defining well clear separation for unmanned aircraft

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.


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


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


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%


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


• 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


• Background

• Defining Well Clear Separation for Unmanned Aircraft Systems Encountering Noncooperative Traffic

– Approach

– Low C-SWaP UAS Results

– Phase 1 UAS Results

• Summary

Outline


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


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


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


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


Outline

• Background


– Approach



• Summary


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.


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


• 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


Outline

• Background


– Approach



• Summary


• 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


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


• 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


• Aerosonde by Michael Paetzold, source image, Creative Commons Share-Alike License 3.0, text applied on bottom of image on slide 7

Bibliography (images)

https://commons.wikimedia.org/wiki/File:Aerosonde.jpg

https://creativecommons.org/licenses/by-sa/3.0/legalcode

defining well clear separation for unmanned aircraft

Documents