluding "Malcom" is the "most likely" face match. That wouldn't be secure. Access Control and Time & Attendance applications which require precise user-matching often utilize stand-alone infrared face matching terminals (not surveillance cameras). And NOW, ZKTeco has a unique dual-camera design combining both infrared and visible light which provides fast accurate face matching and remote enrollment. Thanks again for listening in and providing your feedback. I'm far from being an expert on video surveillance. Please let me know if you disagree with any of my comments. I, too, enjoy learning new things. Thanks Malcom. I'd welcome the chance to work with you on a project some time.
Are the thermal devices available through either Anixter or CDW-G? Is there a pricing schedule for .edu? We self-integrate for uur access control and cctv systems. We may want to procure two devices for a pilot. What is the expected turnaround from order to delivery?
Integration plans.
We have a large installed base of CBORD ACS (over 50,000 active records) and HID readers (several thousand readers) that would need to be integrated into any large deployment. As a standalone device that does not integrate to cboard and does not require a card tap this device would be routinely bypassing our ACS and allowing access to users who may have been revoked/restricted in the enterprise ACS.
While this may be manageable for our one/two door immediate use case, it will not scale. What are the plans to integrate into CBORD?
Also, does this camera support onvif so that our VMS/NVRs can record the stream being produced by this camera. Is the stream being recorded at the device or in the cloud?
Great question. Central stations monitor streaming video. But our device doesn't stream because its not a camera. It's an access control device. Our device stores and matches credentials (i.e. PIN, face, palm) and creates a "transaction" when a match occurs. Based on the device (or external panel) ACL, the door will unlock. A log is kept on the device of all "transactions" (i.e. device ID, User ID, Date, Time, Body Temp). The log file is then sent to our management software or any 3rd party software (by using our SDK). The devices can be managed remotely from any mobile device (i.e. smartphone, tablet, laptop, PC).
Thanks for writing. Your technical questions are best answered by our SpeedFace product manager Esteban (esteban.p@zktecousa.com) but I'm happy to take a crack at them. All our products are sold through authorized distributors (inc. Anixter and numerous low voltage distributors). We don't have standard EDU discounts, but we do make pricing concessions when needed. Under normal circumstances we maintain local stock in our Atlanta warehouse and select stocking distributors. But last 2 weeks our inventory has been drained and we're looking at 3 weeks turnaround. Although the readers don't require software, for remote management we recommend using our feature-rich ZKBioSecurity software which scales up to 8,000 doors. Kindly note we're not just a biometrics reader vendor. We have a very comprehensive product line which also include a visitor authentication and management software specifically designed for schools called ZKVAMS-EDU. It integrates with the thermal readers. Regarding ONVIF, kindly note the thermal readers are not surveillance cameras. They do not stream video. They do not need a VMS software and NVR. They store & match face "templates". When a match occurs (either PIN, Face, Palm) the "transaction" is stored in the reader and can be output to either our software or a 3rd party software (we have a device SDK).
Great questions! I hope you understand our device is not a "video camera". For your school clients I hope you'll consider our walkthrough metal detector with body temp detection. It can address both metal detection (i.e. weapons, vape pens) and elevated body temp in one device. Parents will love it. Please e-mail me at larry.r@zktecousa.com. Thank you.
Thanks Larry. I appreciate all the answers and we will reach out to you directly.
While I understand that you have a complete offering of Access control devices that integrate with your own ACS, clients with a mature ACS deployment are unlikely to do a full rip and replace in order to deploy your devices. Also many doors would still use more traditional card access with biometric security used at more high value entrances. Additionally clients might want to use both biometric and card access as we do currently in our more restrictive spaces.
We might buy dozens, even hundreds of these devices if they integrated with CBORD. Absent that integration every deployment of a ZKteco reader reduces our security posture as it represents way to bypass our enterprise ACS and/or use the CBORD/HID reader that will have to remain to bypass the ZKteco reader/thermal checks. We might still use the device in targeted deployments but at a much smaller scale.
Still, looking forward to getting one or two of these to pilot.
Thanks for the clarification. Kindly note our devices are merely Wiegand readers, installing no differently than card readers (besides running CAT5e cable to each door for TCP/IP connection). The only real difference of course is answering the question; "How do you manage the biometric templates"? If you cannot use our ZKBioSecurity management software, we do have an SDK to allow the CBORD system to manage our readers. Likewise, if CBORD provides us their API, we can then write the integration. Considering e-mailing me at larry.r@zktecousa.com if you'd like to further discuss. Thanks for your feedback!
Thanks for your inquiry. The SpeedFace readers with body temp detection (models SF1008+ and SF1005-V+) thermal scanning area encompasses a person's forehead and eyes' tear ducts. For more info you can e-mail me at larry.r@zktecousa.com.
Hi, a simple question. Product "thermal sensor" is a termographic camera with a micro-bolometer (seems not to me, as no data is provided) or a simple thermopile or similar sensor pointed straight forward (hopefully on user's forehead as user is very close)?
Hi Tommaso. When using our standalone device (with integrated face detection and thermal sensor) the person places their face within 18 inches of the device. If you're interested how the thermal sensor works, it uses infrared thermopile technology which you can research on the Internet if you care to.
I don't have documentation to share. But Brooklyn Low Voltage Supply tested our SpeedFace reader against a hand-held thermometer and we differed only 0.07 degrees. Fast forward to 2:35 minutes in the following video
I love the Brooklyn Supply guys... but the 94 degree temperature reading from the Mobotix solution (2:04) shows that Mr. Gideon is dying of hypothermia and yet they still love all 3 of their 'solutions'.
I'm thankful for your OCD. It reminds me I need to update my avatar. Previously I had only a temporary IPVM membership. I've since upgraded my membership which required I create a new profile and my name Lawrence appears on my credit card.
I'm not aware of any body temp detection vendor (inc. camera, access control or kiosk) who actually manufacture their own thermal sensors. We all source from a handful of thermal sensor vendors (FLIR being the most famous). I've never before been asked to provide details how thermal imaging occurs with our devices. Customers primarily ask for performance metrics and then request test equipment so the customer can substantiate the vendors' respective claims. All I have to share with you is that our device uses infrared thermopile technology which you can research on the Internet if you care to.
Thanks for your question and reminder to update my avatar.
I'm not aware of any body temp detection vendor (inc. camera, access control or kiosk) who actually manufacture their own thermal sensors. We all source from a handful of thermal sensor vendors (FLIR being the most famous). I've never before been asked to provide details how thermal imaging occurs with our devices. Customers primarily ask for performance metrics and then request test equipment so the customer can substantiate the vendors' respective claims. All I have to share with you is that our device uses infrared thermopile technology which you can research on the Internet if you care to.
So you have no documentation backing up your statements. People need to understand what part of the face is measured for temp. I don't think this is unreasonable to ask.
Your question is not unreasonable. But based on my customer experience past 2 months with SpeedFace+, I've found the primary statement customers expect backed up is that SpeedFace accuracy is +/- 0.6 degrees Fahrenheit when a person positions their face within 18 inches of the reader (in a temperature-controlled room).
As mentioned, our device uses infrared thermopile technology (ITT). You can search the Internet to learn what parts of the face are scanned by ITT for temperature detection if you feel it necessary to close sales with your customers.
I tried asking a variation on this question during a webinar the other day and as questions were being reviewed afterwards it was skipped over....
How and what is being measured to determine the temperature? Also, repeatedly during the webinar I heard it referred to as body temperature - is it body temperature or skin temperature?
I anticipate it being asked of me should I attempt to sell this solution and would need to know the answer to even consider the promotion of these.
I'm not aware of any body temp detection vendor (inc. camera, access control or kiosk) who actually manufacture their own thermal sensors. We all source from a handful of thermal sensor vendors (FLIR being the most famous). I've never before been asked to provide details how thermal imaging occurs with our devices. Customers primarily ask for performance metrics and then request test equipment so the customer can substantiate the vendors' respective claims. All I have to share with you is that our device uses infrared thermopile technology which you can research on the Internet if you care to.
"All I have to share with you is that our device uses infrared thermopile technology which you can research on the Internet if you care to. "
From Internet
MEMS thermopile is super-sensitive, accurate to +/-1.5°C It can measure the surface temperature of an object between -40 up to +80°C in the target area with an accuracy of +/-1.5°C and a resolution of 0.06°C.
I can't see how it can be used for body temp detection with +/-1.5C accuracy
It's a tool, which when used properly can be effective. Although ZKTeco offers a model having thermal camera and black body like many famous surveillance camera companies, we primarily promote our SpeedFace+ standalone reader, instead. When using SpeedFace+, users position their face within 18 inches of the device (similar how you interact with a card access reader) and we've seen consistent +/- 0.6 degrees Fahrenheit results. If you feel customers will benefit, ideally its best to test a unit. Prior to testing, consider watching a live demo or ask a trusted party about their experience with it.
BLVS is one of the most highly respected distributors in the low voltage industry. Watch their experience testing Dahua, Mobotix and ZKTeco SpeedFace body temperature detection readers.
ZKTeco, you cannot use undisclosed to promote your company.
Secondly, BLVS did a demonstration, not anything close to a fair, controlled test. For example, they tested Dahua and Mobotix without their blackbodies.
When you make comments like this that are so obviously flawed, it raised major questions about your company.
This is NOT a test. It is barely a product demo and mostly a sales pitch for all 3 devices demoed. ZKteco is undermining their own credibility by citing this youtube video as a test(imonial).
I'm not claiming the BLVS video is a testimonial. As a trusted distributor, there's value in seeking their non-biased opinion on the many products they carry. Doubtful BLVS would stock & support products they don't stand behind. Anyone familiar with BLVS would agree.
I have talked to you enough that I am generally convinced that you mean well but do not know what you are talking about. You honestly should not be in the role that you are in. The things you say are what I would expect from an entry-level regional salesperson, not the CEO of the USA division. You cannot come here and say such wildly misguided things promoting your production and not be called out for it.
I am starting to come to the conclusion that ZKteco doesn't know how their own product works.
What happens in a company uses this to screen employees for work and they deny someone a day's work because this unit says their temp is high? What happens when the employee sues the company and ZKteco gets dragged into the suit? Are they going to show the BLVS video as proof that the thermal unit is accurate?
Michael, you raise an excellent concern regarding "policy" which pertains to the use of ZKTeco and all vendors' products incorporating body temp detection. Most likely your trusted surveillance camera vendor has also incorporated body temp detection.
Body temperature recognition is one of several methods restricting access, just as walkthrough metal detectors and biometrics can be used to restrict access. Policies and procedures need to be established by endusers which address the many different possible scenarios which can occur.
I accidentally had the "post undisclosed box" checked. I don't think it's flawed suggesting IPVM members lacking experience with a vendor/product reach out to a trusted party (inc. BLVS) for their opinion or experience. I understand it's hard trusting an unfamiliar vendor or product.
I like to be forthcoming about products I sell my customers. Temperature is a tool and not a be-all end-all solution.
I deal with a lot of IT people that end up also playing security roles for their company. Many of them do a lot of reading on the product technologies we offer them and there are is a lot of conflicting information on thermal products. How and what is measured to determine the temperature is a legitimate question.
When we reach out to a manufacturer to get answers to questions such as this and they are not able or willing to answer the question, it raises red flags about the product.
If the sensor is accurate to withing 1.5 degree C, but the manufacturer of the product says it is accurate to within .6 degree C, something besides anecdotal evidence of "...we've seen consistent +/- 0.6 degrees Fahrenheit results..." would be useful. We look to the manufacturers for information useful for selling the product, but the information provided has to be consistent with the other technical information. If not, they have to provide other compelling information to justify their claims.
The whole temperature screening process seems to me to be a road that has the possibility of future lawsuits and dissatisfaction from end-users if it does not live up to the hype. I want long-term customers and not people I can grab their money and run from, never to be seen again. For us to offer anything, it has to do what it says OR the customer has to understand its technical limitations so that we are managing the customer expectations. To do that, the manufacturer HAS to provide accurate information.
I have used some ZK Teco access products in the past and have been very satisfied with them They have performed well for the projects they are on and have been cost-effective. So far, I could not offer this product since I do not have enough information to properly inform my customers.
There's both hype and skepticism in the market as more and more surveillance camera vendors, access control vendors, kiosk vendors, and private-labelers add thermal sensors to their products and jump in and compete for their share. You can often tell newcomers (at least in the USA) because they don't even publish measurements in Fahrenheit or inches/feet . . . or they cite health benefits derived by using their products.
One thing in common all these vendors have is that none are designing or manufacturing thermal sensors besides Flir and a handful of others.
Having legal concerns regarding body temperature detection is reasonable. But considering the global demand for these type products, I think it's well worth the time researching it and determining if it compliments your current business offering.
Thanks for your comment.
I'm happy to assist you with a Speedface evaluation.
Can you use this as a stand alone screening device to check for temp and mask compliance only and not enroll any face, palm or passcode into the system? Then utilize the relays to notify screeners to do secondary checks to verify if necessary.
Hi John. Thank you for your question. Your application is very popular. Yes. SpeedFace can operate standalone while detecting and reporting upon both registered users and unregistered users (visitors). In our daily webinar/live demo we demonstrate your application by triggering a red or green LED controlled by each person's body temp and/or presence of mask. We also offer an SDK if you'd like to export the audit log to an external database. When using our device management software you can additionally schedule e-mail alerts.
IPVM conducts reporting, tutorials and software funded by subscriber's payments enabling us to offer the most independent, accurate and in-depth information.
Comments (46)
Undisclosed Integrator #1
Our sales team coerced our operations group into evaluating some of the ZK product line a few years ago to expand our lineup offerings. Of all the products we tested, the most disappointing was their biometric camera (not the same as reviewed in this article). It was a low quality build. After seeing it up close and examining the construction of the device, I cannot imagine their quality has improved since that time. The entire company focus is on low quality "security" products.
Create New Topic
Malcolm Rutherford
I have to say I was very surprised by the statement:
"… why visible light, most traditional facial recognition solutions, they pretty much rely on infrared technology because infrared gives you that accuracy."
While this certainly isn't the case. These days, I would say pretty much every facial recognition solution relies on artificial intelligence facial recognition but then, effectively, the measuring the various data points upon the face to actually produce a facial signature that will be encoded and stored in the database for matching. Nothing in there regarding infrared at all. Know what I particularly expected to be given the, typically, degraded resolution of an infrared image compared to visual light image.
Create New Topic
Larry Reed
Thanks for your comment, Malcom. From my perspective (as a 15-year biometric reader manufacturer) I think of competing biometric reader manufacturers (not face recognition software companies taking feeds from common surveillance cameras which use visible light). The most popular applications utilizing biometrics are 1) Time & Attendance and 2) Access Control. Since these applications require precise matching, the devices normally use an infrared light source. This is also the reason why you can find very cheap ordinary visible light camera modules online for less than $50 whereas infrared camera modules are much more expensive. As I'm sure you're aware, accurate recognition requires 1) Clear image and 2) Fast-accurate matching algorithm. My remark regarding "image quality" was contrasting infrared vs visible light cameras.
Create New Topic
Undisclosed End User #2
Numerous questions
Purchasing questions.
Are the thermal devices available through either Anixter or CDW-G? Is there a pricing schedule for .edu? We self-integrate for uur access control and cctv systems. We may want to procure two devices for a pilot. What is the expected turnaround from order to delivery?
Integration plans.
We have a large installed base of CBORD ACS (over 50,000 active records) and HID readers (several thousand readers) that would need to be integrated into any large deployment. As a standalone device that does not integrate to cboard and does not require a card tap this device would be routinely bypassing our ACS and allowing access to users who may have been revoked/restricted in the enterprise ACS.
While this may be manageable for our one/two door immediate use case, it will not scale. What are the plans to integrate into CBORD?
Also, does this camera support onvif so that our VMS/NVRs can record the stream being produced by this camera. Is the stream being recorded at the device or in the cloud?
Thanks for the presentation.
Create New Topic
Larry Reed
Thanks for the clarification. Kindly note our devices are merely Wiegand readers, installing no differently than card readers (besides running CAT5e cable to each door for TCP/IP connection). The only real difference of course is answering the question; "How do you manage the biometric templates"? If you cannot use our ZKBioSecurity management software, we do have an SDK to allow the CBORD system to manage our readers. Likewise, if CBORD provides us their API, we can then write the integration. Considering e-mailing me at larry.r@zktecousa.com if you'd like to further discuss. Thanks for your feedback!
Create New Topic
Undisclosed Integrator #3
Does anyone have specs on the thermal camera used on the ZKteco panels? What part of the face/head is measured? Thank you
Create New Topic
Larry Reed
I don't have documentation to share. But Brooklyn Low Voltage Supply tested our SpeedFace reader against a hand-held thermometer and we differed only 0.07 degrees. Fast forward to 2:35 minutes in the following video
Create New Topic
Undisclosed Integrator #5
This all reeks of snake oil.
Create New Topic
Undisclosed Manufacturer #6
[Note: Poster is from ZKTeco]
BLVS is one of the most highly respected distributors in the low voltage industry. Watch their experience testing Dahua, Mobotix and ZKTeco SpeedFace body temperature detection readers.
Create New Topic
Steve Stowe
I like to be forthcoming about products I sell my customers. Temperature is a tool and not a be-all end-all solution.
I deal with a lot of IT people that end up also playing security roles for their company. Many of them do a lot of reading on the product technologies we offer them and there are is a lot of conflicting information on thermal products. How and what is measured to determine the temperature is a legitimate question.
When we reach out to a manufacturer to get answers to questions such as this and they are not able or willing to answer the question, it raises red flags about the product.
If the sensor is accurate to withing 1.5 degree C, but the manufacturer of the product says it is accurate to within .6 degree C, something besides anecdotal evidence of "...we've seen consistent +/- 0.6 degrees Fahrenheit results..." would be useful. We look to the manufacturers for information useful for selling the product, but the information provided has to be consistent with the other technical information. If not, they have to provide other compelling information to justify their claims.
The whole temperature screening process seems to me to be a road that has the possibility of future lawsuits and dissatisfaction from end-users if it does not live up to the hype. I want long-term customers and not people I can grab their money and run from, never to be seen again. For us to offer anything, it has to do what it says OR the customer has to understand its technical limitations so that we are managing the customer expectations. To do that, the manufacturer HAS to provide accurate information.
I have used some ZK Teco access products in the past and have been very satisfied with them They have performed well for the projects they are on and have been cost-effective. So far, I could not offer this product since I do not have enough information to properly inform my customers.
Create New Topic
Larry Reed
Thanks Steve.
There's both hype and skepticism in the market as more and more surveillance camera vendors, access control vendors, kiosk vendors, and private-labelers add thermal sensors to their products and jump in and compete for their share. You can often tell newcomers (at least in the USA) because they don't even publish measurements in Fahrenheit or inches/feet . . . or they cite health benefits derived by using their products.
One thing in common all these vendors have is that none are designing or manufacturing thermal sensors besides Flir and a handful of others.
Having legal concerns regarding body temperature detection is reasonable. But considering the global demand for these type products, I think it's well worth the time researching it and determining if it compliments your current business offering.
Thanks for your comment.
I'm happy to assist you with a Speedface evaluation.
You can e-mail me at larry.r@zktecousa.com
Create New Topic
john hubner
Can you use this as a stand alone screening device to check for temp and mask compliance only and not enroll any face, palm or passcode into the system? Then utilize the relays to notify screeners to do secondary checks to verify if necessary.
Create New Topic
Larry Reed
Hi John. Thank you for your question. Your application is very popular. Yes. SpeedFace can operate standalone while detecting and reporting upon both registered users and unregistered users (visitors). In our daily webinar/live demo we demonstrate your application by triggering a red or green LED controlled by each person's body temp and/or presence of mask. We also offer an SDK if you'd like to export the audit log to an external database. When using our device management software you can additionally schedule e-mail alerts.
You can register for a webinar/live demo at SpeedFace+ Body-Temperature LIVE Webinar - All available dates - Access Control, Biometrics, Entrance Control Manufacturer Access Control, Biometrics, Entrance Control & Visitor Management Manufacturer
Create New Topic