Prototyping Collision Avoidance for suas

Transcription

1 Prototyping Collision Avoidance for Michael P. Owen 5 December 2017 Sponsor: Neal Suchy, FAA AJM-233 DISTRIBUTION STATEMENT A: Approved for public release; distribution is unlimited.

2 Trends in Unmanned Aircraft Systems New and Emerging Applications Continued UAS Market Growth FAA Commercial UAS Exemptions Law Enforcement Search and Rescue Package Delivery Agricultural Continued Demand for Access Threats and Challenges FAA under pressure as clamor for small commercial drones grows Senators to question FAA chief after missed drone deadline Businesses pressure FAA on drones FAA releases 1325 UAS reported sightings from between Nov and Jan k+ Registered UAS Continued pressure and challenges integrating UAS into the NAS Collision Avoidance Prototyping - 2

3 Current Small UAS Rule Limitations Current Restrictions Limited to visual line of sight Can only operate in daylight Segregated from manned aircraft 1:1 operator to UAS ratio Maximum altitude 400 ft Weight < 55 lbs. Max speed 100 mph Line of sight limit ~1/2 mile Moving beyond small UAS rule requires additional tools to validate safety and performance Collision Avoidance Prototyping - 3

4 Technologies for Enabling Safe Operations Beyond Current Rule Surveillance Options Decision Support Architecture & System Analysis Onboard ADS-B, EO/IR, Radar, Acoustic External Radar, EO/IR Sensor networks Collision avoidance Path planning Geofencing Mission-specific guidance Centralized vs. distributed Reliable command & control Levels of autonomy Low-cost testbeds enable the development of technologies necessary for safe and autonomous operations Collision Avoidance Prototyping - 4 EO/IR = Electro-optical / Infrared

5 Testbed Architecture Sensors & Payloads Platforms System Test Suite ADS-B, Algorithms, Video Telemetry/ Command Link Decision Support Logic Sensor Emulator Sim-over- Live Aircraft SUAS Information Messaging Bus Architecture (SIMBA) Virtual Pilot Workstation COTS GCS Common Air Picture Prototype Operator Displays & Control Ground Sensor Information/Feed Ability to explore distributed and centralized control Architecture permits rapid integration of new sensors and systems on low-cost platforms System under test implemented on platform or on ground Supports live, simulation, and sim-over-live environment Collision Avoidance Prototyping - 5 COTS GCS = Commercial Off-the-Shelf Ground Control Station

6 Flight Test Setup at Fort Devens, MA Fort Devens: Turner Drop Zone 0.5 Mile 0.8 Mile R4102 A/B Restricted Airspace MIT LL Collision Avoidance Prototyping - 6

7 Outline Testbed motivation and overview Prototype collision avoidance capability Flight demonstration Summary Collision Avoidance Prototyping - 7

8 Airborne Collision Avoidance System (ACAS X) FAA TCAS Program Office funding since 2009 Next Generation System RTCA SC-147 developing MOPS ACAS Xa 2018 ACAS Xu 2020 Dynamic Uncertainty State Uncertainty ACAS X Goals: Improve safety & unnecessary alerts Support reduced separation operations Extend collision avoidance to UAS Streamline development process Additional Surveillance Sources Improved and Flexible Threat Logic Standard Interface ACAS X approach leveraged and adapted to create capability Collision Avoidance Prototyping - 8

9 Prototype Logic Overview Autonomous Onboard Collision Avoidance Raspberry Pi 3 ACAS-X horizontal collision avoidance logic, adapted for - collision avoidance Minimum Objective: 150 ft horizontal separation Manned aircraft avoidance (subject to Devens airspace constraints) Minimum Objective: 750 ft horizontal separation Onboard real-time threat assessment Onboard auto-response Appropriate for fixed-wing and multi-rotors Horizontal and vertical maneuvers August demonstration focused on horizontal Collision Avoidance Prototyping - 9

10 Surveillance Architectures Evaluated Shared Telemetry External: Cloud-Based Radar Feed Telemetry Link Telemetry Link Testbed ASR-9 Radar Tracker and Geographic Filter Cellular Data Link Telemetry Link Telemetry Link Testbed EO/IR Onboard Surveillance ADS-B In Onboard Radar External: Cloud-Based Cellular Surveillance Cellular Linked GPS Cloud Server and Tracker Cellular Data Link Telemetry Link Testbed Telemetry Link Testbed Collision Avoidance Prototyping - 10

11 Outline Testbed motivation and overview Prototype collision avoidance capability Flight demonstration Avoidance of cooperative Avoidance of manned aircraft Avoidance of non-participant Summary Collision Avoidance Prototyping - 11

12 Multi-UAS Autonomous Collision Avoidance 700 ft MSL 700 ft MSL Ft Devens Turner Drop Zone Restricted Airspace Scenario Notes Simple scenarios to highlight collision avoidance timing v. collision avoidance designed to allow reduced separation compared to manned aircraft Collision avoidance outcomes dictated by timing of each aircraft N Equipment: Onboard GPS Onboard Collision Avoidance Equipment: Onboard GPS Onboard Collision Avoidance Intruder Surveillance Source: Shared Telemetry Collision Avoidance Prototyping - 12

13 Multi-UAS Autonomous Collision Avoidance: Results Horizontal separation (ft) Shared telemetry enables robust collision avoidance between operating in same vicinity Intruder Surveillance Source: Shared Telemetry Achieved 586 ft minimum separation 150 ft minimum separation objective Collision Avoidance Prototyping - 13

14 Autonomous Multi-UAS Avoidance of Manned Aircraft 1000 ft MSL Ft Devens Turner Drop Zone Restricted Airspace Manned Aircraft Scenario Notes Automatic response and return to mission on both Space limitations and timing logistics dictate nominal mission 1000 ft MSL N 1500 ft MSL Pilot deviation or radar surveillance noise may prompt to maneuver differently Nominal separation 150 ft Manned Cessna Intruder Surveillance Source: Cloud-Based Radar Feed Equipment: Onboard GPS Onboard Collision Avoidance Not avoiding other Equipment: Onboard GPS Onboard Collision Avoidance Not avoiding other Equipment: Transponder No ADS-B Out No Collision Avoidance Collision Avoidance Prototyping - 14

15 Autonomous Multi-UAS Avoidance of Manned Aircraft: Results Flight Test Recorded Data Airborne perspective of RV7 Radar surveillance enables autonomous avoidance of intruder aircraft Additional refinement on prototype radar tracker warranted Achieved 1207 (blue) and 1982 ft (red) horizontal separation from manned aircraft 750 ft minimum separation objective Intruder Surveillance Source: Cloud-Based Radar Feed Collision Avoidance Prototyping - 15

16 Demonstration of Autonomous Avoidance During Simulated Search and Rescue 700 ft MSL Ft Devens Turner Drop Zone Restricted Airspace Scenario Notes is performing automated search of eastern portion of Turner Drop Zone Intruder models a non-participant in the search operations blundering through operational area 700 ft MSL Intruder GPS Message Relay 713 Miles TDZ MITLL N JHUAPL - Baltimore 713 Miles Equipment: Onboard GPS Onboard Collision Avoidance Performing automated search Equipment: Cellphone reporting GPS location No Telemetry Link to Testbed Manually piloted to harass Intruder Surveillance Source: Cloud-Based Cellular Surveillance Collision Avoidance Prototyping - 16

17 Demonstration of Autonomous Avoidance During Simulated Search and Rescue: Results Flight Test Recorded Data Horizontal Separation Horizontal separation (ft) 5 passes of intruder 150 ft minimum separation objective exceeded in each case Automated search coverage still covers majority of area Intruder Surveillance Source: Cloud-Based Cellular Surveillance Prototype logic enabled safe avoidance of intruder and continuation of search and rescue mission Collision Avoidance Prototyping - 17

18 Outline Testbed motivation and overview Prototype collision avoidance capability Flight demonstration Summary Collision Avoidance Prototyping - 18

19 Collision Avoidance R&D Path Forward Initial flight tests Logic Optimization Expand adaptation for terminal areas, higher altitudes Incorporate - surveillance Regional secure surveillance network Smart-phonebased surveillance Additional - surveillance options Standardize Xu logic for smalls Link requirements for cloud-based surveillance Collision Avoidance Prototyping - 19

20 Summary Lincoln Laboratory Testbed architecture enables rapid prototyping and evaluation of collision avoidance technologies Demonstrated prototype collision avoidance system leveraging ACAS X logic Baseline collision avoidance architecture will require Optimization and tailoring of collision avoidance system Surveillance requirements development EO/IR, Radar (onboard or ground-based), UTM, Standards development for safety and separation requirements Collision Avoidance Prototyping - 20 UTM = UAS Traffic Management

21 Legal Notices DISTRIBUTION STATEMENT A. Approved for public release: distribution unlimited. This material is based upon work supported under Air Force Contract No. FA C-0002 and/or FA D Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the U.S. Air Force Massachusetts Institute of Technology. Delivered to the U.S. Government with Unlimited Rights, as defined in DFARS Part or 7014 (Feb 2014). Notwithstanding any copyright notice, U.S. Government rights in this work are defined by DFARS or DFARS as detailed above. Use of this work other than as specifically authorized by the U.S. Government may violate any copyrights that exist in this work. Collision Avoidance Prototyping - 21