defining well clear separation for unmanned aircraft
TRANSCRIPT
AIAA AVIATION 2019ATOMS-15, UAS Traffic Management IV
Christine Chen
Defining Well Clear Separation for Unmanned Aircraft Systems Operating
with Noncooperative Aircraft
Noncooperative DWC - 2CCC 06/20/19
Legal Notices
DISTRIBUTION STATEMENT A. Approved for public release. Distribution is unlimited.
This material is based upon work supported by the National Aeronautics and SpaceAdministration under Air Force Contract No. FA8702-15-D-0001. Any opinions, findings,conclusions or recommendations expressed in this material are those of the author(s) and donot necessarily reflect the views of the National Aeronautics and Space Administration.
© 2019 Massachusetts Institute of Technology.
Delivered to the U.S. Government with Unlimited Rights, as defined in DFARS Part 252.227-7013 or 7014 (Feb 2014). Notwithstanding any copyright notice, U.S. Government rights inthis work are defined by DFARS 252.227-7013 or DFARS 252.227-7014 as detailed above.Use of this work other than as specifically authorized by the U.S. Government may violate anycopyrights that exist in this work.
Noncooperative DWC - 3CCC 06/20/19
Detect and Avoid
NMAC
Well Clear volume*
Suggestive Guidance to Remain
Well Clear
Directive Guidance Issued to Avoid
Collision
AcquireIntruders
Cooperative orNoncooperative
Aircraft
Avoidance Maneuver
Detect and Avoid is the capability to remain well clear and avoid collisions
NMAC – Near Mid-Air CollisionCooperative – Transponder/ADS-B equipped
RTCA-228 Phase 1 Concept of DAA:
*Notional VolumesDoes not indicate any temporal component that may exist
Noncooperative DWC - 4CCC 06/20/19
Well Clear
• DAA Well Clear is the UAS technical means of compliance for satisfying FAR 91.111 and 91.113
• Detect and Avoid (DAA) Systems provide surveillance, alerts, and guidance to Unmanned Aircraft Systems (UAS) to help them maintain well clear of other aircraft– Designed as an alternative means of compliance for see-and-avoid
regulations
– DAA systems are essential for safe integration of UAS into the National Airspace System (NAS)
FAR 91.111: ...not operate so close to another aircraft as to create a collision hazard
FAR 91.113: Vigilance shall be maintained … so as to see and avoid other aircraft … pilots shall alter course to pass well clear of other air traffic
FAR – Federal Aviation Regulations
Noncooperative DWC - 5CCC 06/20/19
Previous Work
• In 2015, Sense and Avoid Science and Research Panel (SARP) developed a Well Clear Definition for unmanned aircraft based on distance and time– RTCA SC-228 adopted definition
within Phase 1 DAA MOPS (DO-365)
– Phase 1 UAS is assumed to be equipped with ADS-B In, airborne active surveillance, and an air-to-air radar
– Definition driven by TCAS II interoperability, but applied to all intruders regardless of equipage
h = +450 ft
HMD = 4000 ft
P(NMAC|LoWC) = 2.2%*
𝝉𝝉𝒎𝒎𝒎𝒎𝒎𝒎* (estimated time to 4000 ft) = 35 sec
Phase 1 DAA well clear adopted by FAA
NMAC – Near mid-air collision (HMD ≤ 500 ft, VMD ≤ 100 ft)LoWC – Loss of Well ClearHMD – Predicted Horizontal Miss Distance
* Recommended P(NMAC|LoWC) = 5%
Noncooperative DWC - 6CCC 06/20/19
Phase 2 MOPS DAA Well Clear Objective
• Phase 1 DWC was designed to encompass TCAS II RA alerting thresholds− Resulting DWC is very safe but unnecessarily large for noncooperative
aircraft, which do not have TCAS
• Noncoop DWC will enable low C-SWaP UAS operations by reducing noncooperative surveillance requirements compared to RTCA SC-228 Phase 1− Low C-SWaP UAS are too small or budget-constrained to carry the large,
high-power radar required by the Phase 1 MOPS
• Noncoop DWC is anticipated to be applicable to both Phase 1 UAS and low C-SWaP UAS encountering noncooperative aircraft
Objective: Define alternative DAA Well Clear (DWC) for UAS encountering noncooperative aircraft (Noncoop DWC)
MOPS – Minimum Operational Performance StandardsRA – Resolution AdvisoryLow C-SWaP – low cost, size, weight, and power
Noncooperative DWC - 7CCC 06/20/19
• Typically operate at 500-10,000 ft MSL with speeds at or below 100 kts
• Extended operations in airspace classes D, E (non-terminal), or G (non-terminal) with transit operations in classes B and C
• Missions include air quality monitoring, aerial imaging and mapping, and law enforcement
• Can carry ADS-B and TCAS but may not be able to carry the Phase 1 radar (> 50 lbs)
Low C-SWaP UAS
Shadow (RQ-7B)Aerosonde
Noncooperative DWC - 8CCC 06/20/19
• Background
• Defining Well Clear Separation for Unmanned Aircraft Systems Encountering Noncooperative Traffic
– Approach
– Low C-SWaP UAS Results
– Phase 1 UAS Results
• Summary
Outline
Noncooperative DWC - 9CCC 06/20/19
DAA Well Clear Analysisfor Noncooperative Intruders
• Objective: identify and assess DAA Well Clear (DWC) candidates based on safety and operational suitability metrics
• Approach: use realistic encounters* in fast-time simulation to evaluate unmitigated and mitigated performance against noncooperative intruders
Unmitigated• Unmitigated analysis does not include
response to a Detect and Avoid system • Evaluates inherent latent collision risk
without mitigations • Narrows tradespace
Mitigated• Mitigated analysis includes DAA response
using DAIDALUS algorithm• Validates actual risk of each DWC
DWC4 - P(NMAC|WCV)
0.02
0.02
0.02
0.020.03
0.03
0.03
0.03
0.04
0.04
0.04
0.05
0.05
0.06
0.06
0.070.08
0.090.1
0 2 4 6 8 10 12 14 16 18 20
tpz (sec)
1000
1500
2000
2500
3000
3500
4000
4500
5000
5500
6000
HM
D (f
t)
DAA – Detect and Avoid
* One million encounters
Notional
𝝉𝝉𝒎𝒎𝒎𝒎𝒎𝒎*
DAIDALUS(With modified parameters for each DWC)
Notional
Noncooperative DWC - 10CCC 06/20/19
Low C-SWaP UAS Encounter Model
Encounter models generate random aircraft trajectories that are statistically representative of noncooperative trajectories observed from radar data
VFR – Visual Flight RulesNAS – National Airspace
• Shadow B• MQ-19
Aerosonde
Aerosonde Shadow
Altitudes (feet) Speeds (knots)
Ownship Type
UAS Trajectory Intruder Trajectory
Initialize aircraft geometry
Freq
uenc
y
Aerosonde Shadow
Estim
ated
PD
F
Estim
ated
PD
F
Noncooperative DWC - 11CCC 06/20/19
Phase 1 UAS Encounter Model
Ownship speed is the main difference between the encounter sets (Low C-SWaP: 40-100 kts, Phase 1: 40-250 kts)
VFR – Visual Flight RulesNAS – National Airspace
• Shadow B• Aerosonde• Cessna 208• Cessna 510• Reaper• Global
Hawk• Socata
Trinidad
Altitudes (feet) Speeds (knots)
Ownship Type
UAS Trajectory Intruder Trajectory
Initialize aircraft geometry
Freq
uenc
y
Estim
ated
PD
F
Estim
ated
PD
F
Noncooperative DWC - 12CCC 06/20/19
Metrics*
• Safety metrics indicate whether desired separation is achieved− Risk ratio and loss of well clear ratios: 𝐏𝐏(𝐍𝐍𝐍𝐍𝐍𝐍𝐍𝐍 𝐨𝐨𝐨𝐨 𝐋𝐋𝐨𝐨𝐋𝐋𝐍𝐍|𝐞𝐞𝐞𝐞𝐞𝐞𝐨𝐨𝐞𝐞𝐞𝐞𝐞𝐞𝐞𝐞𝐨𝐨, 𝐰𝐰𝐰𝐰𝐞𝐞𝐰𝐰 𝐦𝐦𝐰𝐰𝐞𝐞𝐰𝐰𝐦𝐦𝐦𝐦𝐞𝐞𝐰𝐰𝐨𝐨𝐞𝐞)
𝐏𝐏(𝐍𝐍𝐍𝐍𝐍𝐍𝐍𝐍 𝐨𝐨𝐨𝐨 𝐋𝐋𝐨𝐨𝐋𝐋𝐍𝐍|𝐞𝐞𝐞𝐞𝐞𝐞𝐨𝐨𝐞𝐞𝐞𝐞𝐞𝐞𝐞𝐞𝐨𝐨, 𝐰𝐰𝐰𝐰𝐞𝐞𝐰𝐰𝐨𝐨𝐞𝐞𝐞𝐞 𝐦𝐦𝐰𝐰𝐞𝐞𝐰𝐰𝐦𝐦𝐦𝐦𝐞𝐞𝐰𝐰𝐨𝐨𝐞𝐞)
− Ratio less than 1 indicates that the mitigated system reduces the risk of NMAC or LoWC; e.g., risk ratio of 0.1 indicates 90% reduction in risk
• Operational suitability metrics indicate the appropriateness and severity of alerts required to remain well clear− Alert ratio: 𝐏𝐏 𝐍𝐍𝐀𝐀𝐞𝐞𝐨𝐨𝐞𝐞 𝐞𝐞𝐞𝐞𝐞𝐞𝐨𝐨𝐞𝐞𝐞𝐞𝐞𝐞𝐞𝐞𝐨𝐨, 𝐰𝐰𝐰𝐰𝐞𝐞𝐰𝐰 𝐦𝐦𝐰𝐰𝐞𝐞𝐰𝐰𝐦𝐦𝐦𝐦𝐞𝐞𝐰𝐰𝐨𝐨𝐞𝐞)
𝐏𝐏(𝐍𝐍𝐍𝐍𝐍𝐍𝐍𝐍|𝐞𝐞𝐞𝐞𝐞𝐞𝐨𝐨𝐞𝐞𝐞𝐞𝐞𝐞𝐞𝐞𝐨𝐨, 𝐰𝐰𝐰𝐰𝐞𝐞𝐰𝐰𝐨𝐨𝐞𝐞𝐞𝐞 𝐦𝐦𝐰𝐰𝐞𝐞𝐰𝐰𝐦𝐦𝐦𝐦𝐞𝐞𝐰𝐰𝐨𝐨𝐞𝐞)
− Alert ratio measures the alert frequency relative to the nominal NMAC frequency; an alert ratio of 1 indicates the mitigated system only alerts when absolutely necessary
− Severity: Counts of Caution and Warning alerts
NMAC – Near mid-air collision (HMD ≤ 500 ft, VMD ≤ 100 ft)LoWC – Loss of Well Clear
Encounter Timeline:
NMAC
Well Clear volume*
Corrective Alert NoncooperativeIntruder
Warning Alert
Avoidance Maneuver
*Additional metrics were computed but are not shown here
Noncooperative DWC - 13CCC 06/20/19
Outline
• Background
• Defining Well Clear Separation for Unmanned Aircraft Systems Encountering Noncooperative Traffic
– Approach
– Low C-SWaP UAS Results
– Phase 1 UAS Results
• Summary
Noncooperative DWC - 14CCC 06/20/19
Unmitigated Analysis:Trade Space Down Selection
DWC1 DWC2 DWC3 DWC4 Phase 1HMD* 2000 ft 2200 ft 1500 ft 2500 ft 4000 ft𝜏𝜏𝑚𝑚𝑚𝑚𝑚𝑚* 15 s 0 s 15 s 25 s 35 sh* 450 ft 450 ft 450 ft 450 ft 450 ft
• DWC Candidates chosen based on trade study [1] of potential DWC using unmitigated collision risk (P) and maneuver initiation range (MIR) as metrics:− DWC1 achieves smallest MIR of
candidates with 5% unmitigated collision risk
− DWC2 is simple because it does not have a time component
− DWC3 was proposed for terminal area UAS operations
− DWC4 achieves an unmitigated collision risk smaller than 5%
HMD – Horizontal Miss Distance
𝜏𝜏𝑚𝑚𝑚𝑚𝑚𝑚*
*
[1] M. G. Wu, A. C. Cone, S. Lee, C. Chen, M. W. M. Edwards, and D. P. Jack, "Well Clear Trade Study for Unmanned Aircraft System Detect and Avoid with Non-Cooperative Aircraft," AIAA Aviation Conference, Atlanta, GA, 2018.
Noncooperative DWC - 15CCC 06/20/19
System Operating Characteristic for Low C-SWaP Encounters
• SOC allows simultaneous evaluation of safety and operational suitability
• Risk and LoWC ratio are largely insensitive to DWC definition
• HMD appears to have the largest effect on alert ratio− DWC1 and DWC3 have the
same 𝜏𝜏𝑚𝑚𝑚𝑚𝑚𝑚*, but DWC1 alerts more frequently
HMD – Horizontal Miss Distance
0 20 40 60 80 100 120
Alert Ratio
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
Ris
k or
WC
V R
atio
Risk RatioWCV Ratio
DWC3
DWC1
DWC4
DWC3
DWC1
DWC4
DWC2
DWC2Phase 1
Phase 1
DWC1 DWC2 DWC3 DWC4 Phase 1HMD* 2000 ft 2200 ft 1500 ft 2500 ft 4000 ft𝜏𝜏𝑚𝑚𝑚𝑚𝑚𝑚* 15 s 0 s 15 s 25 s 35 s
LoWC Ratio
Ris
k or
LoW
CR
atio
Noncooperative DWC - 16CCC 06/20/19
• DWC 1, 2, 3 are largely insensitive to reduced surveillance ranges
• DWC 4 and Phase 1 experience large increases in risk ratio and loss of well clear ratio when surveillance range is reduced (see 2 NM blue bars)
Effect of Surveillance Range
NMAC Risk Ratios Loss of Well Clear Ratios
New DWC candidates support surveillance ranges down to 2 NM
DWC1 DWC2 DWC3 DWC4 Phase 10
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
Ris
k R
atio
No Limit
4 NM
3 NM
2 NM
DWC1 DWC2 DWC3 DWC4 Phase 10
0.1
0.2
0.3
0.4
0.5
0.6
LoW
C R
atio
No Limit4 NM3 NM2 NM
Note: Different y-axis scales
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
0.6
0.4
0.5
0.3
0.2
0.1
0
Noncooperative DWC - 17CCC 06/20/19
Outline
• Background
• Defining Well Clear Separation for Unmanned Aircraft Systems Encountering Noncooperative Traffic
– Approach
– Low C-SWaP UAS Results
– Phase 1 UAS Results
• Summary
Noncooperative DWC - 18CCC 06/20/19
• Trends similar to results for low C-SWaP UAS– HMD appears to have
the largest effect on alert ratio
System Operating Characteristic for Phase 1 UAS Encounters
HMD – Horizontal Miss DistanceFOV – Field of View
0 20 40 60 80 100
Alert Ratio
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.220.22
Ris
k or
WC
V R
atio
Risk RatioWCV Ratio
Phase 1
Phase 1
DWC4
DWC4
DWC2
DWC2
DWC1
DWC1
DWC3
DWC3
DWC1 DWC2 DWC3 DWC4 Phase 1HMD* 2000 ft 2200 ft 1500 ft 2500 ft 4000 ft𝜏𝜏𝑚𝑚𝑚𝑚𝑚𝑚* 15 s 0 s 15 s 25 s 35 s
0.10
0.02
0.00
0.05
0.10
0.15
Limited FOV Full FOVR
isk
Rat
io
Induced Unresolved
Phase 1 NMAC Risk Ratios
Ris
k or
LoW
CR
atio LoWC Ratio
Noncooperative DWC - 19CCC 06/20/19
Effect of Ownship Speed on Safety
• Sensitivity of DWC definition to ownship speed:− HMD* has the larger
impact at low speeds
− 𝝉𝝉𝒎𝒎𝒎𝒎𝒎𝒎* has little effect on safety at low speeds
− Smaller 𝝉𝝉𝒎𝒎𝒎𝒎𝒎𝒎* reduces the safety ratios at high speeds
40 – 100 knots 100 – 150 knots
150 – 200 knots 200+ knots
𝝉𝝉𝒎𝒎𝒎𝒎𝒎𝒎 may not be necessary in DAA Well Clear definition for noncooperative intruders
Loss of Well Clear Ratios
Noncooperative DWC - 20CCC 06/20/19
• Performed unmitigated and mitigated analysis of four candidate well clear definitions for low C-SWaP UAS against noncooperative intruders– NMAC and LoWC risk not sensitive to DWC parameters– Safety and operational suitability not dependent on 𝜏𝜏𝑚𝑚𝑚𝑚𝑚𝑚*
• Indicates 𝜏𝜏𝑚𝑚𝑚𝑚𝑚𝑚* may not be necessary
• Future work:– Development and validation of sensor requirements based on the noncoop
DWC– Additional human factors evaluation of noncoop DWC– Additional safety analyses in the presence of sensor noise
Conclusions
NMAC – Near mid-air collision (HMD ≤ 500 ft., VMD ≤ 100 ft.)LoWC – Loss of Well Clear
SC-228 selected DWC2 (2200 ft, 450 ft, 0 𝛕𝛕𝐦𝐦𝐨𝐨𝐦𝐦*) for low C-SWaP UAS and Phase 1 UAS encountering noncooperative aircraft
Noncooperative DWC - 21CCC 06/20/19
• Aerosonde by Michael Paetzold, source image, Creative Commons Share-Alike License 3.0, text applied on bottom of image on slide 7
Bibliography (images)