1

Operational Impact of RA Downlink:

Results of Real-time Simulation

Stanislaw Drozdowski & Doris M. Dehn EUROCONTROL, Brussels, Belgium

Jürgen Teutsch National Aerospace Laboratory of the Netherlands (NLR)

Bernd LorenzEUROCONTROL, Budapest, Hungary

2

Presentation Overview

1. Setting the Scene

� TCAS

� Controller/Pilot tasks

� FARADS Project

2. RADE-2 Experiment

� Objectives

� Participants

� Experimental design

� Results

3. Conclusions & discussions

3

Traffic Alert & Collision Avoidance System (TCAS)

� Last defence against mid-air collisions

� Majority of commercial aircraft equipped

� Issues Resolution Advisories (RA) to the pilots

� Following RAs is mandatory

� TCAS = ACAS

4

Controller – pilot tasks: Nominal situation

� Responsible for separation provision

� Issue clearances or instructions

� Follows ATC clearances and instructions

5

Controller – pilot tasks: RA Issued

� No longer responsible

� Shall not issue clearances or instructions

� Follows the RA

� Reports RA to ATC (as soon as possible)

X

What if RA is not reported?

6

Feasibility of ACAS RA Downlink Study (FARADS)

Überlingen accident

2002 2003 2004 2005 2006

Strategic Action Plan

RA Downlink Experiment 1

FARADS Technical Study

RA Downlink Experiment 2

FARADS Latency Study

FARADS Safety Study

7

RA Downlink Experiment 2A: Objectives

� Assess the impact of RA downlink on

� controller performance (clearances to RA aircraft, ability to separate

traffic)

� understanding of the traffic situation

� perceived workload

� Investigate controllers‘ acceptance of RA downlink

8

Date & Location

� 5 October – 1 December 2005

� EUROCONTROL Experimental Centre, Brétigny

� eDep simulation platform

9

Participants

� 12 area controllers from 4 European ACCs

� 3 – 28 years of experience (average 10.8 years)

Maastricht - 6

Rome - 2

Langen - 2

Marseille - 2

10

Experimental variables

� RA downlink (absent vs.

present)

� Timeliness of pilot report

(timely vs. delayed)

� Controller’s role

(executive vs. planner)

11

Experimental variables

� RA downlink (present vs.

absent)

� Timeliness of pilot report

(timely vs. delayed)

� Controller’s role

(executive vs. planner)

Immediate report

Report after COC

12

Experimental variables

� RA downlink (present vs.

absent)

� Timeliness of pilot report

(timely vs. delayed)

� Controller’s role

(executive vs. planner)

Executive

Planner

13

Control variable: Cause of RA

High vertical rate

Correct clearance and implementation

but ACAS predicts conflict due to

high vertical rate

Controller or pilot error:

Incorrect clearance or clearance

incorrectly implemented

14

Measurements

� Controller Performance

� Number of clearances to RA aircraft

� Number of losses of separation

� Situation Awareness (SA)

� Recollection of RA event

� SA on-line probe

� SASHA questionnaire

� Workload: NASA Task Load Index

� Controller Acceptance

� Concept

� Procedures

� HMI

15

� 48 simulation runs

� Simulation (traffic, RA event, pilot behaviour) judged as realistic by

participants

Realisation of experimental plan

12

1212

12 RA Downlink & Timely Pilot Report

RA Downlink & Delayed Pilot Report

No RA Downlink & Timely Pilot Report

No RA Downlink & Delayed Pilot Report

16

Results: Contradictory clearances

� Two contradictory clearances issued to aircraft involved in RA

� Both in condition without RA downlink

� One in timely, one in delayed condition

17

Example of contradictory clearance (1)

� Conditions:

� No RA downlink

� Delayed pilot report

� RYR339 climbing FL310

� RA triggered due to

RYR339’s high vertical

rate

� RYR339 – limit climb RA

� EIN435 – climb RA

18

Example of contradictory clearance (2)

� EIN435 responds to the

climb RA

� Causes a follow-up

conflict with KLM4547

� STCA triggered

� Controller unaware of RA

tells EIN435 to maintain

FL320

19

Results: Losses of separation

� Losses of separation

exclusively in follow-up

conflicts (i.e. involving

at least one “RA aircraft”)

� No evidence that RA downlink supports planning of post-conflict

situation

� No evidence for “cognitive tunnelling” on RA event

Timely Pilot Report

Delayed Pilot Report

0

4

8

RA downlinkNo RA downlink

20

Results: Recollection of RA event

� SA memory probe with 10 items on various aspects of RA event

� No effect of RA downlink, if RA was caused by High Vertical Rate

� Higher scores with RA downlink, if RA was caused by

pilot/controller error.

High VerticalRate

SA

mem

ory

pro

be

tota

l sco

re

0

2

4

6

8

10

Downlink No Downlink

n.s. p = 0.02

Other Reason

SA

mem

ory

pro

be

tota

l sco

re

0

2

4

6

8

10

Downlink No Downlink Downlink No Downlink

n.s. p = 0.02

21

Results: SA on-line probe

� Pilots of aircraft not involved in the RA requested level or route

changes, traffic information, etc.

� No evidence that RA downlink limits controller focus to RA event

Optimal

No RA

Optimal

RA

Suboptimal

No RASuboptim

al

RA

02468

1012141618

Executive

Planner

22

Controller Acceptance

� Majority of participants see benefits in RA downlink

� Increased SA

� Lower likelihood of contradictory clearances

� However…

� Benefits restricted to RAs caused by pilot/controller error

� Some concerns regarding pilot/controller responsibility

23

Conclusions

� Contradictory clearances to RA aircraft only observed if RA

downlink was not provided

� Better recollection of RA events with RA downlink; however,

effect restricted to those caused by controller/pilot error

� No evidence for “cognitive tunnelling” as a consequence of RA

downlink

� No evidence that RA downlink helps planning of post-conflict

situation

� General acceptance of RA downlink by participants

www.eurocontrol.int/rawww.eurocontrol.int/ra--downlinkdownlink

Email: raEmail: ra--downlink@eurocontrol.intdownlink@eurocontrol.int