Mr. William R. Geer
Sandia National Laboratories
Media Relations and Communications
PO Box 5800, MS 0165
Albuquerque, NM 87185-0165
Phone: 505-844-6601
Fax: 505-844-0645
E-mail: wrgeer@sandia.gov
Number of Human Subjects projects reported: 12
| SNL-00-17 | "Facial Verification System Evaluation" |
| SNL-01-16 | "Adaptive Awareness for Personal and Small Group Decision Making" |
| SNL-02-03 | "Populating Computational Models of Cognition with Knowledge Elicitation Data to Detect Man-Machine State Discrepancies: Knowing and Responding When the Human and Machine Are Not on the Same Page" |
| SNL-02-05 | "Acoustic Biosensors for Bacteria Detection " |
| SNL-02-13 | "Evaluation of Toolkit for Recapature / Recovery Project" |
| SNL-02-15 | "The Relationship between Sleep and Breathing" |
| SNL-02-18 | "Flow in Organizations: Experience at Work" |
| SNL-03-01 | "FY03-FY05 Micro Optical Radar (MOR) Facial Recognition" |
| SNL-03-0221 | "Integrated Microfluidic System for Oral Diagnostics" |
| SNL-03-03 | "Effectiveness of Augmented Cognition Technology as a Decision Tool: A Validation Study " |
| SNL-03-04 | "FY04 Facial Image Collection Database" |
| SNL-03-14 | "Enabling Technology for Human Collaboration" |
"Facial Verification System Evaluation"
Principal Investigator: Mr. Larry J. Wright, Sandia National Laboratories
Project started in: 2000
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 07/25/03
IRB approval number: N/A
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 50
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
Facial verification systems use distinguishing characteristics of the face to verify a person's identity in automated entry control applications. An image of the face is captured using a video camera, and distinguishing features are extracted and compared with previously stored features. If the two match within a specified tolerance, a positive identity verification results and the person is allowed to enter the facility.
Although facial systems have been proposed and studied for a number of years, commercial systems have only been available for the last few years. Developers have had to overcome two difficult problems: (1) wide variations in the presentation of the face (e.g., head tilt and rotation, presence or absence of glasses, facial hair changes, facial expression changes) and (2) lighting variations (e.g., day vs. night, location A vs. location B).
The purpose of the test is to determine the performance of the Lau Technologies Portal, a commercial face verification system configured for physical access control. Performance will be characterized in terms of false reject rate, false accept rate, and enrollment and verification times. In addition, user acceptance of the device will be measured by administering a survey to the test participants. Other important characteristics to be observed during the test include ease of setup, resistance to spoofing (attempts to defeat the system), template storage requirements, and system cost. As configured for the test, the system consists of a face terminal, a processor/controller for the portal, and controlled lighting.
Test subjects will be recruited from employees and contractors at Sandia. Although there are no inclusion/exclusion criteria, recruitment of subjects will be limited to a convenient sample of personnel who work in Building 821 or regularly travel through that facility. Participants will be asked to provide their height and whether or not they wear glasses. This information is required to understand how these factors might impact the performance of the system.
There will be no financial compensation for participating in the study, and there are no known health or personal safety risks associated with use of the facial verification system. The activities involved in this study are similar to using an automated teller machine.
To protect the privacy of subjects and the confidentiality of the data, only test project personnel will have access to information that ties image and performance data to particular test subjects, and computer database files will be password protected. The results of the study, may be published for scientific purposes, but such results will be limited to summary data only and will not provide name, picture, or any identifiable references to subjects. However, any records or data obtained in this study may be inspected by the sponsor, by any relevant governmental agency (e.g., DOE), by the Sandia Human Studies Board, or by the persons conducting this study, provided that such inspectors are legally obligated to protect any identifiable information from public disclosure, except as otherwise authorized or required by law.
"Adaptive Awareness for Personal and Small Group Decision Making"
Principal Investigator: Dr. Carmen M. Pancerella, Sandia National Laboratories
Project started in: 2001
This project ended in fiscal year 2003.
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 09/04/02
IRB approval number: N/A
Explanation of IRB approval:
IRB approval was valid until 9/3/2003.
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 2
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
Objectives:
Many situations call for the use of sensors to monitor physiological and environmental data. In these situations, it is beneficial to have intelligent agents analyze the large amounts of sensor data, recognize cues from the data, and communicate the results to humans and other computers. This project will pilot an awareness and warning tool, comprised of heterogeneous sensors, small light-weight processors, embedded intelligent software, and a wireless network connecting these processors and computers. This tool has broad applicability to emergency teams, military squads, individual exercise and fitness monitoring, health monitor for sick and elderly, and environmental monitoring in public places.
Methodology:
To realize this tool, researchers propose to develop a sensor-independent architecture for integrating self-awareness information across a distributed team (e.g., military squads, approximately 10 soldiers working together in a relatively small geographical distance) and the design approaches, concepts, and software agents for a tool that supports squad-level decision making.
Involvement of Human Subjects:
Actual testing of the system with humans will be limited to ensuring that the sensor is functioning correctly. All intelligent algorithm training and testing will be done with simulated data. As such, the time that the sensor will be attached to subjects will be short, perhaps five minutes or less, as required for the software/hardware development and testing. The subject will attach the sensor as required to determine that the sensor interface is correctly transmitting the data. However, no personally identifiable information will be collected on any subject. Actual personal sensor data will not be archived in the system.
The risks associated with this study are no greater than those encountered in daily life (minimal).
"Populating Computational Models of Cognition with Knowledge Elicitation Data to Detect Man-Machine State Discrepancies: Knowing and Responding When the Human and Machine Are Not on the Same Page"
Principal Investigator: Dr. James C. Forsythe, Sandia National Laboratories
Project started in: 2002
This project ended in fiscal year 2003.
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 02/13/02
Explanation of IRB approval:
IRB approval was valid until 2/12/2003.
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 2
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
The purpose of this research is to improve human decision support in high-performance systems through the development of computational models of cognition. A relatively novel technique that will be applied to improve the fidelity of the computational models is a knowledge elicitation methodology developed at Sandia National Laboratories. Using a protocol of semi-structured interviews and mathematical techniques to derive associative networks, computational models will be populated with more accurate and cognitively plausible representations. Simulations of the constructed model will be compared with human performance data to examine ways of detecting man-machine state discrepancies. The anticipated outcome is a cognitive model that has a closer understanding of the user that can support decision-making in complex, time-pressured domains.
Two to four student interns will participate in this study and will be trained in an Airborne Warning and Control System (AWACS) operator-like task environment. All data collected associated with the participants (e.g., video, computer data, consent forms etc.) will be stored in a locked cabinet and marked as unclassified controlled information (UCI). The interface for this task will be a personal computer. Training will consist of 10 to 15 hours/week of self-paced practice with the task for four to six weeks. Using standard performance measures associated with AWACS operation, the objective is to raise expertise to a stable level of performance. At this point, they will be justifiably categorized as subject matter experts (SMEs).
Using information gathered from literature and semi-structured interviews with the SMEs, a cognitive model will be developed representative of SME AWACS performance. The technological objective for this phase is to establish a capability within the computer to detect when operator's mental models are inconsistent or diverge from the known state of the world. The SME will be utilized to test prototypes of this technology as the technology is refined. Test sessions will be self-paced and the duration will be one to two hours. Test sessions will consist of individual trials lasting 10 to 15 minutes. Trials will involve unique vignettes in which the participant must manage a squad of airborne assets against differing configurations of threats.
For this phase of the study, each SME will participate in two to four iterations of this phase over a period of two weeks. It is anticipated that subjects will each participate in four to six studies of this nature, with studies occurring two to four weeks apart. Each test session will be recorded using video. Afterward, test sessions will be reviewed to identify instances in which subject behavior implied an inconsistency between their mental model and the actual state of the world. These data will then provide a basis for evaluating technical system performance and to derive experimental variables for future research.
The potential risks for participants in this study are similar to those associated with moderate computer work and face-to-face meetings.
"Acoustic Biosensors for Bacteria Detection"
Principal Investigator: Dr. Susan M. Brozik, Sandia National Laboratories
Project started in: 2002
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 01/07/03
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 0
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
The purpose of this cooperative research agreement between SNL and Minnesota Manufacturing and Mining Company (3M) is to develop a commercially viable prototype of an acoustic biosensor for detecting low levels of bacteria in near real-time. This biosensor will be based on acoustic sensing technology developed by Sandia and will include a handheld electronic device and a disposable fluidic cartridge that houses the sensor transducer.
Since a critical need exists for rapid, accurate testing for biological warfare agents, the overall benefit of this project is the development of biosensors for government/national use in detecting anthrax. Secondly, the technology developed will then be applied toward the development of a sensor for the detection of nosocomial infections, a critical need in the health care industry.
Phase 1 (Feasibility) – During this phase, Sandia and 3M will ascertain if the shear-horizontal surface acoustic wave (SH-SAW) biosensor is capable of measuring less than 2,000 spores (model of anthrax) per milliliter of fluid in representative samples. Antibodies raised against anthrax spores will be attached to the sensor platform as part of a recognition layer. To test the response of the sensor, B. thuringensis and B. subtilis spores will be used as anthrax simulants. No testing of anthrax spores will take place at Sandia or 3M, nor will any personnel from either establishment handle anthrax spores. Spore samples in water will be introduced to the sensor platform via a microfluidic chamber. For “real-world” samples, a “contaminated” sample will be introduced to the sensor. The “contaminated” sample will be a simulant spore in a solution of protein, dust, or other similar contaminant. We do not anticipate using nasal samples from humans in Phase 1.
Phase 2 (Development) – The goal of this phase is a prototype of the SH-SAW sensor system appropriate for clinical testing. During this phase, SNL will continue developing an anthrax detector using simulant spores. In addition, Sandia will begin developing the SH-SAW platform for he detection of Staphylococcus aureus (leading cause of nosocomial infection). As in the above experiments, antibodies and/or other receptors will be attached to the sensor platform as a recognition film. A non-pathogenic strain of Staph aureus will be used to test the device. Initially the bacteria will be introduced to the sensor in a water or buffer sample. SNL will then test “real-world” samples consisting of nasal swabs obtained from human subjects.
Protocol for obtaining nasal swabs:
The subjects’ anterior nares will be swabbed with a sterile rayon swab. Two swabs will be taken from each subject; one from each nostril. The sampling will be performed by inserting the rayon swab into the anterior tip of the subject’s nostril and rotating the swab for six complete revolutions along the nares mucosal surface. The swab will then be placed into a sterile test tube containing one milliliter of room temperature Buffered Peptone Water (BPW; Hardy Diagnostics) and mixed vigorously for 30 seconds using a standard lab vortexer. A known concentration of the test bacteria will be added to some of the samples. The test sample will then be introduced to the sensor device for detection of Staph aureus (non-pathogenic strain).
Subject Recruitment:
Because this is a feasibility and prototype study, the subject pool is limited to male and female members of the research team who are in good general health (as self-identified) and have no physical disabilities that would increase risk of harm. Subjects will be recruited verbally. 3M will conduct similar experiments at their facility, recruiting subjects from their personnel who are working on this project. Involvement of those subjects will be reviewed by the 3M Institutional Review Board (IRB), and a copy of that IRB’s approval will be provided to Sandia’s Human Studies Board).
Timeline:
The duration of this project is 27 months with Phase 2 (human testing) to begin after the first three months of the project. Sample collection of nasal swabs from human subjects will take less than a minute, and collections will occur once or twice a week over a period of two to three months.
"Evaluation of Toolkit for Recapature / Recovery Project"
Principal Investigator: Mr. James E. Pacheco, Sandia National Laboratories
Project started in: 2002
This project ended in fiscal year 2003.
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 03/27/03
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 2
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
In the event that an asset falls under the control of hostile forces, measures must be taken to regain control of it as soon as sufficient forces are on hand to optimize the chances of recovery with minimal loss of life and property. The actions required to regain control of nuclear resources are known as recapture/recovery operations. Sandia National Laboratories (SNL) has identified technologies applicable to recapture/recovery operations, and is currently working with our military partner to assess current recapture/recovery capabilities for a typical military base. A “toolkit” has been proposed to recapture facilities that consists of thermal, mechanical, and explosive tools. Most of the breaching tools proposed in the toolkit have been tested and documented by SNL and outside vendors over the past 25 years to breach similar types of surfaces.
The purpose of this project is to develop a facility-specific toolkit to assist in recapture operations. The toolkit contains mechanical and thermal breaching tools and other equipment to assist in the operation. This project will evaluate the effectiveness of these tools by conducting timed tests using SNL employees and contractor personnel. No personally identifiable information will be collected.
Hazards associated with testing and using these tools include: spattering melted steel, hot spots, sparks, loud noise, sharp edges, and airborne dust and dirt. Personnel conducting the tests will be trained in the safe use of the tools being tested, and will be required to use appropriate safety equipment during these tests.
The first phase effort will evaluate and test existing non-explosive entry methods on mock-up personnel and vehicle access doors at SNL.
No explosive tests are planned for this phase (FY02-03), which will measure cutting rates of these tools to breach various door and wall surfaces. The performance of individual operators will not be measured. One to three subjects will actually use the tools.
These tests are expected to require minimal physical exertion by the participants. This assumption is based on the estimated effort required to make one to five cuts through each proposed barrier with the tools proposed. There will be breaks between the series of cuts to allow subjects to rest and for the PI to collect data and make notes. Subjects will be informed that even though the tests are timed, the objective is to measure the rate at which the tools cut through various surfaces under normal operating conditions. Subjects will also be encouraged to work at a comfortable rate and to stop if they feel stressed. Heat stress associated with wearing personal protective equipment during the summer will be addressed in the preliminary hazard assessment, safe operating procedures, or other Environment, Safety and Health (ES&H) documentation. To minimize the heat stress on subjects wearing personal protective equipment, we anticipate using cooling fans, installing temporary shading, or conducting the tests inside when possible. The tests will be videotaped, photographed, and documented, and these materials may be used in future training sessions.
Subjects may be female or male, in good general physical health, as self-identified, and have no physical disabilities that would increase risk of harm. The minimum age to participate in the study will be 18 years old. Subjects who have experience conducting similar tests or experience with the safe operation of thermal cutting tools, hand or power tools will be given preference. The research team will describe the project to all potential subjects and have them read and sign a standard form indicating their consent to participate. Subjects will be encouraged to ask questions throughout the study and will be reminded that participation is voluntary.
The risks associated with this study are the hazards caused by the use of the tools, which include spattering melted metal, hot sparks, hot spots, loud noise, air-borne dust/dirt, and sharp edges. However, the tests are considered to be minimum risk because the participants will be trained in the use of the tools being tested and will be provided with, and required to use, appropriate safety equipment. The tools will be used in the manner they were designed - as they are used in standard industrial procedures. Any new information developed during the study that may affect a participant’s willingness to continue participation will be communicated to the participants.
There is no financial compensation or other benefit for participation in this study, other than personal satisfaction from helping to advance knowledge in this specific area. Our military partner (as well as DOE, DoD, and other agencies interested in recapture capabilities) will benefit by having confidence in the effectiveness of the toolkit.
Data will be stored on electronic spreadsheets in Excel and Word documents. The computer storing these files is password controlled and inside a secured area, and access to the database will be restricted to project personnel. All personally identifiable information (if any) will be stored in a filing cabinet in Building 821, room 3094. At the end of the study, results will be documented in a Sandia report. All the data and final information will be retained by the project leader for a minimum of three years after completion of the study. This study will not produce sensitive data nor is any part of this study classified.
"The Relationship between Sleep and Breathing"
Principal Investigator: Dr. Barry J. Krakow, Sleep and Human Health Institute (SHHI)
Project started in: 2002
Funding for Human Subjects Research: No Funding Sources Reported
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 05/07/03
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 37
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
Objectives:
This study will investigate the use of nasal strips in human subjects with insomnia and sleep-disordered breathing (SDB)symptoms or disorders in a randomized controlled design. It proposes to determine whether the nasal strips, without other treatment, can decrease or eliminate insomnia symptoms.
Potential Benefits:
Entering this protocol is likely to educate the subject on the value of treating insomnia, so it is not likely to delay other treatments that the subject might wish to pursue for sleep problems. Subjects may benefit directly by discovering that nasal strip treatment of sleep disordered breathing (SDB) leads to improved sleep, and their motivation for further treatment of sleep disorders may be increased. There is no financial compensation for participation.
Methodology:
The current proposal will be a randomized controlled trial studying sleep maintenance insomniacs with SDB. Two groups will be studied to contrast the effects of nasal strips versus a control group. The nasal strip (treatment) group will be educated on the relationship between SDB and insomnia and will be instructed in the use of nasal strips. The control group will monitor their sleep and breathing symptoms only. Both groups will prospectively monitor changes in breathing and sleep during a four-week interval, with the option to continue for an additional four to eight weeks. All wait-list control participants will be invited to participate in treatment at the end of their four-week control period.
Involvement of human subjects:
This project will require the direct (in vivo) involvement of human subjects in all aspects of the protocol. These subjects will be recruited from the general population at SNL.
Potential Risks:
Potential physical risks associated with nasal strip use are minimal and include irritation of the skin, either through allergic reaction to the material (fairly rare) to rawness from improper removal of the strips. If subjects are known to have allergies to latex, this may trigger the allergic reaction because latex is included in the packaging of the strips, but not on the strips directly. Such individuals can actually use the strips if they can learn how to remove the strips from the packaging without getting latex on their fingers. Removal of strips requires emphasis on water and heat (preferably a shower or warm, wet washcloth) to break the adhesive seal on the skin, followed by gentle tugging on the strip to remove gradually.
Another potential non-trivial risk is breach of confidentiality; however, several precautions are incorporated into the study design to minimize this risk. All data will be collected by staff at the Sleep and Human Health Institute (SHHI) in Albuquerque. Hard copy data (including questionnaires and email reports) will contain personal identifiers and will be stored in locked file cabinets after the data has been entered into a coded database (stored in password-protected computers). Current SHHI computers are operated through the University of New Mexico (UNM) School of Medicine facilities, which have appropriate firewall protection to prevent unauthorized access from anyone outside UNM. Access to the database will be limited to appropriate SHHI staff. No part of this study is classified, and the data is sensitive to a degree similar to any medical record.
"Flow in Organizations: Experience at Work"
Principal Investigator: Dr. Ryan W. Quinn, University of Michigan
Project started in: 2002
This project ended in fiscal year 2003.
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 09/18/02
Explanation of IRB approval:
IRB approval was valid until 9/17/2003.
Additional IRB approvals from other institutions:
Type of Review: Expedited
Approving Institution: University of Michigan Behavioral Sciences IRB
Most recent approval: 05/09/02
IRB approval number: IRB file #7091
Explanation of additional approval:
IRB approval was valid until 5/9/2003.
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 121
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
One purpose of this study is to learn how SNL engineers can achieve a state of optimal experience in their work that psychologists call “flow.” This is a state in which people both perform their best and feel their best. Interviews with engineers, scientists, and technicians revealed that most technologically-oriented knowledge workers experience flow at work and would like to experience it more often. Some of the results of this study will help individuals focus on things they can do to achieve a flow state more often in their work. A second purpose of this study is to help identify and thus reduce the amount of time that SNL personnel spend on unproductive activities, such as administrative paperwork or bureaucratic procedures.
The approach to studying engineers’ high performance experiences that this will use is called the Experience Sampling Method (ESM). Subjects participating in an ESM study receive a signaling device (e.g., a beeper) and a survey booklet at the beginning of a five-day workweek. During each day of the week, the signaling device will emit a signal at four random times. After the participants hear the signal, they either open the survey book and fill out a short survey, or (if the signal went off at a particularly inopportune time) they complete the activity they were engaged in and then open the survey book and fill out the survey. Previous ESM surveys of a similar length have taken respondents three to four minutes to complete. Thus, this study will collect data on engineers’ experiences in 20 randomly selected work activities.
The intent is to recruit 120 subjects from respondents to a e-mail message sent to randomly selected Sandia personnel. Participation is completely voluntary, and personally identifiable information will be retained only until the beepers are returned. At that point all data will become untraceable back to a particular subject. There is no financial or other compensation for participation.
This study has been approved by both the Institutional Review Board of the University of Michigan and the Sandia Human Studies Board.
"FY03-FY05 Micro Optical Radar (MOR) Facial Recognition"
Principal Investigator: Mr. Robert D. Habbit, Sandia National Laboratories
Project started in: 2003
Funding for Human Subjects Research: No Funding Sources Reported
This project does not involve the use of multiple protocols/subprojects.
Identifier or number: N/A
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 01/20/03
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 0
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
Objectives:
The driving factor influencing this effort is the national need for an automated, rapid, non-invasive, personal identification system. Currently, there is no automated system capable of identifying known good actors, known bad actors, or unknown actors at public entry points such as border crossings and mass transit terminals. Thus, everyone is a suspect and subject to screening, resulting in precious screening resources spent on would-be known good actors and less screening of would-be known bad actors and unknown actors. Facial recognition is recognized as the least invasive, quickest method for personal identification. Current facial recognition systems use distinguishing physical characteristics (2D photographs) to identity subjects. Unless the exact parameters of the photo are measured and recorded, it is difficult to compare two different photographic acquisitions in different locations. A 3D imaging system would provide all of the attributes of a 2D photograph plus an accurate 3D geometric image. The 3D images provide all of the necessary information to establish “control” of the scene as well as additional information such as distance from the tip of the nose to cheek. It is expected that near 100 percent positive identification of known good actors can be achieved and potentially similar results for known bad actors. This study involves collecting and processing 3D images of humans to advance sensor/algorithm technologies for facial biometric identification applications.
Methodology:
Sandia National Laboratories (SNL) has developed a technology capable of acquiring full-scene 3D geometrically accurate images. The base technology was flown on the Space Shuttle to acquire images and modal response of the International Space Station truss structure and solar panels. Prototype sensors have been built to demonstrate acquisition of 3D images of military targets at distance on land and underwater, but none of these sensors are designed for or are capable of obtaining images with the precision necessary for facial recognition. The electronics could be re-designed to meet the requirements for facial recognition; however, the cost of those electronics would render wide-spread application of the sensors cost prohibitive. This project asserts the electronic functions used in the previous systems could be reconfigured into a solid-state application specific integrated circuit (ASIC), thus, significantly reducing the cost to produce the sensor while fully meeting or exceeding the precision required for facial recognition. In addition, the research team asserts that robust efficient algorithms can be designed to process, identify, and store images in a near real-time environment. The ultimate goal is to place this sensor in public locations where positive identification of both good and bad actors is required. This would enable faster access for known good actors, positive identification of know bad actors and, most importantly, enable more thorough screening of unknown actors.
Ionizing radiation, radioactive substances, or chemical substances: None
Involvement of human subjects:
Procedure:
All testing will be in vivo. The procedure for acquiring 3D range images is identical to acquiring photographs. The only difference is the illumination (light) source, and the receiver is time encoded such that the time of flight of the light can be measured. This measurement allows the calculation of range. The range information coupled with the optical properties of the lens allow one to calculate the exact position in space of the subject, thus, producing an accurate 3D geometric image. 3D imagers use a modulated or pulsed coherent light source. The form of the illumination source can be a laser, laser diode array or light emitting diode (LED) array. In all cases, the illumination levels are below the eye safe levels established by SNL Laser Safety Program and/or the Occupational Safety and Health Administration (OSHA).
During the early phases of the project, both Sandia-developed sensors and commercial laser radar systems may be use to acquire images. These initial images are important for algorithm development. Ultimately, the micro optical radar (MOR) sensor will come to fruition and will be used for image acquisition.
Each Participant will be asked to pose and/or walk through identified areas. A typical acquisition session should take no longer than one-half hour. Like early photography, the subject may be asked to pose for up to one minute during each acquisition. The expected project duration is expected to be three years.
Protection of Data:
To protect the confidentiality of the data, participants will be identified by a number assigned at the initial acquisition and only project personnel with a need-to-know will have access to information that ties image data to a particular test subject’s name. Also, computer database files will be password protected. These protection measures will be explained to all participants.
"Integrated Microfluidic System for Oral Diagnostics"
Principal Investigator: Dr. Anup K. Singh, Sandia National Laboratories
Project started in: 2003
Funding for Human Subjects Research: No Funding Sources Reported
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 12/09/02
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 0
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
Analysis of saliva and other oral fluids has great potential in diagnosis of oral and systemic diseases, in monitoring levels of environmental toxins and drugs-of-abuse, and in preliminary screening for exposure to biological and chemical warfare reagents. The use of saliva and other oral fluids as diagnostic samples may be preferred in many applications over other bodily fluids because of the ability for fast and inexpensive sampling in clinical as well as non-clinical settings. Although, clinical studies have demonstrated usefulness of saliva and other oral fluids in some applications for detection of markers of oral and systemic diseases, the development of diagnostic technology and devices has not met the expectations. Moreover, the current assay techniques are time-consuming, require relatively large amounts of samples, and are not amenable to automation and portability.
The objectives of this application are the following: 1) Develop an integrated microfluidic system for simultaneous multi-analyte detection in saliva and other oral fluids. The prototype will be capable of rapid and sensitive analysis of minute sample volumes and will attain higher selectivity by performing multiplexed analysis. 2) Analysis of mediators of oral and periodontal diseases found in oral fluids (i.e., P. gingivalis endotoxin, interleukins, and bone breakdown fragments) using the proposed device. 3) Validation of the prototype in a longitudinal human trial of patients at low and high risk for the development of periodontal diseases. Success of these objectives should provide for technology transfer and utilization of this device in clinical periodontology as well as potential applications in preliminary screening for systemic diseases and exposure to biological warfare agents.
"Effectiveness of Augmented Cognition Technology as a Decision Tool: A Validation Study"
Principal Investigator: Dr. James C. Forsythe, Sandia National Laboratories
Project started in: 2003
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 05/13/03
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 34
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
This study examines whether an augmented cognition decision tool created by Sandia National Laboratories (SNL) can enhance users’ ability to make accurate decisions about the potential for insider activity within the electronic environment at Sandia. The approach is based on computational cognitive modeling framework developed at SNL, including the premise that an insider will leave behind clues – behaviors that, when seen in conjunction with one another, would alert someone to that person’s suspicious activities. The problem is that these clues are spread out across time, people, and scenarios – no one person is able to gather all of the information and put together the “big picture” of that person’s behaviors. The tool being evaluated allows a user to sift through huge volumes of data to spot these suspicious combinations of behaviors more quickly and help human decision makers understand this big picture.
Participants comprise three distinct groups: subject matter experts (SMEs) who consulted on building the model, SMEs who did not assist in building the model, and SNL personnel who have no knowledge of the model and have no particular expertise in detection and/or prevention of insider crimes. Participants will be asked to make decisions about whether or not there should be some concern about the existence of an insider within a given simulated population of people using a year’s worth of simulated data for 100 hypothetical SNL employees in a total of 22 different databases. The method by which subjects will encounter this data differs across three tool use conditions (with tool, raw data but no tool, and scenarios but no tool). There will be several dependent variables collected during the course of these sessions including average rate of decision time, accuracy of decisions made, and subjective measures of decision ease. Each of these variables will be measured at the conclusion of each tool use condition.
Subject participation is expected to last approximately eight hours and will occur over three to five sessions, depending on subjects' availability.
Potential risks to subjects include the fatigue normally experienced when engaging in cognitively demanding tasks for given periods of time. In order to counter this fatigue, there will be plenty of opportunities for breaks.
There is no financial compensation or other tangible benefit for participation in this research project. SNL employees and on-site contractors will be provided a project and task number to charge their time spent participating in this project. Overall benefits of this research include the possibility that use of a tool similar to the one you will be using in this study might help detect treasonous or other insider activities against the United States earlier than in the past. In addition, this kind of technology might also be applied in the private sector to prevent insider crimes like embezzlement and workplace violence.
"FY04 Facial Image Collection Database"
Principal Investigator: Mr. Larry J. Wright, Sandia National Laboratories
Project started in: 2003
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 04/02/03
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 209
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
As part of work included in Technical Support Working Group (TSWG) Facial Recognition Technology work, Sandia National Laboratories (SNL) will conduct a photographic study in support of the National Institute of Standards and Technology (NIST) Facial Recognition Verification Testing (FRVT) 2004. The goal of the project is to stimulate development of better methods to recognize humans using facial characteristics for use in security system applications. Participants in the study will have their picture recorded on both video and still photography digital cameras. The collected images will become parts of a database delivered to NIST for use in evaluating biometric algorithms used for identifying people by matching their different facial images. SNL will use the videotaped data in continued support of biometrics testing for at least the next 5 to 10 years and research projects that are currently in progress, and the digital still images will be delivered to NIST and/or the Defense Advanced Research Projects Agency (DARPA).
Computer algorithms can be used to recognize people in different types of images. For example, a person might be recognized by matching a picture of their face or the iris of their eye against a database of similar images from known persons. This study seeks to develop data sets and performance evaluation methods in support of the NIST FRVT 2004. In order to compare the relative effectiveness of competing approaches to recognition of humans from sensory, or visual input, the FRVT program will need data sets that capture a variety of different source data on the same group of people. For example Algorithm A might use color images of the face to recognize humans, whereas Algorithm B might use a shape model acquired from a video camera image. To compare the relative effectiveness of the methods, it is necessary to have different types of images acquired on the same people.
Briefly, the data collection process is performed as follows: each participant will pose for several pictures of the face taken with two separate digital still cameras under two different lighting conditions, then a video camera will record a walking sequence. Each participant will slowly walk toward the video camera and then stop and pose. The participant will move both his head and then eyes up and down, and from side to side while the camera is recording. The photographic and digital cameras do not pose any risk to the human volunteers, and participants should not experience any discomfort beyond that typically associated with having a picture taken.
To test facial identity algorithms to be used in security related devices, it is necessary to record human volunteers on different days and at different times of day, over a period of time. The photographic data collection process is estimated to last approximately one year, with a possible extension of up to five years depending on funding. Each participant’s facial features will be measured and the distances between the subject’s eyes, for example, will be recorded. This data, along with the original images will be provided to NIST.
The collection of images and associated data from each participant will be given a date and a session identification (ID) number. The ID number will only be used to organize the database information at SNL. No identifying information such as a name, organization number, or social security number will be stored with the data. However, some images from the database may appear as examples in NIST research publications, without names attached.
There are several benefits to be derived from this project. The primary benefit is to the national research community and society as a whole. This study will aid the development of biometric algorithms to be used in security systems for identification of the population, resulting in increased safety for the population. Also, the results of this evaluation will be used in several analytical tools that are currently being used to evaluate and upgrade the physical security of DOE facilities. Lastly, this information will likely be transferred to other federal agencies [e.g., Federal Aviation Administration (FAA), National Institute of Justice (NIJ), DoD, DARPA, etc.] for use in their respective protection activities.
"Enabling Technology for Human Collaboration"
Principal Investigator: Dr. Adele B. Doser, Sandia National Laboratories
Project started in: 2003
This project ended in fiscal year 2003.
Funding for Human Subjects Research:
This project does not involve the use of multiple protocols/subprojects.
Institutional Review Board (IRB) Review:
Type of Review: Expedited
Approving Institution: Sandia National Laboratories
Most recent approval: 07/10/03
Number of human subjects who participated in this project/protocol/subproject in the last reporting period: 5
Reporting period for number of human subjects:
Fiscal Year 2003
Type(s) of Human Subjects Involvement:
As wireless systems continue to improve, and personal processing systems such as palm pilots become commonplace, one can envision a future melding of technologies allowing for a broader range of applications that will affect everyday life. By 2010, we see a convergence of information and communications technology that will allow design and operation of a completely novel human-machine system (see “The Observant Computer,” Rebecca Zacks, Technology Review, April 2003). People will have small, wearable processors (armband size) or "Pals", which may continuously integrate information about the human into one central adaptive group system architecture we call "MENTOR." Physiologic, personality, attention, and interpersonal factors are among the kinds of data we believe might serve as bases for algorithms enabling higher group performance. Humans and computers are thus not separate systems, but a single complex adaptive system with both autonomous explicit self-models and goals in the context of the human group's defined mission. The MENTOR System will enable human teams to operate at consistently superior levels of performance. It is intended as an immersive environment for teams working in high consequence, high stress, ambiguous, and time critical situations, especially those for which it is impractical to adequately train or prepare.
Objectives:
The goal of this project is to develop new methods to assist groups working in high consequence, high stress, time-critical high-cognitive load tasks, especially those for which it is impractical to adequately train or prepare. The scheme developed will be used to improve task performance in such situations. The belief is that the incorporation of physiological data can be used to accomplish this goal.
Test Protocol:
Subjects will be asked to participate in an interactive video game that requires that, to be successful, participants must function in a collaborative group environment. During the sessions, participants will be asked to wear any or all of the following non-invasive sensor equipment: a small device strapped onto an arm (<500 g), a hat with sensors (<1 kg), skin electrodes (a few grams each), a chest strap (~50 g). The following data will be collected from the participants for use in the study: electroencephalogram (EEG), electrocardiogram (ECG), electromyogram (EMG), electro-oculogram (EOG), skin conductance, respiration, tilt (head, right, and left hand), and pulse.
While wearing the equipment, participants will be asked to engage in an interactive group video game in a conference room setting. Subjects will first participate in individual “training sessions” to learn the game. No special preparation is required on the part of the participant. Wireless collection schemes will not be employed, so participants will be connected to table-based equipment and will need to remain seated for the length of each session (approximately one hour).
The purpose of the video game scenario is to observe the evolutions occurring as the group becomes accustomed to working together. No special questions or psychological tests will be conducted on the participants, except for those questions that arise during the course of the video game scenario. The purpose is to observe the personal and psychological interactions of the group, rather than to direct them.
Subject Recruitment:
The only exclusion criteria are being a direct report to the PI or project manager or having previous experience playing the video games. Subjects will be allowed to view the test equipment before signing the consent form and will be reminded upon seeing the equipment that they may withdraw from the experiment at any time. The subjects should be able to wear the devices for approximately one hour without having to leave the room (15-minute “compatibility sessions” will be conducted prior to the first recording of any actual data to determine if the subject has any physical discomfort from the equipment). This information is provided on the consent form. Subjects who communicate that they have discomfort will be encouraged to reconsider their decision to participate. In addition, potential subjects who discover they are allergic to adhesives used in medical procedures will also be asked to reconsider their participation.
Risks/Benefits:
Since all sensors are non-invasive, and all equipment is electrically isolated, the risk for potential physical harm is less than or equal to daily life or the performance of routine physical exams/tests. The discomfort of the study would be less than that experienced in most non-invasive medical procedures. Possible risks include temporary discomfort from wearing sensors and headgear or potential local allergic reactions from adhesives.
Data Collection and Storage:
Aside from physiological data (which is not personally identifiable) the personal information gathered in this study will include name and gender. Name information will be used in the collection-phase of the study only and will not be included in the final database. Additional personal information may be identified from videotapes. Except as required by law, no person outside the development group will have access to any personally identifiable information or videotapes collected in the study. An exception to the above is as follows: if there is more than one member of each gender included in the study group, the gender of the participant may be included in the final report.
Once the data collection phase is complete, all references to subject names will be stricken from the database. Subjects will only be identified by identification numbers in the final report. These ID numbers will be four digits long, generated at random via computer. Only the research team will have access to name/number information during the collection phase. Data may be presented individually, but there will be no direct identification of the individuals from whom data was obtained.