Certify "The #1 Fever Detection Solution" Tested

Derek Ward
Published Nov 18, 2020 15:39 PM

"Certify" claims to be "The #1 Fever Detection & Thermal Scanning Solution in the Marketplace" and "found in more than 75% of NFL stadiums", but how well does it really work?

IPVM Image

IPVM purchased and tested the Certify SnapXT Pro terminal, examining the following:

  • How accurate are forehead and inner canthus measurements?
  • Does the device measure temperatures as-is, or does it average them to appear "normal"?
  • How often does it miss elevated temperatures?
  • How accurate are measurements vs an IR thermometer?
  • How does subject distance to the terminals impact temperature measurements?
  • Do environmental factors impact performance?
  • How useful is the Certify cloud portal?
  • How accurate is the device's mask detection and can it be spoofed?

"Found In More Than 75%

A Certify executive posted that these devices can be 'found in more than 75% of NFL stadiums':

IPVM Image

He included a picture of the system deployed by the Colts, excerpted below:

IPVM Image

Executive Summary

The Certify SnapXT Pro missed moderately elevated temperatures (~100°F) while making lower and elevated temperatures be reported as more normal. Measurement accuracy declined further with Certify's "Multiple Scan" mode enabled, reporting multiple subjects at the same temperature, despite differences of 1-3°F when measured with an IR thermometer.

On the positive side, the device has an online portal with the ability to easily export reports, add users/members, create questionnaires for guests, and alter device settings remotely. Additionally, there is US-based tech support that can remote directly into any connected device on the portal should issues arise.

Certify's online portal and local support are strong differentiators likely to attract organizations looking to deploy many fever screening terminals. However, the device's inaccurate group screening mode and missed elevated temperatures are likely to lead to significant numbers of missed detections.

Vs. Dahua / Hikvision / ZKTeco

Compared to Dahua, Hikvision, and ZKTeco temperature measurement terminals, the Certify SnapXT Pro suffers from more missed detections and had greater issues with subjects of varying distances. It also has roughly the same price as the other tablets at ~$2,000 after selecting a tablet mounting option.

Vs. Sperry West / Alibaba Terminals

Compared to the Sperry West and Alibaba terminals we have tested, Certify SnapXT Pro had similar measurement results, though with the inclusion of remote monitoring, reporting and configuration. These models (at least off Alibaba) are priced much lower than Certify's Pro offering at ~$500 online, or less in larger quantities.

Missed Elevated Temperatures

With subjects' temperatures elevated over Certify's threshold of 100.4°F to nearly 101°F, the reported measurements were ~2-3°F lower than a calibrated Extech thermometer and outside Certify's own 0.3°C / 0.6°F accuracy specs. Missed elevated temperatures could be seen on either the forehead or inner canthus.

IPVM Image

Slight Measurement Difference Due To Distance

Measured temperatures fluctuated slightly depending on subject distance, by ~0.5°F.

IPVM Image

Temperature Averaging In Group Mode

The Certify SnapXT Pro allows users to toggle on a "Multiple Scan" or Group mode that allows more than one subject to be measured at a time. However, we found that the device averages temperatures in this mode. Shown below, two users are measured at 98.1°F, despite measurements of 98.4°F and 97.7°F respectively.

IPVM Image

The other downside of Certify's group mode is measurements well outside their recommended range. During our tests, subjects were measured as far as ~9'6" from the device. In the example below, the subject is measured at 92.5°F, which was well below the subject's ~97.7°F temperature measured by IR thermometer.

IPVM Image

Blackbody Testing

Blackbody tests were conducted with the device in single and group user modes, and temperature normalizing can be seen throughout, which is a root cause of missed elevated temperatures. From our test with a blackbody and a subject with a steady 36.5°C body temperature, we noted the following:

  • Certify kiosk showed a ~1°C lower measurement with subject's showing elevated temperature due to averaging algorithm.
  • Certify kiosk "averaged" close blackbody and human measurements.
  • Paper target was measured higher when in group mode than in single user mode.
  • Subject temperature was unusually "cooled" as paper blackbody go warmer.

IPVM Image

IPVM is able to demonstrate Certify's temperature averaging while in group mode using our paper cutout over the blackbody and a subject leaving the scene. In the beginning of the example below, the subject is measured at 97.5°F and blackbody at 97.3°F. As he exits the scene, the temperature measurement on the blackbody decreases to 96.8°F, a 0.5°F difference.

IPVM Image

Hats And Glasses Not Recommended

In our testing, we found that if the warmest part of the face (whether it be the inner canthus or forehead) was blocked, measured temperatures would drop. However, if both were worn and this area could not be seen, the terminal produced extremely low measurements looking for the next warmest spot on the face. Certify recommends that hats and glasses be removed prior to temperature measurements being taken.

IPVM Image

Adjustable Wall Mount

Certify has a recommended mounting height range of ~5', but unlike most other terminals we have tested, it includes an adjustable wall mount which may be moved to accommodate different heights. The device can be pitched up or down by ~10°, demonstrated below:

IPVM Image

Mask Detection Spoofing

Mask detection was easily spoofed by hands, misworn masks and everyday objects such as phones or coffee cups. Additionally, if the cup of coffee is hot, the device will screen a subject with an elevated temperature.

IPVM Image

Certify Web Portal

Certify has an online portal that allows users to add and configure devices, users, members and other items. In the video below, we demonstrate the following:

  • What the dashboard looks like and how to review temperatures and images.
  • Adding a device/kiosk to the account.
  • Reviewing and exporting reports.
  • Configuring the settings on a kiosk, and what the exclusive settings are on the online portal versus the local device interface.
  • Additional settings/options on the portal.

Certify.me Cloud is a subscription-based service that users can pay for per-device. Their subscriptions start at a "Standard" package at $9.95 a month up to $49.95 a month for "All Access". Upon purchasing a Certify kiosks, users get a free month's access of the portal.

IPVM Image

Local Device Configuration

Device configuration can be done locally, with many of the same options present as there were in the online portal. However, there is one crucial setting found locally, and that is temperature calibration. Per Certify, calibration only needs to be done when the device is initially set up. The video below demonstrates the device's calibration, as well as how to access admin menus.

When compared to the Telpo's local interface, the Certify kiosk's software has much more in terms of options and configurations. Certify has dedicated options for access control, expanded audio and visual settings and various other settings not found on Telpo's kiosk.

IPVM Image

Certify's Odd Recommendations and FAQ

Through Certify's FAQ, we found a few odd and contradictory statements and examples. For example, they have defined installation environment diagrams, though in the FAQ, Certify answers that surround temperature does no affect calibration or body temperature readings.

IPVM Image

In the FAQ, Certify also asks that subject's removed masks before being measured, though the device is capable of measuring subject temperatures wearing masks.

IPVM Image

Certify also had and answer for return policy in their FAQ, which states that there isn't one.

IPVM Image

Size Comparison

The Certify SnapXT Pro is similar in size to other full-sized tablets such as the Hikvision MinMoe and ZKTeco 8" models.

IPVM Image

Thermal Sensor

Below is an up-close image of the Certify SnapXT Pro thermal sensor. After researching the thermal sensor, we found it to be a GUIDE CB360, which has 120x90 resolution.

IPVM Image

Made in China

The Certify SnapXT Pro is Made in China. The country of origin sticker can be easily applied and removed.

IPVM Image

Version Tested

The Certify SnapXT Pro kiosk was tested using software version 3.3.126 on android version 7.1.2.

Comments (5)
Undisclosed Manufacturer #1
Nov 18, 2020

The fraud continues. Albeit with a little more salad dressing than most.

Anything that "averages" temperatures is a fraud.

Let's all sing in unison...A FRAUD, A FRAUD, A FRAUD!!!

"wrongful or criminal deception intended to result in financial or personal gain"

Michael Gonzalez
Nov 18, 2020
Integrated Security Technologies, Inc.

Handy little tab on the country of origin sticker...yikes.

Undisclosed Integrator #2
Nov 18, 2020

"The country of origin sticker can be easily applied and removed."

Is this more a reflection on how dumb the buyers are? So easily fooled, beguiled that they believe blindly? Seriously, the seller takes such little action to hide the deception because they feel the buyer is just that easy?

It's like people believing someone that would say, "The virus is just gonna disappear. One day - like a miracle - it will disappear."

John Honovich
Jan 24, 2021

This company is now running ads on Instagram claiming to detect temperatures on the backs of people's heads:

IPVM Image

John Honovich
Mar 06, 2021

Update: Certify has pulled this product from the market after an IPVM peer reviewed study, Washington Post report, and an FDA warning letter to Certify.

IPVM Image

Certify falsely claimed to WaPo that multi-person scanning was not activated for any customers:

Neisser said the company is “in the process of removing any multi-scanning functionalities from the SnapXT Pro HID.” She acknowledged those features are “built into the software” but said they are “not a core part of the product” and have “not been activated for any customers up to this point.”

We know this is untrue because we bought one of their products and was able to use and include it in our test that we released 4 months ago:

IPVM Image