drone net architecture for uas traffic...
TRANSCRIPT
Drone Net Architecture for UAS Traffic Management
ICARUS GroupMulti-modal Sensor Networking
Experiments
November 27, 2017
ICARUS Group - “Drone Net”Drone Net - Multi-Modal Sensor Network For Small UASShared Air Space Safety And SecurityUAS Traffic Management (RT Catalog of Aerial Objects)Focus on Rural and Urban UTM Scenarios
Sam Siewert Drone Net, ICARUS Group Slide 2
https://utm.arc.nasa.gov/ Kopardekar, Parimal, et al. "Unmanned aircraft system traffic management (UTM) concept of operations.“ AIAA Aviation Forum. 2016.
Significance
Slide 4
Motivation – Large Numbers of sUAS– Droneii, FAA, Sandia, ASSURE– Counter UAS Challenge– senseFly Catalog of Uses
Problem – Default solution – Part 107 for sUAS and beyond– ADS-B for sUAS insufficient, infeasible– RADAR/LIDAR feasibility
Drone Net hypothesis– Networked, multi-modal
(passive/active), information and sensor data fusion for de-confliction
– EO/IR + acoustic, spectral fusion, machine learning
– Compare to and validate with LIDAR/RADAR, ADS-B
sUAS ATM Proposed Solutions1) Blacksage, 2) Droneshield3) Dedrone, 4) Gryphon, 5) AARONIA, 6) UTM, 7) LATAS, 8) UTM partners, 9) ERAU Drone Net
Test, Compare, Train, and Stimulate UTM Development
Drone Net - UAS Traffic Management Concepts
© Sam Siewert, ICARUS Group – Drone Net Concept, 11/1/2017
Air-column Test Range[Drone Net NodeSensor Network]
1 Km
1 Km
ADS-B[Ping 2020i]
ADS-B truth
Machine Vision & Learning[Real-Time and Simulation]
sUAS
Compliant Flight Configuration
Passive Sensing Net
AcousticArray
EO/IR Narrow Field
Performance
ROC, PR, F-measure,new metrics
S or X-band RADAR
Active Sensing
Ground LIDARLimited Range
Detect, Track, Classify, Identify, Localize
Cluster &GP-GPU
GA Traffic
MAVlink
UAS LIDAR For Proximity
Operations
NASDatabase
All-skyHemispherical
Navigation Logtruth
Slide 5
Urban Scenario Roof MountUTM Scenarios such as Parcel Delivery
Embry Riddle flight line provides lots of light aircraft traffic
Campus (semi-Urban) environment
Wildlife – insects, bats, birds, etc.
Sam Siewert Drone Net, ICARUS Group Slide 6
STEM 125 EMVIA Research LabDrone Net Sensor Data Capture and Processing
EO/IR, Acoustic and Embedded Systems Integration and Test
UAV, Instrument Integration and Test
General Research for Embedded Machine Vision and Intelligent Automation
Sam Siewert Drone Net, ICARUS Group Slide 7
7.5 TFLOP, 10TB RAID-10Machine Learning
GPS
DSRCWireless
Access Point
Non-compliant sUAS
All-skyHemispheric
al
Local Drone NetMachine Learning Server
Drone NetMaster DBMS
Acoustic array
ADS-BTx/RxCompliant sUAS
LIDAR
EO/IR with IMU
EO/IRwith IMU
All-skyHemispheric
al
Acoustic array
EO/IRwith IMU
ADS-B Rx ADS-B Rx
Drone Net Passive Instrument Experimental Configuration
Slide 8
EO/IR - Software Defined MSIRGB/Panchromatic Visible Cameras1 LWIR Camera with ZnSe WindowJetson Tegra K1/X1/X2, 802.11 Wireless DSRC, USB3 Hub, Power, NEMA Enclosure
Sam Siewert Drone Net, ICARUS Group Slide 9
CPU0
CPU1
CPU2
CPU3
Low-Power CPU
PowerManagement
Controller
wake-up lines
Kepler GPU(192 Stream Co-Processor
Cores)
Memory Controller
Memory
Real-timeClock
I/O Controllers
USB3Port1Port2VDD_CPU
VDD_GPU
VDD_CORE
Mini PCIe
PCIe8
Channel Decoder
4.5-18VDC, 1.1W
NTSC
Narrow Field8-14 bit LWIR
8-12 bit panchromatic
Machine Vision
EthernetController
12VDC5000
mAmp
802.11Controller
Jetson - Tegra X1/X2 System-on-Chip
Wide Field All-sky camera andMPTS acoustic array
NarrowField
Visible
Slide 10
MV/ML Flight EO/IR Frames[OEM Snapshot for prototype,MV/ML future enhancement]
MV/ML Ground EO/IR Frames[Detection, Classification and Identification
Subset of frames fromContinuous 10Hz baseline]
MATLABGeometric Analysis
& Re-Simulation
OEM NavigationLog Data
HF NavigationLog Data
[future enhancement]
ADS-B Log Data[sUAS, GA compliant
identification]
MV/MLDetection Performance
Receiver Operator CharacteristicPR (Precision/Recall), etc.
Human ReviewDetection, Classification, and Identification
{TP, FP, TN, FN}
Localization Error &ADS-B Identification, Detection
{TP, FP, TN, FN}
Simulated HFOV, VFOVAnd Cross Section of Tracked sUAS
Synthetic Frame Generation
Time Correlated Frame Retrieval
HF truthOEM truth
Optical Navigation truth
Frame Compare
ADS-B truth
Actual
sUAS Not sUAS
PredictedsUAS 250 TP 43 FP
Not sUAS 0 FN 3 TN
Slide 11
Needs Debugging – Literally!Many Insects Detected in Visible to LWIR
Opportunity to work on Bird / Aviation Interaction Testing
Sam Siewert Drone Net, ICARUS Group Slide 12
2017/18 Team – ERAU ARI SponsoredERAU – Drone Net
– Dr. Sam Siewert, PI, SE/CE– Dr. Iacopo Gentilini, Co-I, ME– Dr. Stephen Bruder, Co-I, EE– Dr. Mehran Andalibi, Co-I, ME/AE
ERAU Graduate and Undergraduate– Jonathan Buchholz (ME Robotics)– David Olson (EE)– Garrison Bybee (SE)– Jonathan Buchholz – MS CESE RA (2018/19, 19/20)
CU Boulder – Embedded Systems Graduate– Vijoy Sunil Kumar – MS, ESE– Aasheesh Dandupally – ME, ESE– Soumyatha Gavvala – ME,ESE– Omkar Ajit Prabhu – ME, ESE
Research Collaboration Participants– ERAU Prescott Aviation Science, ATC and UAS programs– University of Alaska (ACUASI, Fairbanks)
Industry Advising/Collaboration Participants– Randall Myers, Mentor Graphics (PCB, CAD, Systems Fabrication)– Google (Applied, Faculty Research Proposal)
Sam Siewert Drone Net, ICARUS Group Slide 13
Next StepsCollaboration with ERAU Prescott Aviation Science– Jennah Perry (Assistant Prof. ATC)– Integration of Drone Net Detection, Tracking (Localization) and
Identification with ATC Grid FSS Management (FAA System Ops)ATC sUAS Integration Operations Concept VideoParallel talk submissions to AUVSI Denver 2018
– Curtis James (Chair, Applied Aviation Sciences, Prof. Meteorology)– NSF MRI for Dual-pole X-band RADAR (Ranger X)
Feasibility comparisons of sUAS Tracking and Identification with Active vs. Passive MethodsRange, Accuracy, Precision comparisons
ERAU UTM Collaboration - host field tests and sUAS Fly-ins– Summer 2017 (Educational Outreach Program)– ARI Research - Demonstration at Daytona– Plan Aviation Science Joint Testing in Prescott (target 2019)
Sam Siewert Drone Net, ICARUS Group Slide 14
SummaryDrone Net Will Produce Significant Detection, Tracking, Classification and Identification Data
Nodes Linked Together to SAN/NAS Database and File System on a Campus, Airport, Facility
Locations Can Uplink and Share Detection Signatures to Cloud for Improved Classification and Identification
Combined Machine Vision, Machine Learning and Big Data Analytics Challenge
Sam Siewert Drone Net, ICARUS Group Slide 15
Drone Net Related Talks & Publications1. S. Siewert, et al., “Drone Net, a passive instrument network driven by machine vision and machine
learning to automate UAS traffic management”, (submitted), AUVSI Xponential, Denver, Colorado, May 2018.
2. S. Siewert, S. Bruder, I. Gentilini, M. Andalibi, “Drone Net Architecture for UAS Traffic Management Multi-modal Sensor Networking Experiments”, (in preparation), IEEE Aerospace Conference [program], Big Sky, Montana, March 2018.
3. S. Siewert, et al. for ICARUS Group, “Drone Net – Big Data, Machine Vision and Learning Challenge and Opportunity”, invited speaker, 5th Annual Global Big Data Conference, Silicon Valley, August 29-31, 2017.
4. S. Siewert, et al. for ICARUS Group, “Drone Net: Using Tegra for Multi-Spectral Detection and Tracking in Shared Air Space”, [proposal, recorded talk], GPU Technology Conference, Silicon Valley, May 8-11, 2017.
5. S. Siewert, M. Vis, R. Claus, R. Krishnamurthy, S. B. Singh, A. K. Singh, S. Gunasekaran, “Image and Information Fusion Experiments with a Software-Defined Multi-Spectral Imaging System for Aviation and Marine Sensor Networks”, AIAA SciTech 2017 [program], Grapevine, Texas, January 2017.
6. S. Siewert, V. Angoth, R. Krishnamurthy, K. Mani, K. Mock, S. B. Singh, S. Srivistava, C. Wagner, R. Claus, M. Vis, “Software Defined Multi-Spectral Imaging for Arctic Sensor Networks”, SPIE Proceedings, Volume 9840, Algorithms and Technologies for Multispectral, Hyperspectral, and Ultraspectral Imagery XXII [program, h-index], Baltimore, Maryland, April 2016.
Sam Siewert Drone Net, ICARUS Group Slide 16