2-7--ADOT_ECD_DLSSU_RFP--NEC_Ease_of_Use_FINAL_2022-04-18.pdf

An NEC Response to the Arizona Department of Transportation – Enforcement and Compliance Division ECD Drivers License Security Software RFP – BPM004292

Section 7 – Ease of Use

Page 7-1

7 Ease of Use RFP References: APP Portal, Financial Questionnaire, Proposal Requirements for Drivers License Security Software Overall Solution, Ease of Use; ADOT ECD DLSS RFP, Scope of Work, Section 10

As part of our application development process, NEC has an integrated review process to analyze both the user interface (UI) and user experience (UX), ensuring both ease of use and simplicity of design. Our approach to accessibility, which aligns with related Arizona policies noted in Section 2, “Solution Adherence,” ensures compliance with accessibility guidelines and optimizes the user experience by minimizing friction, thereby enhancing user productivity. In addition, all applications use HTML 5 for web client access, thereby supporting a broad user base and ensuring compatibility with all major browsers. The following sections describe NEC’s general approach to accessibility followed by an overview of the functionality provided by the specific web client applications proposed for the ADOT ECD DLSS project.

7.1 NEC’s Approach to Accessibility

* * * B E G I N C O N F I D E N T I A L I N F O R M A T I O N * * *

Wherever applicable, NEC’s software solutions comply with both national and international principles of accessibility.

Overall, this means the universal design of our solutions meets the Section 508 and Web Content Accessibility Guidelines (WCAG) standards. Our software user interfaces employ the following principles:

1. Equitable Use – Overall design is useful and applicable to people with diverse abilities.

2. Flexibility in Use – The user interface (UI) design accommodates a wide range of individual references and abilities.

3. Simple and Intuitive Use – Use of the design is easy to understand, regardless of the user's experience, knowledge, language skills, or current concentration level.

4. Perceptible Information – The design communicates necessary information effectively to the user, regardless of ambient conditions or the user's sensory abilities.

5. Tolerance for Errors – The UI design minimizes consequences of accidental or unintended actions.

6. Low Physical Effort – The design can be used efficiently and comfortably with minimum fatigue.

7. Size and Space for Approach and Use – Appropriate size and space is provided for approach, reach, manipulation, and use regardless of user's body size, posture, or mobility.

Depending on the intended use, NEC’s software applications may be developed and launched based on 1) fluent design guidelines and code examples for Microsoft Windows applications; or 2) web architecture where software logic is divided between a server and a web browser client using HTML 5 and XML W3C. Both approaches result in intuitive app experiences for both regular users and users with disabilities. These include:

• Design Basics – Navigation in NEC apps is based on a flexible model of navigation structures, navigation elements, and system-level features. Command elements are the interactive UI elements that enable the user to perform actions, such as sending an email, deleting an item, or submitting a form.

Page 7-2

• Layout – Responsive tips and application programming interfaces (API) for a UI that looks great on every screen size.

• Controls and Patterns – All the UI building blocks that users need, from radio buttons to navigation controls.

• Style – Applications that stand out with color, typography, layout, and presentation.

• Motion – Applications that include animations, notifications, and transitions.

• Shell – Tiles for launching the app and displaying notifications.

• Input and Interactions – Optimized for touchscreen displays, cursor position devices, and command keys, as well as surface dial, stylus pen, and other input types.

• Usability – NEC understands that one in eight users may have a disability. Therefore, our applications are accessible as appropriate for their given uses. Microsoft Windows PC hardware and software potentially provide an array of assistive technologies for command, control, display, and annunciation.

NEC's software can be installed and operated on several Microsoft Windows platforms or accessed through most standard web browsers and can be used in multiple ways by a variety of users.

7.2 NeoFace® Reveal

As incidents of stolen and false identities rises, public safety agencies and commercial enterprises have become keenly aware of the need for fraud detection in their identity issuance workflows to ensure the integrity of driver license and other credentialing programs. Without the use of technology to aid the identification of individuals, many cases of fraud would remain undiscovered.

The latest version of NEC’s NeoFace® Reveal application uses the most advanced face recognition technology available today, allowing operators to scan, detect, and search for poor-quality photographs against databases of known individuals and evidence from unsolved cases. Reveal typically returns search results within seconds, helping expedite investigations to solve identify cases of fraud.

To improve image quality for matching, Reveal provides a comprehensive set of image processing enhancements which enable users to improve detail and remove background noise, standard enhancements include crop, rotate, brightness, contrast, hue, intensity, smoothness, sharpness, histogram equalization, illumination, aspect ratio correction, and de-interlacing.

7.2.1 Benefits of NeoFace Reveal

The NeoFace Reveal software solution provides the following benefits:

• Employs NEC’s high-performance face matching algorithms, delivering some of the fastest and most accurate results available, as concluded by the National Institute of Standards and Technology (NIST).

Improved tolerance for off angle poses providing high accuracy results without 3D rotation tools.

Optimized algorithm training to account for subjects wearing masks.

Improved accuracy for aging allowing for recognition across longer timespans.

Award-winning algorithms power NeoFace Reveal’s detection processes, allowing rapid face identification and image enhancement.

Page 7-3

• Facilitates fraud detection and investigations by providing an innovative and intuitive web-based user interface.

• Delivers a feature- and functionality-rich application that enables investigators to:

View transactions and access face quality control, candidate verification, and database maintenance functions from a single dashboard.

Import and capture face images from photographs and videos.

Improved video processing, including auto face detection and extraction to provide the best quality of image for matching.

Enhance and correct images to improve potential search results and ensure high-quality images are enrolled in the database.

Perform searches using features extracted from captured face images, with or without enhancement.

Review search results and compare the probe image against potential candidate images.

Use demographic search tools to query the photo database based on sex and age range.

Manage exceptions during automated enrollment to detect potential cases of fraud and ensure issuance of only one identity per individual.

Integrated photo lineup creations and online witness sessions for subject identification.

7.2.2 NeoFace Reveal Dashboard

A dashboard is displayed upon launching the NeoFace Reveal application. The dashboard shows the immediate status of work in progress and allows users to readily navigate to available transactions per the pertinent workflow queue. The dashboard consists of individual tiles that provide information on the following:

• Count of pending quality control (QC) jobs and recent QC statistics.

• Count of currently pending QC and verification reviews.

• Statistics on current and previous job/transaction volumes.

• Statistics on opened new cases and their statuses.

Page 7-4

Figure 7-1: NeoFace Reveal – Dashboard

The Database (DB) Cleansing queue (lower right section of the above figure) provides a quick overview of the transactions flagged as potential matches during the database import process. The statistics summarize the status of DB cleanse transactions in the verification process.

7.2.3 Face Quality Control

If NeoFace Reveal is unable to automatically enroll face images due to poor image quality or other reasons, the face images can be held in a queue for user review. Reasons for review include insufficient adherence to quality metrics for image enrollment and the inability to detect a face on a frontal photo, or the detection of multiple faces in the submitted image.

Figure 7-2: NeoFace Reveal – Face Quality Control Edit

Page 7-5

As part of the editing process, if the user determines that the face image should not be enrolled in the database, the user can reject the face image. If the face image is rejected, an error notification can be logged with user-inputted text.

7.2.4 Browsing Photo Databases

The NeoFace Reveal application provides users with the ability to browse the photo repository to retrieve records based on demographic characteristics as described in Section 7.2.6.3, or through a similarity search using face recognition. Resulting images can be viewed and included in photobooks.

7.2.4.1 Results Review

The results of demographic searches are presented on Results Review pages:

Figure 7-3: NeoFace Reveal –Results Review

7.2.5 Photobooks

The user can quickly view the images in the Results Review pages for selection and inclusion in a specified photobook. Images can also be selected from the results for inclusion in a specified case photobook. Photobooks allow users to create collections of photos for grouping and viewing, for up to a maximum of 200 images per photobook. Images can either come from browsing the databases or may be manually uploaded to the photobook as external sourced images. Photobooks can also be used as a base image set for adding photos to photo lineup boards.

Photobooks can be one of three types:

• Global – Contains a collection of face images that will be accessible to other users for investigative purposes.

Page 7-6

• Personal – Contains a collection of face images from the user's agency that will not be accessible to other users for investigative purposes.

• Case – Contains associated images linked to a specific case, including images that may appear in lineups and related witness sessions.

After assembling images, the Photobooks screen allows users to review all photos within the collection and examine each image for additional subject details. Photos are arranged in the order that they were added to the photobook, but images can be manually rearranged as desired. Similar to the Search Results view, the Photobooks screen displays images with additional details and actions available upon image selection.

7.2.6 Case Management

NeoFace Reveal uses a case structure to store and associate information related to an investigation. From the Case Management screen, users can perform several actions to maintain and update the case. Users can:

• Create new cases

• View and add evidence images, documents, and demographic information related to the case

• Select any evidence image for further image enhancement and submit for search

• Create case-based photobooks

• Create case-based photo lineup boards

• Merge cases

• Link cases

• Print case reports

• Assign cases to users

7.2.6.1 Evidence Images

NeoFace Reveal stores face images for investigation as evidence images. Reveal supports JPEG, PSD, GIF, RAW, HDR, BMP, TIFF, and PNG image types for capture.

On the image capture screen, users can upload individual images from a local directory and enter related information for each image. Descriptive data can be entered to provide additional context for the evidence image.

The following data fields are available when entering new evidence images:

• Evidence Number – Unique identifier for images within the case. Each image has one corresponding evidence number. The system automatically generates sequential numbers in this field, beginning with 01. Users can manually override the evidence number by entering a number of their own.

• Method of Collection – Specifies the method and process of collection for the image.

• Location – Description of location where the image was obtained.

• Collected By – Identifier for the investigator who collected the image.

• Notes – Additional notes related to the image (optional).

Once uploaded, Reveal displays all evidence images in a pick list for additional review and processing.

Page 7-7

7.2.6.2 Image Enhancement

Once an evidence image is imported, face data is extracted and manual image enhancements can be applied to improve the image's suitability for face matching. NeoFace algorithms measure system confidence that the image contains a face and that the quality of the extracted image is sufficient. The system confidence is returned as a reliability metric. This metric provides users with real-time information on the suitability of the face image.

Enhancements may improve matching accuracy for poor-quality images or bad capture angles. Reveal provides a comprehensive set of enhancement options to improve image quality and matching ability.

Standard Enhancements

• Crop – Select a sub-section of the image for continued processing

• Rotate – Rotate the image up to 360 degrees to orient the face in a vertical direction

• Resizing – Resize an image to focus on the face.

• Brightness – Increase or decrease overall image brightness

• Contrast – Increase or decrease overall image contrast

• Color Balance – Adjust the color balance of an image

• Adjusting Hues, Saturation, or Lightness (HSL) – Correct the color hue, saturation, and the amount of light expressed in an image

• Inverting Colors – Invert the colors of an image

• Smooth – Perform pixel smoothing to reduce noise and edge effects

• Sharpen – Sharpen pixels to better define edges

• Marking Landmarks – View and mark annotations on facial aspects such as scars or tattoos

• Marking Red Eye – Set the threshold values for red eye correction

• Changing Image Magnification – Zoom in or out of the image

• Adjusting Illumination – Adjust the amount of illumination on an image

• Histogram Equalization – Equalize pixel values to create a level histogram

• Noise Reduction – Reduce background noise in the image

• Flip – Flip and mirror image in horizontal and vertical directions

• De-Interlace – Remove interlacing from analog video still images

7.2.6.3 Demographic Filters

Once the user isolates a face image and submits it for search, the results list will contain faces with characteristics similar to the submitted search face. If desirable, the user can apply demographic filters to narrow the search:

• Sex – Only search images of the specified sex

• Age Range – Only search images within the specified age range

Page 7-8

7.2.6.4 Search Verification

Face image searches that result in potential matches are queued for user verification. The resulting Candidate List window displays the probe image with a candidate list sorted by biometric matching score, allowing users to compare the search image with the matched images. Each potential candidate is displayed in a summary view with the matching similarity score.

Figure 7-4: NeoFace Reveal – Candidate Verification

7.2.6.4.1 Verification Tools

When verifying, the following additional image review tools and view modes are available:

Review Tools

• Zoom – Adjust image sizes to focus on specific image details

• Pan – Adjust center of zoomed image when larger than display area

• Rotate – Rotate image to align in a vertical direction

View Modes

• Side by Side – View images in a side-by-side configuration. Images can be locked so that tools applied to one image are automatically applied to the other.

• Overlay – Overlay images on top of each other, allowing users to review the images through horizontal and vertical sliders and transparency blending.

Page 7-9

7.2.6.4.2 Verification Steps

Verification occurs in a series of steps:

1. Verify 1 – First-level verification by an initial user. This process can be repeated a second time by a different user to confirm cases of no fraud.

2. Verify 2 – Second-level verification used by more senior investigators to determine whether additional investigation is required. Individual candidates can be reviewed without the results of the Verify 1 process being displayed.

Based on user comparison, a no fraud or investigate decision can be marked for each candidate. At any time, users can cancel the verification process without a decision, leaving it in the queue for another examiner.

7.2.6.4.3 Morphological Comparison

In accordance with the Facial Identification Scientific Working Group (FISWG) guidelines for confirmation of facial matches, NEC has included a step-by-step morphological comparison process on 18 specific points of facial identification. During the comparison, users can select individual features from the matching faces, measure coordinates to ensure consistent aspect rations, highlight similar facial features, and annotate facial sections. Each feature provides user confirmation of the match, and all information is consolidated into a comparison report that details the reasons the investigator determined the faces were from the same individual.

Page 7-10

Figure 7-5: NeoFace Reveal – FISWG-Compliant Morphological Analysis on 18 Attributes

7.2.6.5 Photo Lineups

The photo lineup screen allows system users to select photos based on similar demographics for comparison to a known suspect image. The suspect image will always be included in the photo lineup board, highlighted to indicate its position in the lineup. Users can drag and drop images from the general gallery or individual photobooks into the photo lineup. Image placement can be adjusted at any time to update the display order. Users can also press a Shuffle button to prompt the system to automatically and randomly re-sequence photo lineups.

Once all images are populated to the photo lineup board, the user can generate a witness review.

Page 7-11

Figure 7-6: NeoFace Reveal – Photo Lineups Board

7.2.6.6 Witness Review

Once a photo lineup board is populated, the user can initiate a witness review and identification session. During witness sessions, a witness reviews a list of images and identifies the suspect among them.

Witness sessions can be printed and reviewed in the following formats.

• Image Grid View – The image grid view provides an array of selected photo lineup board images for witness review. From the grid view, witnesses can view all images at the same time, manually selecting an image as the identified suspect.

• Individual Image View – The individual image photo lineup view presents images to the witness one at a time.

The witness examines the photos and makes a decision for each suspect. This alternate identification process can create a positive or negative confirmation on each image.

7.3 Insights Workbench

The Insights Workbench application provides a set of applications available from a secured web browser central location for system administration, monitoring, and reporting of all Facial Recognition System (FRS) management functions, such as user management, transaction/workflow management, system configuration and settings, and reporting.

Figure 7-7: Insights Workbench System Management Console

Page 7-12

7.3.1 System Dashboard and Monitoring

Insights Workbench provides a top-level view of all transactions being processed by the FRS. A variety of views can be presented, including bar graphs, pie charts, snap shots, and tables. These can be customized by the agency to provide specific perspectives as well as the different types of FRS workloads (e.g., criminal versus civil identification, latent investigation, etc.). The ability to drill down to the transaction level is also supported.

Figure 7-8: Insights Workbench Dashboard

Figure 7-9: Insights Workbench Transaction Status Monitoring

Also, Insights Workbench provides advanced system monitoring features, such as the ability to monitor throughput SLAs and critical parameters that indicate the health of the system. The administrator can also set threshold levels

Page 7-13 for each service and configure alerts and automated time synchronization across all FRS equipment using a centralized network time protocol (NTP) server. In addition to displaying a visual alert, the alert mechanism can integrate with email systems and text messaging systems (SMS) to notify administrators.

The system administrator can also use Insights to stop and start services and monitor services at various levels across the entire grid:

• Overall FRS services

• Individual component services

• Individual server-level resources (CPU/memory/disk/network)

• Detailed match processing

• Database statistics

Figure 7-10: Insights Workbench – Resource Monitoring Dashboard

7.3.2 User Management

Insights Workbench user management provides the ability to establish and maintain user IDs, passwords, and functional authorizations. The functions also permit the system administrator to selectively set up and manage multiple classes of users with configurable permissions per class. Data access functions such as purge, expungement, deletion, and sealing are only available to a supervisory-level group. Users not in this group will be unable to access or view data access functions. The system administrator can also grant or deny individual function authorizations to override group-level restrictions on a user account.

The administration framework uses a strict role-based access control (RBAC) model. Each role can have one or more functions and each user can be associated with one or more roles. Only users with requisite roles are granted access to a function.

7.3.3 Reporting

Comprehensive reporting and auditing functions are key features of Insight Workbench for FRS operations. Both predefined and ad hoc reporting is available, enabling authorized users to review system use, user activities, and database statistics. All audit trail data is automatically logged and stored in the database, providing easy access to

Page 7-14 management and transaction reports, and offering the ability to track operational inconsistencies. Following is a partial list of Insights reporting capabilities:

• Selective viewing and printing of FRS logs, organized by any combination of dates, times, users, transaction types, file names, device names, transaction results, and problem reports.

• Report output in an electronic, editable format, presented by timeframe, transaction type, and use patterns.

• An exceptions report showing events such as abnormal system occurrences (including shutdowns), unsuccessful logon attempts, component failures, transaction aborts, and viruses.

• Storage and resource usage reports such as total size of repository, used space, and free space.

• Image quality reports that include quality statistics and quality control (QC) rates in specified time periods.

• Transaction reports that include statistical information such as the number of transactions and hit rate by search type in a specified timeframe.

• Ad hoc reports that enable administrators to view the number and percentage of transactions that succeeded or failed parser and image quality checks.

Figure 7-11: Insights Workbench Reports Menu

Page 7-15

Figure 7-12: Insights Workbench Sample Report

7.3.4 System Configuration and Management

Insights enables the system administrator to:

• Add or update system configuration parameters (such as those related to transaction management, interface locators, etc.)

• Perform data backups and restorations

• Monitor job statuses, manage their execution, and filter or search for jobs.

• Batch ingest transactions

• Synchronize the time across all FRS elements

• Manage table data (such as eye color and hair color)

• Maintain case files, unsolved files, corrective actions, and case merges.

• Set up workflows and transaction type associations

• Establish transaction priorities based on a predefined scale (for example, 1–9, where 1 is the highest priority)

• Enable second-level search verifications

• Specify the default number of candidates and candidate images to return for verification.

The administrator can set a default priority for each transaction type. Insights also permits the system administrator to manage verification threshold scores and QC criteria separately for enrollments and investigation searches.

Page 7-16

Figure 7-13: Insights System Parameters – Modifying Settings

7.3.5 Hardware and Site Requirements

Implementation of Insights Workbench client access will require the following:

• HTML 5–compliant browser

• Network connectivity between the site location and the Azure cloud via the ADOT network and point-to-point

VPN tunnel

• Recommended bandwidth of 1 Mb

7.4 Mobile WAY

NEC’s Mobile WAY (which stands for “Who Are You?”) is a smart mobile application that enables law enforcement, border patrol, and public security officers to perform biometric identifications, verifications, and more in their fields of operation. Officers can use their Android or Apple smartphones to determine a person’s identity by matching their face against a biometric-enabled database at a remote site or even on the smartphone itself. NEC’s unsurpassed and unparalleled biometric algorithms assist in the capture of high-quality images for biometric searches, and results are returned instantly to the smartphone.

Page 7-17

Figure 7-14: NEC’s Modern Mobile WAY to Capture and Search Biometrics

Mobile WAY uses secured cloud technology and mobile communications to connect the officer’s smartphone with the mobile app to any participating remote biometric database. Residing in the cloud, NEC’s Mobile ID Platform Services (MIPS) will receive, route, and secure both the biometric search requests and the results, performing real-time auditing and reporting of all user activities and processed transactions.

NEC is developing Mobile WAY to use FirstNet, Microsoft Azure Government cloud, and NLETS infrastructure for the necessary networking and data processing. The following block diagram illustrates the architecture.

Figure 7-15: Azure Hosted NEC MIPS for Connectivity and Interoperability

The overall concept of operations is that the officer will download the Mobile WAY app from a commercial app store and install it on their secure commercial smartphone or all-in-one biometric mobile device. While the Mobile

Page 7-18

WAY app is free to download, a paid subscription is required for operation. With the Mobile WAY app, the officer can capture face images, read credentials, or enter bio-demographic data, and then submit a search to MIPS.

MIPS is the essential middleware that connects the Mobile WAY app to the targeted biometric systems enabling the user to submit search requests and in turn receive search results. MIPS provides more than overall data exchange, transaction monitoring, and workflow management. It also provides user and subscription management to ensure that licenses are created and distributed, and that users are authenticated and authorized to log on and use the mobile app with targeted biometric system search processing. MIPS is hosted in the secured Azure cloud for scalability, extensibility, and availability. Finally, MIPS secures all transactions and users to ensure compliance with agency security standards.

Mobile WAY will be available as either an Apple iOS app or a Google Android app. It is designed and will be tested for mid- to top-tier commercial smartphones compatible with and available through FirstNet. NEC assumes that law enforcement agencies will acquire and manage the FirstNet smartphones on their own.

* * * E N D C O N F I D E N T I A L I N F O R M A T I O N * * *

7.4 Mobile WAY