Source: TRIDENT SENSING LLC submitted to NRP
AIRBORNE WILDFIRE MAPPING UTILIZING AN ARRAY OF THERMAL INFRARED CAMERAS AND AUTOMATED MULTI-APERTURE/MULTI-PLATFORM FIRE MAP PROCESSING
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
COMPLETE
Funding Source
Reporting Frequency
Annual
Accession No.
1026231
Grant No.
2021-33610-34697
Cumulative Award Amt.
$99,829.00
Proposal No.
2021-00956
Multistate No.
(N/A)
Project Start Date
Jul 1, 2021
Project End Date
Feb 28, 2022
Grant Year
2021
Program Code
[8.1]- Forests & Related Resources
Recipient Organization
TRIDENT SENSING LLC
8475 WEBER DR
KAMAS,UT 84036
Performing Department
(N/A)
Non Technical Summary
Climate change is increasing the frequency and severity of western wildfires. The threat to people and infrastructure is growing with more population living in the wildland-urban interface. Advances in technology provide the opportunity to improve wildfire mapping accuracy, timeliness and dissemination. An un-stabilized airborne fire mapping sensor is proposed utilizing an array of miniature thermal infrared and color cameras providing near hemispherical downward-looking coverage. A solid-state Inertial Navigation System (INS) is included to provide accurate sensor orientation for geo-referencing fire mapping data. The sensor is small and light enough to be conformal mounted on every airborne firefighting asset making all participating aircraft fire mapping nodes. An innovative image processing algorithm will automatically combine multi-aperture and multi-platform sensor data onto a common time-based 3D fire map. A prototype test sensor will be test flown to validate sensor sensitivity and INS pointing accuracy data. Fire imaging data will be processed and depicted on a time-based 3D map.The successful development and commercialization of the proposed technology will provide Incident Commanders improved time-based data for managing wildfires that threaten people and infrastructure. The end goal is to improve the efficiency of fighting wildfires. Damage costs will be reduced and lives will be saved by improving the situational awareness of firefighting personnel and the general public.
Animal Health Component
40%
Research Effort Categories
Basic
10%
Applied
40%
Developmental
50%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
1227210202050%
1227210201025%
1227210208025%
Goals / Objectives
Project Title: Airborne Wildfire Mapping Utilizing An Array Of Thermal Infrared Cameras And Automated Multi-Aperture/Multi-Platform Fire Map ProcessingGoals:o Design, manufacture and evaluate through flight test a small, un-stabilized airborne wildfire mapping concept demonstration sensor (Reference WBS Element 1)o Design and demonstrate an innovative image processing algorithm that will automate fire mapping and automatically combine multi-aperture and multi-platform sensor data onto a common time-based 3D fire map. (Reference WBS Element 2)oDevelop Framework for Sensor/Platform Interoperability, including NIROPS and satellite platforms. (Reference WBS Element 3)o Develop a Phase II design concept for a small, low cost and lightweight multi-aperture sensor package and associated direct to map processing that can be deployed on any airborne asset, to include future USFS Unmanned Air Systems (UAS), with minimal aircraft modifications. (Reference WBS Element 4)
Project Methods
1 TIR CAMERA CORE SENSITIVITY AND INS ACCURACY TESTING: This task supports Objective #1 - TIR and INS sensor suitability for the fire mapping mission. An additional set of sensors will populate a lab integration unit to complement Trident Sensing's flying demonstrator. The lab integration unit will use 3D printed parts and serve as a hardware integration and software development unit. It will be populated with the second ship set of sensors. During flight test, the lab unit components will serve as spares for the flight test sensor. The flight test sensor contains 2 overlapping TIR camera cores, 1 Color/NIR camera core, INS/WAAS GPS and supporting processors and memory. The SBIR Phase I effort will pick-up with Trident Sensing's IR&D effort at the flight test phase and will concentrate on obtaining scientific data.1.1 TEST SENSOR DESIGN: This task captures all HW and SW efforts to develop the flight test camera system. This effort was completed under Company IR&D.1.1.1 MECHANICAL: The mechanical design task encompasses all design and manufacturing of the flight test sensor suite including aircraft mounting scheme. This effort was completed under Company IR&D.1.1.1.1 This task also supports creation of a lab "brass board" unit for HW/SW integration activities.1.1.2 ELECTRICAL: Develop system interface specification and complete all electrical interconnects required for system functionality.1.1.3 SOFTWARE: The SW development task encompasses all design and coding of software to support the flight test sensor camera control and data capture using the Python object oriented language.1.2 FUNCTIONAL TEST: Testing will be performed by the PI, SW lead and technician to verify test unit functionality. The test will be supported by functional leads on an as needed basis.1.2.1 CALIBRATION: Measure each camera core alignment relative to INS frame of reference for SW geo-referencing corrections. Generate corrections for flat field distortions.1.2.2 LAB TEST: Verify sensor control in lab environment. Capture attitude, position and camera imagery data.1.2.3 AIRCRAFT INSTALL: Install flight test sensor on aircraft.1.2.4 AIRCRAFT GROUND TEST: Verify full functionality as installed on aircraft in ground test environment. Perform safety of flight EMI/EMC source victim testing. Verify safe electrical loads. Provide flight test readiness review recommendation.1.3 FLIGHT TEST: This activity will quantity the TIR sensor sensitivity and INS accuracy as well as gather data for the fire detection algorithm validation. The test will be flown by a consultant chief pilot utilizing a leased general aviation aircraft. 20 flight hours are budgeted to this task. The PI will coordinate the ground test crew activities to include laying out the IR test target grid and tending the test targets during the overflights.1.3.1 IR TEST TARGET BORROW OR BUILD: This task encompasses the effort to obtain or build four IR test targets from 0.0156 M2 to 1 M2. IR targets will be similar to a charcoal briquette barbeque grills but with a controlled footprint. This task will be performed by a technician.1.3.2 FLT #1: FUNCTIONAL CHECK FLIGHT: Flight #1 will verify flight test sensor operation using IR target Scenario #1.1.3.3 FLT #2-5 TIR SENSITIVITY /IMU ACCURACY: Flight #2-#6 will gather TIR camera sensitivity and INS accuracy data in straight and level and maneuvering flight at multiple altitudes.1.3.4 FLT#6-7: B/U OR LIVE FIRE TEST: Flights #6 & #7 are back-up flights. If not needed as a back-up to Task 1.3.3, the flight test aircraft will be flown over a local wildfire to obtain real world fire mapping data. It is preferable to overfly a fire that has recently been mapped by the USFS PHOENIX fire sensor for comparison purposes.1.3.5 UAS DEMONSTRATION: Mechanical and electrical integration of sensor to representative multi-rotor UAS. Fly two test flights on IR targets to gather test data and UAS lessons learned for Phase II design.1.4 DATA ANALYSIS DATA ANALYSIS: Quantify TIR camera sensitivity and INS accuracy from flight test data. Qualitatively evaluate sensors for blooming, smear, jitter etc.1.4.1TIR IMAGERY SENSITIVITY: Calculate NADIR/slant detection ranges for the various IR targets.1.4.2 INS ACCURACY: Calculate IR target location from INS pointing and GPS position data. Compare to surveyed IR location and calculate pointing accuracy in straight and level and maneuvering flight.2 AUTOMATIC FIRE DETECTION ALGORITHM: This task supports Objective #2 - Implement automatic fire detection and mapping algorithms for processing fire data and registering the data to a common multi-platform 3D reference frame. This task will be conducted by the PI and commence at project kick-off. All image processing will be conducted off line using a high level scientific/mathematical programming environment such as MATLAB. Some minimal SW support is programmed for this effort.2.1 ALGORITHM DEVELOPMENT: Develop multi-camera and multi-platform image processing algorithm that will combine multiple camera images.2.2 MATLAB IMPLEMENTATION: Implement data processing algorithm coded in MATLAB.2.3 EVALUATION ON FLIGHT TEST DATA: Test data processing algorithm using actual flight test data offline in MATLAB. Combine multiple camera images into a single time based 3D image frame. Evaluate fire map data playback using 3D GIS platform such as GeoCollaborate, ARCGIS, Google Earth or Avenza.3 INTEROPERABILITY WITH SATELLITES/NIROPS DATE: This task investigates how to incorporate existing sensors and platform imaging data into the common time-based 3D fire map. A representative platform/sensor will be selected and data sharing quality metrics will be assigned. Real world data will be incorporated into the 3D fire map. Ideally this will be NIROPS or satellite mission data coinciding with the live fire flight test missions in Task 1.3.4.3.1 SELECT PLATFORM/SENSORS: Select a platform and sensor. Assign quality metrics based on thermal capability and spatial resolution.3.2 IMPLEMENT IN MATLAB: Incorporate data into MATLAB fire map processing code and overlay on the common map.4 PHASE II SYSTEM DESIGN CONCEPT: Perform high-level prototype sensor design using lessons learned from Phase I. This effort will feed into the Phase II proposal effort.4.1 MECHANICAL: Utilize Phase I lessons learned for compact prototype mechanical packaging4.2 ELECTRICAL: Utilize Phase I lessons learned for prototype electrical design. Determine if custom boards or carrier cards will be required in the prototype sensor.4.3 SOFTWARE: Utilize Phase I lessons learned to architect prototype software package.4.4 MAP PROCESSING ALGORITHM: Utilize Phase I lessons learned to propose enhancements to the map algorithm.5 TECHNICAL REPORT: PI authors technical report documenting Phase I work, results, conclusions and recommendations.

Progress 07/01/21 to 02/28/22

Outputs
Target Audience:The target audience of this research is the wildfire mamagement communityandUSDA researchers under theTopic 8.1.6: Developing Technology that Facilitates the Management of Wildfires on ForestLands. A sub group is the remote imaging community. This includes scientific research groups such as NIFA and NASA. Additionally, the results of this research will presented at the USDA/US Forest Service and NASA co-sponsored Spring 2022Tactical Fire Remote Sensing Advisory Committee (TFRSAC). Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?This project provided valuable real world hands-on experience to five undergraduate and graduate interns in the fields of mechanical/manufacturing engineering, computer engineering, and software development. How have the results been disseminated to communities of interest?The results of the USDA NIFA sponsored SBIR Phase I project were briefed on 19 May 2022to theSpring 2022 - Tactical Fire Remote Sensing Advisory Committee Meeting. TFRSAC is sponsored jointly by NASA, USFS, and USGS. What do you plan to do during the next reporting period to accomplish the goals?This is the Final Report for SBIR Phase I. No further reporting anticipated until SBIR Phase II

Impacts
What was accomplished under these goals? 1)Design, manufacture and evaluate through flight test a small, un-stabilized airborne wildfire mapping concept demonstration sensor: Two test sensors (primary airborne and lab integration/spare) were manufactured and tested in a relevant airborne environment. Quantitative testing to measure sensor sensitivity and geo-location accuracy was performed using thermal infrared targets placed on level ground and in mountainous terrain. Qualitative testing was performed over a controlled burn in the Apache-Sitgreaves National Forest near Show Low AZ. The Concept Demonstration sensor was sensitive enough to detect sub-pixel hot spots in the range of 6%-10% of ground sample area. Geo-location accuracy of approximatel 10 meters was demonstrated up to 2,000 ft AGL and approximately 13 meters at 4,000' AGL. Both sensitivity and geolocation accuracy were suitable for wildfire tactical airborne mapping. 2) Design and demonstrate an innovative image processing algorithm that will automate fire mapping and automatically combine multi-aperture and multi-platform sensor data onto a common time-based 3D fire map: Data reduction was performed using our Company developed innovative direct to map processing. Flight test data were processed directly to fire map products and overlaid on Google Maps. This demonstrated the feasibility of direct-to-map data products. 3) Develop Framework for Sensor/Platform Interoperability, including NIROPS and satellite platforms: Initial work was performed to identify the parameter weighting schema for combining data from multiple platforms with varying sensor sensitivity, accuracy and timeliness. Combining data from multiple platforms remains a Phase II objective. 4) Develop a Phase II design concept for a small, low cost and lightweight multi-aperture sensor package and associated direct to map processing that can be deployed on any airborne asset, to include future USFS Unmanned Air Systems (UAS), with minimal aircraft modifications: Adjustments were made to the Concept Demonstrator design for Phase II. Improvements include better data synchronization, moving the sensor dome map processing to the control box, and adding GPU map processing.

Publications

  • Type: Conference Papers and Presentations Status: Published Year Published: 2022 Citation: Spring 2022 - Tactical Fire Remote Sensing Advisory Committee Meeting 19 May 2022 TACFI-RS: TACtical Fire ÿ¢ÿ¿ÿ¿ Remote Sensor SBIR Phase I Results Title: Airborne Wildfire Mapping Utilizing an Array of Thermal Infrared Cameras and Automated Multi-Aperture/Multi-Platform Fire Map Processing


Progress 07/01/21 to 02/28/22

Outputs
Target Audience:Targeted Audiences: - USDA US Forest Service / NASA jointTactical Fire Remote Sensing Advisory Committee (TFRSAC) - Civil Air Patrol Headquarters ? - Air National Guard Headquarters Changes/Problems:This Phase I Concept Demonstrator project has gone smoothly. Only small variances from the originally prposed program plan were required: 1) Less flight time was required to achieve the flight test objectives. This resulted in lower aircraft rental and chief pilot costs. 2) Due to the difficulty in finding a suitable commercial UAS (Unmanned Air System), we elected to use non-program funds to build a small UAS. This saved the project from direct charging UAS rental and UAS pilot costs. 3) Direct material costs were higher than proposed for multiple reasons: - The semiconductor shortage caused redesign of circuit board and wiring. This resulted in additional parts orders. - We elected to add a narrow field of view thermal camera to complement the two wide field of view cameras. This allowed the team to gather additional sensor suitability data at higher altitudes. What opportunities for training and professional development has the project provided?Our team includes mechanical engineering, software and computer engineering interns. These interns are obtaining real word experience on a complex multi-disciplinary project. How have the results been disseminated to communities of interest?Phase I project results will be briefed to the Spring 2022 Tactical Fire Remote Sensing Advisory Committee (TFRSAC) hosted by NASA Ames in May 2022 What do you plan to do during the next reporting period to accomplish the goals?The remaining Phase I period of performance will be spent reducing data and on detailed final reporting.

Impacts
What was accomplished under these goals? 1) WBS Element 1: - We designed and manufactured two Concept Demonstration Sensors. The primary sensor was installed and flown on a single engine general aviation aircraft. The second sensor served as a lab integration unit. - Three thermal infrared test targets were borrowed from the National Interagency Fire Center (NIFC) in Boise and two additional larger targets were locally manufactured. - Three flight test events were conducted using the thermal infrared targtets measuring sensitivity and geolocation accuracy in flat terrain and rugged terrain. Additionally a targte was placed under a dense conifer canopy to measure the capability to geolocate largely blocked targets. - Two test flights were flown over a prescirbed burn on the Apache-Sitgreaves National Forest near Show Low AZ. 2) WBS Element 2: - Innovative fire map processing algorithms were implements in software to demonstrate multi sensor and multi platform map processing. This work is on going. 3) WBS Element 3: - We are developing a framework for sensor/platform interoperability. This work is on going 4) WBS Element 4: - We are developing the Phase II design concept. This work is largely complete.

Publications