|
|
 |
|
What are Biometrics?
|
|
The scientific analysis of unique human traits or characteristics for verifying identity.
|
 |
|
Are my fingerprints secure?
|
|
YES! Your fingerprints may never be used as a means as identification
any other place but in CheckPrint T/A. Once you have registered your fingerprints
the original captured image is destroyed leaving only a Print ID
model. A Print ID model has no referanceable information to the owner
of the print other than as a means of confirming identity with a newly
scanned finger.
|
 |
|
What modes does the unit work under identification or verification?
|
|
The BioClock units can operate in either identification or verification
modes. Identification is the same as 1 to many sometimes noted as 1:M
or 1:N. In this mode the employee is not required to remember a pin number,
or swipe a card. They simply place their finger on the sensor and they
will be identified based on their enrollment. Verification is the same
as 1 to 1 sometimes noted as 1:1. In this mode the employee is required
to enter a pin number or swipe a badge card. The unit then uses this information
to help in confirming the identity of the employee.
|
 |
|
What is considered a valid enroll?
|
|
A valid enroll can be as little as 1 finger per person or as many as 10
fingers per person. If you enroll a single finger per person, the unit
can store 250 enrollments. If you enroll all 10 fingers per person, the
unit can store 25 enrollments.
|
 |
|
What is the recommended number of fingers to enroll per person?
|
|
In normal office conditions we recommend to enroll 2 fingers per person
on opposite hands. If your environment is more hazardous, where employees
may cut there fingers for example. You may want to enroll all 10 fingers.
Each finger enrolled makes up a single fingerprint template.
|
 |
|
How many fingers can be enrolled in a single unit? And what happens if I go above that number?
|
Currently 250 fingerprint templates can be used in identification mode.
The EER is
set at 1 to 250. However this does not mean that because you enroll 300
fingers that everything falls steeply downhill. The algorithm will perform
as best as possible with all enrolled fingers, regardless of the number
of prints. The 250 fingerprint template limit is at this time what Sense
is confident to endorse. As additional levels can be attained the units
can be updated and additional fingerprints can be added with the same
confidence. To help further explain, after 250 templates you may start
to have FA based upon finger population. If there are a large group of
prints with similar type patterns it is possible for FA to occur in this
situation. If you continue to have a small amount of similar type patterns
than you could go higher and receive no FA. For example there are 100
fingerprints in the system, 50 with a loop, 10 with a swirl, 20 with a
circle, 20 with a arch. There would be a greater chance of a FA with the
fingers of the loop type. There is a great deal of extenuating factors
involved in what effects the number of similar prints and what attributes
to a FA. In the example above just because there are 50 loop patterns
does not mean that you will receive a FA. If enrollment of the template
was excellent than your chances become minimal that a FA will appear.
|
 |
|
What types of operating conditions or environments can the unit operate under?
|
|
Published operating temperatures for standard computer equipment are between
(0C-43C) (32F - 110F) degrees, maximum relative humidity, non-condensing.
SENSE Holdings Inc. Time Clocks operate with in these parameters. To describe
what would happen if the machine were operated outside of the stated operating
environment is very difficult. The unit could lock up, fail to boot, fail
to recognize individuals, or overheat based on environment.
Under optimal
conditions the unit should function properly when adjusted to the environment.
As stated we have additional cooling and heating elements for the units
to overcome some extreme conditions.
If the machines
are located in an area that would have short term exposure to extreme
weather, such as a door opening and closing from the outside. Under these
conditions the units will operate normally.
|
 |
|
What connections are available on your product today (i.e. Ethernet)?
|
|
Our units communicate TCP/IP across any network that supports this protocol.
They do not require a login to the network, they simply use the network
backbone and a dedicated TCP/IP address as a transport.
|
 |
|
Is your unit a ruggedized unit? What NEMA standard is used?
|
|
We have the ability to build to specific NEMA types per customer request.
|
 |
|
Can you meet a 2-3 second total response time from the time that a person places their finger on the platen?
|
|
With present hardware and software this is not a problem In a 1:100 template
database recognition is 1 second. Average response would produce 2-3 seconds
per user. The response of the system has a lot to due with education of
the end user. I.e. use of the correct finger, and correct placement.
|
 |
|
If you meet the 2-3 response time, what is the maximum number of individual templates you could scan in the data base in order to meet that response time?
|
|
At present testing has been done that confirms 1:120 with and average
of 2-3 second response time. With faster processors and advances within
this year on the algorithm we expect to increase this to 1:1000.
|
 |
|
How would accuracy be affected in a clean environment (such as our office environment) or a dirty environment (such as on the shop floor)?
|
|
There are typically three types of environments in the work place. The
first is "very clean" such as a white-collar business office,
the second "moderately" clean such as a factory warehouse, and
the third "very dirty" such as an auto garage. There will be
no difference between these three environments as far as accuracy in terms
of FA. The only difference between a clean environment and a dirty one
is how hard the sensor might have to work in order to obtain the fingerprint
image. This may result in a higher FR.
|
 |
|
What does FAR, FRR, ERR and FTE mean?
|
|
FAR (False Accept Rate) - A false accept (FA) happens any time a unauthorized
person is accepted by the device. The statistical rate at which this happens
out of 10,000 attempts is considered the false acceptance rate.
FRR (False
Reject Rate) - A false reject (FR) happens any time a authorized person
is rejected by the device. The statistical rate at which this happens
out of 10,000 attempts is considered the false acceptance rate.
ERR (Equal
Error Rate) - The equal error rate is the point where the false reject
rate and false acceptance rate probabilities cross on a graph. This intersection
point is a good measure of the overall performance of the biometric device.
The smaller the EER the better the performance of the biometric device
FTE (Failure
to Enroll Rate) - The average number of persons that are not able to be
enrolled using a biometric device because of ergonomic and environmental
conditions, or the demographics of the individual. This rate can be lowered
by the device's ability to adapt to real world conditions, and proper
training on enrollment.
|
 |
|
What is your FTE?
|
|
Due to the physiological make up of individuals certain people cannot
be enrolled using fingerprint devices. These individuals may have psoriasis,
extremely sweaty fingers, or extremely dry fingers, to name a few of the
conditions. Fortunately as these situations present themselves scientists
and engineers are working to solve the limitations of biometrics so all
individuals can take advantage of the security and ease of use of these
new devices. It is very hard to determine a FTE for an biometric since
the rate will change based upon the sampling of individuals at any one
time. We have found under real world conditions that the BioClock device
currently has a FTE of 1:100 depending on population.
|
 |
|
What is your FAR/FRR (ERR)?
|
|
In a database of 75,000 finger templates the FAR/FRR (EER) is between
.1% to .01%. We like to quote based on EER rather than FAR or FRR rates
due to the better correlation of real world conditions.
|
 |
|
What changes can be made to the unit to improve the EER?
|
|
The match score can be changed from a low of 60 to a high of 100. At 60
the match score is set at normal, this is the optimal value for EER. Changing
this setting does not effect the look up times, but does effect how tolerant
the finger placement on the sensor can be. With a setting of 60 the placement
of the finger can have up to a 5% movement from left to right, and a 5%
movement from top to bottom from the original enrollment.
To raise
the settings to a higher setting the finger placement is less tolerant
to movement or placement (up, down, left, right). For a registered user
to be found on a system set above normal, the placement of the finger
has to be very close to the placement during enrollment. Because placement
is more stringent it is harder to achieve an acceptable match score. This
will lead to a higher FAR.
|
 |
|
What is the maximum number of finger scans that can pass through the unit in one day?
|
|
There is no limit to the number of finger scans that can be processed
in a single day. As long as there is room to store the punch data. Present
hardware allows for 5 gigabytes of storage space. Their would be a tremendous
communications overhead to process this many punches.
|
 |
|
What is the maximum number of finger scans that can be stored in the unit?
|
|
At present size, 125 punches can be stored in 90k of space. Current units
have the ability to store 6.9 million punches. However this would be equivalent
to having 1.7 million employees punching 4 times per day.
|
 |
|
What is the maximum number of templates that can be stored in the unit?
|
|
Leveraging the same 5 gigabytes of storage space used above. At present
size (1.5k), 1000 templates can be stored in 1.5 megabytes of space. At
present the algorithm can reliably perform a 1 to many lookup on 1 to
120 templates. So while the unit can store a very large number of templates
this would not be practical under the present algorithm.
|
 |
|
How many fingers can be enrolled for each employee?
|
|
All 10 fingers on an employee can be stored. Which would result in 10
templates per person.
|
 |
|
How does the presence of grease, grime, dirt, moisture affect the accuracy in scanning the finger?
|
|
The finger should be wiped of all surface grease, grime, dirt, moisture
prior to being placed on the finger sensor. The finger does not have to
be perfectly clean in order for a quality image to be produced.
" Grease/Moisture
- will distort the image beyond reliably quality.
" Dirt/Grime
- could distort the image beyond reliably quality depending on the type
and where it is present in the finger prints pattern.
|
 |
|
Can your unit operate between -28 to 140 degrees?
|
|
Standard electronic components operate between (0C-43C) (32F - 110F) degrees,
maximum relative humidity, non-condensing. Given the extensive use of
standard electronic components used within the units this would be the
proper operating environment.
|
 |
|
How often do the units require maintenance?
|
|
The cooling system of the units should be cleaned once every 6 months.
This is very simple and can be performed in approximately 10 minutes per
unit. This requires the machine to be out of service for this amount of
time. The procedure entails replacing the fan air filters and blowing
out any extra accumulation of dust and debris. Additional fan air filters
can be purchased from your sales representative.
|
 |
|
What is your established mean time between failure?
|
|
Present field testing has proven no maintenance is required prior to 8
months after first installation. At that time very minor maintenance was
performed. I.e. dust removal, fan motor check, power supply check. At
present we have not had to replace any finger print sensors.
|
 |
|
Do you currently have an off-the self software package available that
is marketed with the unit? Does that software cover enrollment, storage
of templates at the device, storage of finger scans at the device? Can
that software operate in a networked environment? Does that software include
audit/control reports?
|
|
Installation included our CheckPrint TA software. Which includes finger
print enrollment, storage, backup, communications, edit of employee punches,
audit trail records,, export to payroll packages, 20 standard reports
with the ability to sub query most of these reports, job costing, 15 user
defined custom fields, scheduling. The main principle of CheckPrint TA
is to perform Time and Attendance functions. It does not hold payroll
or human resources information.
At present
for security reasons CheckPrint TA was developed with a single administrator
in mind. Additional security and network ability is being added in the
next several versions. Additional audit/control reporting will be added
in these releases.
|
 |
|
Does your unit currently have the CE Mark certification? If not, are you
working on securing this certification? When do you anticipate having
this certification complete? Are there any attributes to your hardware/software
solution that would result in you not being approved for the CE Mark certification?
|
|
Because of the nature of the biometric industry our units have not been
UL or CE certified. Once a final design is set in place we fully intend
on acquiring both certifications and do not anticipate any obstacles relating
to hardware or software.
|
| Back to TOP |