Flexible Innovation Testbed (SAFIT TM ) Final Presentation - - PowerPoint PPT Presentation
Safe Autonomy Flexible Innovation Testbed (SAFIT TM ) Final Presentation September 6, 2017 Sally C. Johnson Jesse C. Couch Adaptive Aerospace Group, Inc. Hampton, VA sjohnson@adaptiveaero.com Outline Requirements Capture SAFIT TM
September 6, 2017
Page 2
Page 3
An Unmanned Aircraft System (UAS) platform for safely testing NASA’s unproven autonomy applications
– Learning, adaptation, non-deterministic algorithms – Operation in complex environments – Multi-vehicle cooperation
projects
– Autonomy Incubator – UAS Integration in the NAS – Adaptive Controls and Controls Upset Research – Safety Critical Avionics Systems Research
Page 4
– Design UAS testbed platform tailored to support NASA’s autonomy research – Demonstrate feasibility of key innovative features
– Detailed design of SAFITTM UAS testbed
– SAFIT-WrapTM prototype development and simulation demonstration of
– Procure/integrate key hardware components and demonstrate flow of data – Build prototype of vehicle (under cost sharing)
Page 5
– Design UAS testbed platform tailored to support NASA’s autonomy research – Demonstrate feasibility of key innovative features
Detailed design of SAFITTM UAS testbed
SAFIT-WrapTMprototype development and simulation demonstration of
Procure/integrate key hardware components and demonstrate flow of data Build prototype of vehicle (under cost sharing)
Page 6
– 10 minute hover with 3-lb payload
– 30 minute cruise at 40 mph with 6-lb payload
Page 7
– 10 minute hover with 3-lb payload
– 30 minute cruise at 40 mph with 6-lb payload
performance
Page 8
– 10 minute hover with 3-lb payload
– 30 minute cruise at 40 mph with 6-lb payload
performance
– Pre-planned – Real-time
Page 9
– 10 minute hover with 3-lb payload
– 30 minute cruise at 40 mph with 6-lb payload
performance
– Pre-planned – Real-time
Page 10
wide range of mission scenarios – Vertical Takeoff and Landing (VTOL)
– Conventional Takeoff and Landing (CTOL) configuration
surfaces
power options
– Internal combustion generator vs all electric
– 2 wing panels, tail booms, separable empennage, 4 rotor trunnions – Access panels for payload modules
Page 11
Page 12
4x Hacker A40-10L-14p 4x Castle Phoenix Edge 75A 2x 16000mah 6s2p Lipo (22.2V nom) 2x 15x10E, 2x 15x10EP 2” L 1.6” OD 0.6lb 6.8” x 2.9” x 2.7” 4.2lb
Page 13
Page 14
Page 15
– Traffic avoidance – Obstacle avoidance – Geospatial containment – Flight envelope protection
– Situation Awareness – Alerting status
Page 16
Potential Solution
Partitioning
wrapper
application
Wrapper provides
if needed
Checks outputs for
Plans optimal solution using
Reliable Solution
Page 17
(NMAC) cylinder – Radius: 500 ft – Half-height: 100 ft
a predefined “Well Clear” distance from traffic
– Maneuvering in cluttered environments – Slower speeds than civil transports – Nimble maneuvering
Page 18
– NASA’s UAS Traffic Management (UTM)
– Where multiple UAS are operating
– Onboard separation assurance may be needed for non-normal and off- nominal events
– Suburban and rural UAS traffic
Page 19
Volumes developed
accuracy at low altitudes
detecting conflicts based on ability to turn at 30o per second
traffic avoidance algorithm
Page 20
paths tangentially to obstacles
Page 21
Page 22
Page 23
apart, including multi-vehicle conflicts
– Tuning of building look-ahead time vs. buffer size – Increased look-ahead time may preclude entering curved streets or approaching T intersections
– 8 s traffic look-ahead time required – 4 s traffic look-ahead time resulted in several NMACs and building collisions 70 ft 50 ft UAS maneuvering corridor
Page 24
adequate to protect against Well Clear violations
– Necessary due to navigation/position uncertainty – Initial maneuvers were sometimes insufficient to avoid Well Clear violation
maneuvering in an urban environment
– The Well Clear Volume was shown to protect against NMACs in challenging scenarios
– Appropriate for simple encounters in low traffic density – Shown to be effective in complex multi-vehicle conflicts – Suitable as supplement to UTM
Page 25
Page 26
– Safe flight evaluation of unproven autonomy applications – Full-service support:
Page 27
– to support safe operation of multiple UAS – across a wide range of commercial and research missions – including Beyond Visual Line of Sight operations – certified for commercial UAS operations under a future standard
– Marketed to commercial UAS manufacturers as an optional flight management system – Marketing of high-integrity core functionality for other developers to build upon
Page 28
– Open source software is continually updated with new features, such as
– Hardware and connections are unreliable
– Compass “inconsistency” on new hardware
– Unstable degraded flight
and dangerous even manually flying – Fly-aways
without warning and fly away
Page 29
Ultra-high-integrity
satisfies limited safety properties
testing for correct implementation High-integrity
testing for correct implementation Low-pedigree
Partitioning
must be separated from complex, unproven code
High-Level Executive Low-Level Functions Mid-Level Logic
Formal methods
with unexpected behavior
Page 30
– Develop a UAS testbed capability to support a wide range of NASA’s research projects, including autonomy research – Initiate development of a flight management system for safe implementation
– No easy answers, but we believe a high-integrity version of SAFITTM can help
– Maneuvering autonomously – Single operator handling multiple UAS – Beyond Visual Line of Sight operations