March 30, 2020
Huntsville, AL 35801
DUNS #: 078706159 CAGE code: 6TYK3
COronavirus Measurement and Positive Alert System (COMPAS)
for Disease Infection Monitoring and Pandemic Spread
Developed: 23 March 2020 Submitted by: Barbara J. Reeder
This document and any attached materials are the sole property of DEFTEC Corporation and are not to be used other than to evaluate DEFTEC products and services. This document and any attached materials are not to be disseminated, distributed, or otherwise conveyed throughout your organization to employees without a need for this information or to any third parties without the express written permission of DEFTEC.
COronavirus Measurement and Positive Alert System (COMPAS) for Disease Infection Monitoring and Pandemic Spread..................................................................................................... 1
COronavirus Measurement and Positive Alert System (COMPAS) for Disease Infection Monitoring and Pandemic Spread
A crucial component to monitoring any infectious disease crisis is the ability to know and track such parameters as infection rates and the geolocation of outbreaks, not only to provide situational awareness of a pandemic, but also to help inform decisions regarding asset and resource allocation. The Coronavirus Measurement and Positive Alert System (COMPAS) is based on COTS-based, fielded systems designed by the DEFTEC team to provide a common operating picture during both natural and manmade emergencies, including disease pandemics (Figure 1). COMPAS is capable of receiving inputs from a wide array of data feeds, integrate this information, and, by virtue of its machine-learning based algorithm, provide decision makers with real time, accurate analysis of the operational environment. With this information, incident commanders can not only gain insights into the current state of a pandemic but can also use the COMPAS decision algorithm to determine predicted outcomes based on decisions made in light of the current operational picture. COMPAS will leverage robust, out of the box capability currently deployed in multiple, real-world environments in support of senior leaders at the federal, state, and local levels. COMPAS operates from both Android and iOS operating systems to provide feedback to incident command centers via secure WIFI. The COMPAS solution is scalable, user friendly, and has reasonable upfront and operating costs. Based on its ability to integrate an array of data sources, the COMPAS provides incident commanders with dynamic decision support capability under a host of emergency scenarios. For the purpose of this White Paper, DEFTEC proposes deploying COMPAS in conjunction with FLIR’s EX Series thermal sensor to conduct rapid, high confidence screening of individuals to determine possible Coronavirus (COVID-19) exposure. Leveraging real-world expertise gained by fielding systems similar to COMPAS in support of the USG and local municipalities, DEFTEC will modify its current approach to meet requirements for disease infection monitoring and pandemic spread.
Figure 1 Coronavirus Measurement and Positive Alert System
For the purposes of meeting national and international requirements for information technology capability to track infections during a pandemic, the DEFTEC team proposes employing its COTS- based (COMPAS) solution to provide a unique common operating picture and situational awareness of current COVID-19 infections, including details of individuals exposed to and/or potentially infected by this virus using a commercial FLIR thermal sensor for screening these individuals. COMPAS is capable of handling multiple streams of data from a host of collection sites in real time, storing it, and transmitting it over a variety of secure networks to command centers for use by decision makers. In addition to collection, storage and transmission, the COMPAS machine-learning based algorithm can integrate multiple data feeds to provide a common operating picture of the current state of Coronavirus infections at both the micro (state and local) and macro (national and potentially international) levels In addition, the robustness of the COMPAS system offers commanders with predictive analytic capabilities, providing expected outcomes based on commander decisions in response to the current state of the pandemic. The system’s extensive data management capability offers commanders the ability to readily access data collected previously and integrate it into the current common operating picture, providing the ability to perform trends analysis that enhances the system’s predictive capability. This extensive and powerful data management capability also provides researchers with the ability to populate and test detection algorithms with real world, validated coronavirus data.
Individuals are scanned by the COMPAS thermal imager; the unit contains all needed data and algorithms to test a person without needed any input from other servers or locations. Test results are given to the local operator when complete. At the same time: 1. test result metadata is transmitted to situational awareness servers, and 2. test images are aggregated for subsequent transmission to algorithmic research and optimization servers.
The proposed COMPAS solution focuses on calibrated, accurate measurement of facial temperature variations as an indicator of possible Coronavirus exposure and/or infection. Thermal imaging data will utilize the FLIR EX SERIEC camera. This camera is a well-established, low cost COTS technology for use in the field for thermal imaging applications. All currently available, affordable COTS thermal imagers, including the FLIR EX SERIEC camera, utilize "uncooled microbolometer" image detectors. While affordable, uncooled microbolometers have lower sensitivity and resolution than that of competing technologies. Typically designed for flame and/or industrial hot spot detection, simply re-purposing unmodified COTS thermal imagers for a medical application risks deploying detectors poorly tuned to the subtle, critical differences generated by coronavirus.
The thermal image (of an individual of unknown Coronavirus status) captured by the FLIR camera is transferred to an iPad running the COMPAS app. The app interfaces with a COMPAS application program interface (API) to process and report the test results immediately. All processing and reporting occur locally (at the edge device), with no external communication required.
To minimize misdiagnosis while facilitating use of readily available, affordable COTS thermal imagers, critical elements of the COMPAS approach leverage "Reference High" and "Reference Low" thermal targets; active heated/cooled targets that are included and positioned in each captured
image. Using the original, unaltered image from the thermal camera, the detection algorithm locates the Reference Target pair in each image, sets upper and lower detection range, and adjusts the image "Gamma". Thus, the detection algorithm utilizes the maximum dynamic range the imager can deliver (or in photographic terms, the optimal "exposure" settings), tuned to the characteristic coronavirus temperature range.
Finally, incorporating Reference Targets in each image imparts a stable, permanent benchmark of thermography characteristics to each image, so that doctors and scientists can reliably conduct research, improve true detection rates and reduce false positive/false negative rates.
The DEFTEC team proposes that collection be conducted remotely and automatically to enhance collection efficiency and eliminate system operator exposure. In addition, operation of the FLIR camera at standoff distances of six feet of more without loss of thermal data accuracy also minimizes exposure of collection equipment to potentially exposed individuals.
The COMPAS technical throughput rate/capacity far exceeds what can responsibly, safely be achieved by real-world medical staff. Based on current capabilities, COMPAS will be able to process at least 6 individuals per minute, per imager. However, given the need to maintain safe separation distance between potentially infected individuals, we expect that the practical throughput per sensor station will be 1 to 2 persons per imager, per minute. The COMPAS system is capable and scalable; real-world logistical challenges of safely/responsibly moving humans through the testing area will be the limiting factor in COMPAS throughput, not inherent technical limitations of the system.
Test data and results from COMPAS edge devices is accumulated, processed and displayed on Situational Awareness and Decision Support servers in real-time. These low-latency, small-sized data payloads are critical to understanding and responding to emerging issues as they occur.
Data collected by the deployed thermal sensors will be automatically transmitted to command and control centers via secure WIFI. The DEFTEC team proposes utilizing patent-pending cellular routers to automatically switch between cell towers/providers on the fly depending upon signal strength always providing the best connection available on the backnet channel. This patent pending proprietary technology connects exclusively with the 9-1-1 cellular backnet channel via a use contract with the Canadian cellular service provider, Telus. This Telus backnet channel use contract provides for reciprocity with AT&T, Verizon, T-Mobile, Sprint, Telefonica, and Vodafone, providing service coverage in the U.S., Canada, Mexico, Central and South America, Western Europe and parts of Asia. The secure channel cellular communications sits behind a Palo Alto firewall and provides almost unlimited bandwidth, greatly reducing hacking risks and data security breaches.
In the event of transmission interruption at one or several thermal collection locations, data collection and storage will continue automatically until communication is restored. When connection is re- established, offline data collected in the interim will be automatically transmitted.
Metadata from edge detectors is transmitted immediately as described above to the COMPAS centralized visualization servers for situational and geospatial awareness, correlated event tracking, and emergent threat modeling. The core of the COMPAS data management and visualization capability is the Touch Assisted Command and Control System (TACCS™), a command, control, communications, computers, combat systems and interoperability (C5I) software solution providing an out-of-the-box common operating picture, and both situational awareness and decision support tools for domestic and international end-users. The user interfaces are designed for operations centers, dispatch desks, executive offices and field personnel to support seamless operational decision making at all command levels. TACCS™ is an operator configurable multi-display of ready-reference dashboards available as an on-premises or cloud hosted solution. Its features include alert processing, the ability to integrate numerous open source data feeds, critical infrastructure management, GPS tracking, and live camera feeds, event management and recovery and mitigation planning. TACCS™ gathers, collates, synthesizes and disseminates information in real-time to all appropriate and authorized parties.
TACCS™ is built with a service-oriented open architecture utilizing geographic information systems. In addition, most common data formats are also easily integrated including Web Map Service (WMS), Web Feature Service (WFS), ArcGIS Feature Data, Representational State Transfer (REST) API, Google Maps Webfeed (GeoRSS), and KeyHole Markup Language (KML).
Using this interface, agencies such as DHS and CDC will be able to categorize, define, detect, and manage response to situations and incidents based on agency-specific requirements and policies.
Test images from COMPAS edge-based Thermal Cameras are added to Algorithmic Development and Refinement servers when high-bandwidth connections allow. These high-latency, large-sized data payloads are critical to understanding the disease and methods of increasing detection effectiveness. When refinements are made to detection routines, these enhancements can be pushed to edge devices to increase the effectiveness of all fielded detectors.
Thermal images captured will upload to a centralized library for detection algorithm improvement and refinement, using the growing repository of real-world images as the training and testing library for any and all Machine Learning (ML) and Artificial Intelligence (AI) efforts related to Coronavirus tracking and monitoring. Rather than tying COMPAS to a specific ML or AI algorithm, our premise is to build the library through which any ML/AI can be engaged. With this approach centralized authorities will thus be able to readily identify the best ML/AL algorithms to employ given a specific set of operational circumstances, and to push these improved detection algorithms and settings to individual COMPAS systems as understanding of COVID-19 evolves.
As outlined above, the COMPAS solution offers a comprehensive approach to Coronavirus thermal screening, incident management and response, and emergency decision support. As part of providing the most comprehensive capability to decision makers during this crisis, the DEFTEC team will provide the following support during implementation and continued operation of the COMPAS solution for Coronavirus screening and tracking:
DEFTEC proposes employing the COMPAS system to screen individuals for possible exposure to Coronavirus by calibrated capture of facial heatmaps using the commercial FLIR EX SERIEC thermal imager. Collection can be conducted remotely, and data can be capture and stored automatically. Collected data can be indexed and transmitted through secure networks for integration and analysis.
Using thermal variations to determine individual infectious disease exposure is well established. During the 2003 SARS outbreak in Asia, many governments conducted thermal screenings of arriving airline passengers to identify possible infected individuals.
While this approach provided some baseline screening utility at the time, the proposed COMPAS- based thermal screening approach offers several advantages and advances over previous, more traditional screening methods. The COMPAS-based system uses an open architecture and employs a COTS thermal sensor (the FLIR EX SERIEC) whose detection algorithm can be remotely updated in real time in support of evolving mission objectives. Senor operation and collection can be conducted remotely by the user, and the sensor can operate efficiently at standoff distances (six feet or more) to minimize direct exposure of the senor to individuals being screened. The system Incorporates known- cold and known-hot targets into images to verify proper imager calibration, minimize error, and maximize sensitivity over human skin temperature ranges. Images, results, and metadata can be automatically sent to centralized servers for situational awareness and emergency management. With Machine Leaning/AI engines Emergency Management is facilitated through a live picture of on-the- ground truth and conditions in real time.
In addition to the specific application for Coronavirus thermal screening, our proposal provides a package that encompasses service and software to support the entire data capture, transmission, integration, and analysis process in an efficient and automated manner.
The flexibility of the COMPAS system makes it ideal for use at both the micro and macro levels. The system can build an extensive screening network that provides situational awareness from the municipal, state, and national levels. Being a scalable solution, the COMPAS system is ideal to support officials responding to a localized outbreak, officials who can benefit from COMPAS’ predictive and trend analysis capabilities to quickly and efficiently make decisions regarding asset deployment and resource allocation, and researchers who can use COMPAS data and analytics to refine current capabilities and define future technical requirements for infectious disease response. The COMPAS solution provides commanders with tailored, robust emergency management support.
Because there are essentially no limitations on the types and volume of data input, COMPAS is capable of handling virtually all data management and analysis need in response to local, state, or nationwide emergencies.
The proposed COMPAS solution builds on successful fielding of this technology in support of mission critical needs for both USG customers and local governments, including the Metropolitan Washington Council of Governments. With the appropriate security and privacy protections in place, DEFTEC stands ready to provide the details of these applications to government organizations.
While the COMPAS solution builds on readily available COTS technologies and fielded capabilities developed by the DEFTEC team, our approach is a novel creative alternative to HTC methods currently in use. The COMPAS solution overcomes several limitations to current approaches.
The COMPAS is based on mature, fielded, COTS technology that has proven its utility in real world incident response. With proper data sensitivity and proprietary information protections in place, the DEFTEC team will provide the government organization with performance data for its systems currently fielded in support of its customers, including the USG. The DEFTEC team will also work to design and run a COMPAS field test under operational condition relevant to disease monitoring during a pandemic. For this test, several key COMPAS components will be demonstrated and tested including:
A successful test/demonstration of the COMPAS capability will validate the system’s ability to collect thermal data on a representative number of people, store and retrieve this data, and transmit this data over a secure WIFI network to a remote operation station. At the station, COMPAS will integrate this thermal data with other data feeds such as geolocation, time, and, where appropriate and in conjunction with organization’s requirements, personal identifying information on individuals screened. The COMPAS’ output will be a common operating picture, consisting of a map identifying the location of screened individuals, a pop-up capability for each individual providing metadata including detailed thermal mapping of calibrated facial temperature variations during collection.
The DEFTEC team is prepared to demonstrate COMPAS capability during a standalone test developed in conjunction with the agency, or as part of a preplanned test and/or field trial. Depending on requirements expressed by the agency, DEFTEC stands prepared to work with other vendors to demonstrate COMPAS capabilities as part of a system of systems, designed to provide a comprehensive COVID-19 response.
SECTION 3: REGULATORY STRATEGY
COMPAS will comply with regulatory and security requirements, providing access to the solution based on user roles and privileges controlled by the administrative interface. Agency-specific structure models will be set up and maintained up to the minute regulatory requirements will be enforced as directed by the Government. This will be achieved through metadata modelling and application of contextual structures.
DEFTEC and its team members, Fluid Topics, Inc and ESI Convergent, LLC can put together a pricing guideline estimate to include two options.
Basic baseline operations option fully capable of all aspects of the solution design but does not identify the individual being screened to the imaging data sets. It does allow for either Wi-Fi or Cellular communications and is fully self-deployable on the Edge with or without connectivity to any LAN, WAN or cellular. The ESI Convergent, LLC team will provide phone support for any unforeseen issues that arise; however, the design implementation is 100% self-deployable without technical assistance.
Advanced option, which builds on top of the Basic baseline design by including:
This provides the ability to maintain the maximum screening throughput and allows for an Identity to be correlated with sensor data sets in a GDPR compliant manner. This data Library will be applied to both existing and future analytics layers to enable the advanced detection of Human Temperature Controls and other imaging applications.
For a tailored and detailed pricing estimate please contact DEFTEC at George.firstname.lastname@example.org