DOCUMENT
Attachment 2 Questions and Answers from OPS eIndustry Day.pdf
OVERVIEW
Original Source
Contract Opportunity
Related Opportunity
Related Agency
Posted
Aug. 13, 2020
Type
.pdf
Size
0.41MB
DOCUMENT PREVIEW
EXTRACTED TEXT
OPS RFI and Industry Day: TSA25-04-03475
Amendment 0001 8/13/2020
Attachment 2- Questions and Answers from OPS e-Industry Day
OPS e-Industry Day Questions & Answers
This is a list of questions and answers from the OPS e-Industry Day on 7/29. This information is
intended to provide some clarity to interested parties, but everything in this document is subject to
change at any time.
Footwear Screening
o Q: Where did 360 passengers/hour throughput requirement come from?
A: The 360 passenger/hour goal is based on keeping up with the CTs used for accessible
property screening under our current CONOPS.
o Q: Will there be a separate Classified Appendix for Shoe Scanners?
A: Expected detection levels for footwear screening are provided in the RFI. Classified and
Sensitive Security Information discussions may happen with vendors as needed.
o Q: Did DHS S&T conduct a study on what footwear was at checkpoint? Could this be shared with
vendors/industry?
A: Yes. DHS S&T conducted an observational study on types of footwear typically worn
by the traveling public at U.S. airports. Surveys were conducted at six (6) different airports
and were structured to account for time of day/week variability (business vs. leisure
travelers) and seasonal or weather variations. DHS S&T and TSA will coordinate to share
results from that study with industry.
o Q: Can the Shoe Scanner be integrated into the base of existing AIT systems?
A: This would require collaboration with current AIT system or next-generation OPS
systems developers, but it is possible. For a raised base, a ramp is needed to prevent a
tripping hazard.
o Q: What would the height limitation of an AIT and shoe scanner combination be?
A: Although there will not be any initial limitations to the height of the system, limitations
such as ensuring there is enough room for maintenance will likely be listed in a future FRD.
Size, weight, and power restrictions for systems are typically defined by functional
requirements documents. Generally, body scanners must be capable of scanning individuals
up to 77.4 inches in height and the system must be no taller than 118 inches.
o Q: Does the screening need to go beyond just explosives detection? Also, is there a material
discrimination need here?
A: A footwear screening system will require anomaly detection similar to AITs. Material
discrimination is encouraged for the reasons defined in the material discrimination
objective area, but not required.
o Q: Is there any information from the existing AIT that would help the performance of the footwear
scanner and would it be available?
Current AITs currently do not scan footwear.
Material Discrimination
o Q: Does the technology need to localize threat? Can it say yes/no on the whole body?
A: The intention is to identify a material category for an anomaly detected by body
scanner/AIT, so localization is necessary.
o Q: Will there be requirements for classification accuracy? For instance, the machine detects an
anomaly and material discrimination suggests it is paper, but in reality, it is paper?
A: Yes, requirements will have to include accuracy. TSA is currently working through the
specifics and welcomes input from industry.
o Q: If the solution were to use some of the AIT data, or want to access the AIT raw data would that
be available, does TSA own the data from the AIT?
A: TSA is working with AIT system original equipment manufacturers (OEMs) to
determine what raw data can be shared or would be necessary to support material
OPS RFI and Industry Day: TSA25-04-03475
Amendment 0001 8/13/2020
Attachment 2- Questions and Answers from OPS e-Industry Day
discrimination solutions. TSA asks that RFI respondents specify in the RFI what data is
sought.
o Q: Is a solution that provides non-contact material discrimination as part of anomaly resolution (vs.
primary screening) acceptable?
A: The On-Person Screening capability focuses on primary screening, but this may be
useful for the Alarm Resolution Capability Manager Team.
o Q: Could the non-contact material discrimination sensor be positioned automatically by cooperative
robotics in proximity to the anomaly?
A: TSA is open to the idea of utilizing robotics to move a sensor to a body part where an
anomaly is detected; however, such processes cannot slow down the checkpoint line. Alarm
resolution may also be interested in this technology.
Enhanced Detection Performance and Throughput
o Q: Does TSA know what drove the requirement for passengers to stand with their hands above their
heads in the current system?
A: A more spread-out body position helps the system scan the body and identify anomalies.
TSA is interested in new solutions that do not impede detection, but makes the screening
process more comfortable for a wider range of the traveling public.
o Q: Is TSA considering on-screen Alarm Resolution like CT scanners, given the image quality were
A: Though the TSA is interest in on-screen alarm resolution and open to ideas on how this
could work, it cannot function like CT scanners. Due to privacy concerns, on-person
screening equipment must identify anomalies with an automated detection algorithm and
display them on a generic avatar, unlike accessible property screening systems.
o Q: The false alarm triggered by the incorrect arm position is driven by the detection model or the
good enough?
hardware?
A: As long as the body is inside of the systems scanning envelop, this occurrence is likely
triggered by software.
o Q: Can raw data from the AITs be made available?
A: Currently, there is a database of images used for the DHS S&T Kaggle competition
from a millimeter wave (mmW) system. These images can be made available to vendors
that are working on anomaly detection algorithms. The images (not raw data) seemed to
suffice for competition; therefore, TSA did not intend to share any raw data. If other types
of data (other than images) would be helpful, please provide that information in the RFI
response.
Improved Data Visualization
o Q: Is there a specific timeline for common graphical user interface (CGUI)?
A: Due to COVID-19, testing and data collection in the field has been delayed. TSA is
looking to have a CGUI design guide that influences CGUI development in approximately
one (1) year.
o Q: Is the AIT data in a proprietary format or in a Digital Imaging and Communications in Security
(DICOS)-like open format?
A: TSA is aiming to move towards supporting Open Architecture systems, which may
include the use of DICOS imaging format. However, certain aspects of the AIT data will
remain proprietary to its vendors. TSA has images available from the millimeter wave On-
Person Screening system that was used as part of the machine learning Kaggle competition.
These images can be made available to vendors on an as needed basis. TSA also welcomes
collaboration of third-party vendors with AIT OEMs.
o Q: What are TSAs thoughts on how visualization display must take screening on-the-move into
account?
OPS RFI and Industry Day: TSA25-04-03475
Amendment 0001 8/13/2020
Attachment 2- Questions and Answers from OPS e-Industry Day
A: TSA is looking into how Automatic Threat Recognition (ATR) could be applied to a
camera-based system; a passengers moving should not eliminate the possibility of having
an avatar of the passenger. Processing time and tracking of passenger profiles is important
to TSA and will require further collaboration with industry to ensure screening on-the-
move is a feasible capability at airports.
o Q: We are aware of the privacy concerns regarding sensitive detection imagery of passengers. Are
there any guidelines as to how much is too much detail?
A: Systems and visualization processes utilized by TSA should not display any anatomical
detail. Detail for anomalies (non-anatomical detail) can be shown.
o Q: Would a Common Workstation require Open Architecture?
A: TSA is moving towards an Open Architecture system on a whole. At the moment, a
Common Workstation or CGUI would not require Open Architecture, but with Open
Architecture as an end goal, TSA may eventually begin requiring specific vendors to
employ a specific CGUI.
o Q: Would this concept include alarm resolution tools, such as handheld mmW for alarm resolution?
A: The RCA Capability Manager for Alarm Resolution, and team are addressing Alarm
Resolution, which will likely be a more optimal avenue for the handheld mmW system.
Amendment 0001 8/13/2020
Attachment 2- Questions and Answers from OPS e-Industry Day
OPS e-Industry Day Questions & Answers
This is a list of questions and answers from the OPS e-Industry Day on 7/29. This information is
intended to provide some clarity to interested parties, but everything in this document is subject to
change at any time.
Footwear Screening
o Q: Where did 360 passengers/hour throughput requirement come from?
A: The 360 passenger/hour goal is based on keeping up with the CTs used for accessible
property screening under our current CONOPS.
o Q: Will there be a separate Classified Appendix for Shoe Scanners?
A: Expected detection levels for footwear screening are provided in the RFI. Classified and
Sensitive Security Information discussions may happen with vendors as needed.
o Q: Did DHS S&T conduct a study on what footwear was at checkpoint? Could this be shared with
vendors/industry?
A: Yes. DHS S&T conducted an observational study on types of footwear typically worn
by the traveling public at U.S. airports. Surveys were conducted at six (6) different airports
and were structured to account for time of day/week variability (business vs. leisure
travelers) and seasonal or weather variations. DHS S&T and TSA will coordinate to share
results from that study with industry.
o Q: Can the Shoe Scanner be integrated into the base of existing AIT systems?
A: This would require collaboration with current AIT system or next-generation OPS
systems developers, but it is possible. For a raised base, a ramp is needed to prevent a
tripping hazard.
o Q: What would the height limitation of an AIT and shoe scanner combination be?
A: Although there will not be any initial limitations to the height of the system, limitations
such as ensuring there is enough room for maintenance will likely be listed in a future FRD.
Size, weight, and power restrictions for systems are typically defined by functional
requirements documents. Generally, body scanners must be capable of scanning individuals
up to 77.4 inches in height and the system must be no taller than 118 inches.
o Q: Does the screening need to go beyond just explosives detection? Also, is there a material
discrimination need here?
A: A footwear screening system will require anomaly detection similar to AITs. Material
discrimination is encouraged for the reasons defined in the material discrimination
objective area, but not required.
o Q: Is there any information from the existing AIT that would help the performance of the footwear
scanner and would it be available?
Current AITs currently do not scan footwear.
Material Discrimination
o Q: Does the technology need to localize threat? Can it say yes/no on the whole body?
A: The intention is to identify a material category for an anomaly detected by body
scanner/AIT, so localization is necessary.
o Q: Will there be requirements for classification accuracy? For instance, the machine detects an
anomaly and material discrimination suggests it is paper, but in reality, it is paper?
A: Yes, requirements will have to include accuracy. TSA is currently working through the
specifics and welcomes input from industry.
o Q: If the solution were to use some of the AIT data, or want to access the AIT raw data would that
be available, does TSA own the data from the AIT?
A: TSA is working with AIT system original equipment manufacturers (OEMs) to
determine what raw data can be shared or would be necessary to support material
OPS RFI and Industry Day: TSA25-04-03475
Amendment 0001 8/13/2020
Attachment 2- Questions and Answers from OPS e-Industry Day
discrimination solutions. TSA asks that RFI respondents specify in the RFI what data is
sought.
o Q: Is a solution that provides non-contact material discrimination as part of anomaly resolution (vs.
primary screening) acceptable?
A: The On-Person Screening capability focuses on primary screening, but this may be
useful for the Alarm Resolution Capability Manager Team.
o Q: Could the non-contact material discrimination sensor be positioned automatically by cooperative
robotics in proximity to the anomaly?
A: TSA is open to the idea of utilizing robotics to move a sensor to a body part where an
anomaly is detected; however, such processes cannot slow down the checkpoint line. Alarm
resolution may also be interested in this technology.
Enhanced Detection Performance and Throughput
o Q: Does TSA know what drove the requirement for passengers to stand with their hands above their
heads in the current system?
A: A more spread-out body position helps the system scan the body and identify anomalies.
TSA is interested in new solutions that do not impede detection, but makes the screening
process more comfortable for a wider range of the traveling public.
o Q: Is TSA considering on-screen Alarm Resolution like CT scanners, given the image quality were
A: Though the TSA is interest in on-screen alarm resolution and open to ideas on how this
could work, it cannot function like CT scanners. Due to privacy concerns, on-person
screening equipment must identify anomalies with an automated detection algorithm and
display them on a generic avatar, unlike accessible property screening systems.
o Q: The false alarm triggered by the incorrect arm position is driven by the detection model or the
good enough?
hardware?
A: As long as the body is inside of the systems scanning envelop, this occurrence is likely
triggered by software.
o Q: Can raw data from the AITs be made available?
A: Currently, there is a database of images used for the DHS S&T Kaggle competition
from a millimeter wave (mmW) system. These images can be made available to vendors
that are working on anomaly detection algorithms. The images (not raw data) seemed to
suffice for competition; therefore, TSA did not intend to share any raw data. If other types
of data (other than images) would be helpful, please provide that information in the RFI
response.
Improved Data Visualization
o Q: Is there a specific timeline for common graphical user interface (CGUI)?
A: Due to COVID-19, testing and data collection in the field has been delayed. TSA is
looking to have a CGUI design guide that influences CGUI development in approximately
one (1) year.
o Q: Is the AIT data in a proprietary format or in a Digital Imaging and Communications in Security
(DICOS)-like open format?
A: TSA is aiming to move towards supporting Open Architecture systems, which may
include the use of DICOS imaging format. However, certain aspects of the AIT data will
remain proprietary to its vendors. TSA has images available from the millimeter wave On-
Person Screening system that was used as part of the machine learning Kaggle competition.
These images can be made available to vendors on an as needed basis. TSA also welcomes
collaboration of third-party vendors with AIT OEMs.
o Q: What are TSAs thoughts on how visualization display must take screening on-the-move into
account?
OPS RFI and Industry Day: TSA25-04-03475
Amendment 0001 8/13/2020
Attachment 2- Questions and Answers from OPS e-Industry Day
A: TSA is looking into how Automatic Threat Recognition (ATR) could be applied to a
camera-based system; a passengers moving should not eliminate the possibility of having
an avatar of the passenger. Processing time and tracking of passenger profiles is important
to TSA and will require further collaboration with industry to ensure screening on-the-
move is a feasible capability at airports.
o Q: We are aware of the privacy concerns regarding sensitive detection imagery of passengers. Are
there any guidelines as to how much is too much detail?
A: Systems and visualization processes utilized by TSA should not display any anatomical
detail. Detail for anomalies (non-anatomical detail) can be shown.
o Q: Would a Common Workstation require Open Architecture?
A: TSA is moving towards an Open Architecture system on a whole. At the moment, a
Common Workstation or CGUI would not require Open Architecture, but with Open
Architecture as an end goal, TSA may eventually begin requiring specific vendors to
employ a specific CGUI.
o Q: Would this concept include alarm resolution tools, such as handheld mmW for alarm resolution?
A: The RCA Capability Manager for Alarm Resolution, and team are addressing Alarm
Resolution, which will likely be a more optimal avenue for the handheld mmW system.
Show All