IATA Guidance Assessment of Pilot Compliance To TCAS

Guidance Material
Performance assessment of pilot

compliance to Traffic Alert and
Collision Avoidance System (TCAS)
using Flight Data Monitoring (FDM)
Edition 1 January 2019 1

Assessment of Pilot Compliance to TCAS
“DISCLAIMER. The information contained in this document is subject to constant review in the light
of changing government requirements and regulations. No subscriber or other reader should act
on the basis of any such information without referring to applicable laws and regulations and
without taking appropriate professional advice. Although every effort has been made to ensure
accuracy, the International Air Transport Association (IATA), the European Organization for the
Safety of Air Navigation (EUROCONTROL), and any other contributors to this publication shall not
be held responsible for any loss or damage caused by errors, omissions, misprints or
misinterpretation of the contents hereof. Furthermore, IATA, EUROCONTROL, and any other
contributors to this publication expressly disclaim any and all liability to any person or entity,
whether a purchaser of this publication or not, in respect of anything done or omitted, and the
consequences of anything done or omitted, by any such person or entity in reliance on the
contents of this publication.
Other contributors’ opinions expressed in this publication do not necessarily reflect the opinion of
the International Air Transport Association and EUROCONTROL. The mention of specific
companies, products in this publication does not imply that they are endorsed or recommended by
the International Air Transport Association and EUROCONTROL in preference to others of a similar
nature which are not mentioned.
© International Air Transport Association and EUROCONTROL, 2019. All Rights Reserved. No part
of this publication may be reproduced, recast, reformatted or transmitted in any form by any
means, electronic or mechanical, including photocopying, recording or any information storage and
retrieval system, without the prior written permission from:
Senior Vice President
Safety & Flight Operations
International Air Transport Association
800 Place Victoria, P.O. Box 113
Montreal, Quebec
CANADA H4Z 1M1
2 Edition 1 January 2019

Contents
Abbreviations and Acronyms ...................................................................................................................... 4
Objective .............................................................................................................................................................. 5
Introduction......................................................................................................................................................... 5
Monitoring Operational Safety Issues (TCAS RA Event) .................................................................. 6
FOQA/FDM Review Group ........................................................................................................................ 7
Key Performance Indicators ........................................................................................................................ 7
Conclusion........................................................................................................................................................... 8
Recommendations ........................................................................................................................................... 8
Supporting information .................................................................................................................................. 9
TCAS Advisories Classification.............................................................................................................. 9
TCAS Operational Assessment .......................................................................................................... 10
Expected TCAS RA Response ............................................................................................................. 10
Challenges Associated With Pilot Responses .............................................................................. 11
Flight Crew Action..................................................................................................................................... 12
Reporting ...................................................................................................................................................... 12
Opposite Initial Pilot Response (OIPR).............................................................................................. 18
Cooperating with ATC Authorities (Air Navigation Service Providers - ANSPs) ............. 19
ACAS Xa ........................................................................................................................................................ 19
References ....................................................................................................................................................... 20
Appendix A – Eurocontrol operational safety study: TCAS RA not followed ........................ 21

Abbreviations and Acronyms

A/C Aircraft
ACAS Airborne Collision Avoidance System
ANSP Air Navigation Service Provider
AP Auto Pilot
ASR Air Safety Report
ATC Air Traffic Control
CFIT Controlled Flight into Terrain
CPA Closest Point of Approach
FD Flight Director
FDA Flight Data Monitoring
FDAP Flight Data Analysis Program
FDM Flight Data Monitoring
FDR Flight Data Recorders
FOQA Flight Operational Quality Assurance
GPWS Ground Proximity Warning System
ICAO International Civil Aviation Organization
MOR Mandatory Occurrence Report
OIPR Opposite Initial Pilot Response
PNF Pilot Not Flying
RA Resolution Advisory
RARR Resolution Advisory Required Rate
SARPs Standards and Recommended Practices
SOPs Standard Operating Procedures
TA Traffic Advisory
TCAS Traffic Alert and Collision Avoidance System
TAWS Terrain Avoidance and Warning System
VSI Vertical Speed Indicator

Objective
The purpose of this document is to comply with the request of the IATA Safety Group action item 25/10 to
prepare a guidance document on the subject of using flight data monitoring (FDM) to monitor and address
response non-compliance with airborne collision resolution advice issued by Traffic Alert and Collision
Avoidance System II (TCAS II). This Guidance Material was prepared jointly by IATA and EUROCONTROL.
Currently, TCAS II is the only implementation that meets the Airborne Collision Avoidance System (ACAS),
which is the International Civil Aviation Organization (ICAO) terminology and is included in the ICAO
Standards and Recommended Practices (SARPs). Therefore, the term ACAS II is typically used when
referring to the standard or concept and TCAS II when referring to the implementation.
Note: The acronyms TCAS and ACAS are used interchangeably in the document.
Introduction
TCAS is an airborne system designed to provide collision advice against suitably equipped intruders and
to increase awareness of the flight crew of nearby aircraft.
There are currently two types of TCAS equipment in operational use: TCAS I and TCAS II. Both provide the
flight crew with a cockpit display indicating the presence of a transponding ‘intruder’.
TCAS I equipment warns of potential conflicts by providing a traffic advisory (TA), which is announced
“Traffic, Traffic”, but it does not provide any resolution advice.
TCAS II equipment provides a second level of alert called a resolution advisory (RA). This alert directs the
flight crew to make a vertical maneuver to avoid the intruding aircraft.
Note: TCAS II does not provide RAs in the lateral direction.
Currently, there are three versions of TCAS II in operation:
• Version 6.04a is still mandated or allowed on some aircraft operating in US airspace. In Europe
version 6.04a may be found on aircraft outside the current European mandate (i.e. either military or
those below the mandated weight or number of passenger seats thresholds). Version 6.04a is not
compliant with the ICAO ACAS SARPS (Annex 10, volume IV).
• Version 7.0 is still mandated or allowed on many aircraft operating in US airspace and other parts of
the world. In Europe version 7.0 may be encountered on aircraft outside the current European
mandate (i.e. either military or those below the mandated weight or number of passenger seats
thresholds). Version 7.0 was the first TCAS II version to be compliant with the ICAO ACAS SARPS
(Annex 10 volume IV); however, since 2010 (amendment 85) only version 7.1 is compliant.
• Version 7.1 is the only ACAS version meeting the current requirements of ICAO and European
mandates. It was developed based on an extensive analysis of version 7.0 performance to address
two safety issues.

One of the drivers for this paper is the study conducted by EUROCONTROL, including the online survey
carried out by IATA that elicited 3,800 responses from flight crew in 90 countries.
The results show that while 37 percent of respondents experienced an RA in the last 5 years, 15 percent
of them did not follow the RA. The outcome of the survey can be found in Appendix A to this document.
For TCAS II system to work as designed and to resolve a risk of midair collision, immediate and correct
flight crew response to TCAS resolution advisories is important. As any delayed or incorrect flight crew
response negates the effectiveness of the RA, their actions will be the most important single factor
affecting the performance of TCAS system.
Operational experience has shown that the correct response by flight crew is dependent on the
effectiveness of the initial and recurrent training in TCAS procedures. Moreover, in an event of an RA, the
flight crew shall report on the flight completion the circumstances by means of Air Safety Report (ASR) /
Mandatory Occurrence Report (MOR). The reporting of the RA by the flight crew is expected to open the
parenthesis in the airline safety department in the identification of the contributing factors that will help to
assess and understand the root causes of the incidents.
A complimentary approach to the ASR is the Flight Data Monitoring (FDM) system, also known as Flight
Data Analysis Program / Flight Operational Quality Assurance (FDAP/FOQA) programs, which measures
and monitors how aircraft is being flown by the analysis of data downloaded from an aircraft’s on-board
flight data recorders at the end of every flight. Events like TCAS RAs are captured by such programs and
because of their small number, all RAs must be assessed to formulate mitigation actions.
Note: Pilots and flight crew are used interchangeably in the document.
Monitoring Operational Safety Issues (TCAS RA Event)

The structure of the organization has to be taken into account when setting up safety monitoring
arrangements. This document suggests the use of FOQA/FDM for monitoring and follow up of TCAS RA
events. The FOQA/FDM program detects a TCAS RA in the flight data. This data will be used to analyze
RAs associated with close encounters and problematic areas such as approaches to closely-spaced
parallel runways as well as flight crew compliance to RAs, the data can also detect whether the flight crew
are not responding to RAs, delaying response, or responding in the wrong direction. Once the
assessment is made, identification and development of mitigation measures will be put in place.
The following questions will help in assessing, controlling or eliminating such events:
1. Was an ASR/MOR filed by the flight crew?

The analyst should check if an ASR/MOR was filed. If yes, typically this allows the flight crew to be
approached and if necessary debriefed. If no, typically the pilot can only be approached by the
FOQA/FDM gatekeeper (depending on airline policies and procedures). On first contact, the flight crew
must be asked to file an ASR/MOR and it is expected that the pilot will do so. If the flight crew refuses, this
is a regulatory non-compliance with mandatory reporting. Airlines vary in how this scenario is dealt with,
but typically a review group would be convened consisting of the safety manager, a pilot’s representative

and a flight operations manager. The group would decide whether the refusal to file an MOR constitutes
sufficient reason to withdraw the flight crew’s identity protection under the FOQA/FD scheme.
2. Was the response to the TCAS RA within time and performance parameters
set by the company?
The analyst reviews the flight data to determine if the response to the TCAS RA was within the required
time limit and to the required vertical rate was followed. This analysis will be attached to the ASR/MOR in
the company SMS database, allowing the safety officer/manager to determine if the event can be closed
or requires follow up with the pilots. In the latter case, an effective method of debriefing is for the safety
manager to review the flight data animation with both pilots, if possible with a training captain from the
fleet. The debriefing should be conducted in a positive spirit as a learning opportunity. If the lessons are
assimilated, the report can usually be closed. In some circumstances, it may be necessary to reinforce the
lesson with a dedicated simulator training session. In the case where the pilot has not filed a report and
his/her identity remains protected a debriefing will be conducted by the FOQA/FDM gatekeeper.
FOQA/FDM Review Group

The event will then be reviewed at the regular FOQA/FDM review meeting. This group may also recommend
further mitigations such as a notice to all pilots regarding the company procedure for responding to TCAS
RAs and/or including TCAS RAs in the next round of simulator refresher training.
Key Performance Indicators

In order to allow, if needed, the comparison of TCAS performance and pilot compliance it is recommended
that the aircraft operators collect, if feasible, the following data (derived from their FDM systems):
• Number of RAs by type (as defined in Table 1)
• Number of RAs by type (as defined in Table 1) and altitude1
• Number of RAs by type (as defined in Table 1) and aircraft type
• Number of RAs by type (as defined in Table 1) and their duration
The assessment of pilot compliance (Followed – Not Followed – Opposite – Excessive) should be made
using the above-mentioned criteria. The evolution of pilot compliance assessment should be made on
regular basis to determine if there are any training or safety issues.
Furthermore, a key indicator of the number of RAs per flight hour and per leg should be calculated and its
evolution observed as well, as the increase in certain types of RAs (e.g. Level Off RAs) may be an indicator
of underlying operational problems (approaching the cleared level with too high a vertical rate).
1
The following altitude bands should be used: 1000 – 2350 ft AGL; 2350 ft AGL – FL50; FL50 – FL100; FL100 – FL200; FL200 –
FL290; FL290 – FL410 and FL410 and above.

Conclusion
An operator must establish procedures how their flight crews should operate TCAS and respond to RAs.
Broadly speaking, operator’s procedure should cover topics addressed in this document and, especially,
ICAO provisions. That includes but it is not limited to:
• Pilot responses to RAs;
• Pilot compliance with RAs;
• Aircraft operations during an RA;
• TCAS training;
• RA reporting;
• Use of FOQA/FDM for monitoring and follow up of TCAS RA events.
The risk of a midair collision is still present. When an RA is generated correct action must promptly be
taken. In addition to the recommendations listed in this document, initial/recurrent training as well as
simulator training will enhance flight crew understanding of how TCAS system works, how they should
respond to RAs, as well as the limitations of TCAS. Pilot response is a key component of the TCAS
system.
In an event of TCAS RA, the pilot shall report the circumstances by means of ASR / MOR. This data
together with the operational data downloaded from an aircraft’s on-board computer at the end of every
flight (FDM) can be collected, analyzed and used by the operator, to identify and discover underlying
issues that have the potential to negatively affect aviation safety and to enable operators to take
appropriate action to mitigate.
Recommendations
TCAS has been deployed to act as a last resort safety net to mitigate the risk of midair collisions by
providing flight crews with collision avoidance advice.
So that TCAS can deliver its safety objective, Operators are recommended to ensure that
• Their aircraft are equipped with TCAS as mandated and that the equipment is properly maintained;
• Their pilots who operate TCAS-equipped aircraft have received the relevant training;
• Approved pilot training programs are implemented for initial and recurrent training;
• Procedures are in place for pilots to report a TCAS event and/or problems with TCAS performance;
• Procedures are in place to analyze any reported events and problems and then provide feedback;
• When safe, practical, and in accordance with an operator’s approved operating procedures, pilots
should limit vertical rates to 1500 fpm or less when within 1000 ft of assigned altitudes, unless
otherwise instructed by ATC;
• Their pilots understand the potential risks of an improper response to an RA.

Supporting information
TCAS Advisories Classification
TCAS provides midair collision avoidance by detecting and tracking aircraft proximate to own aircraft.
TCAS II can issue two types of alerts which are announced orally and displayed on the relevant cockpit
instruments:
• Traffic Advisories (TAs), which aim to help the pilots in the visual acquisition of the intruder aircraft,
and to alert them to be ready for a potential resolution advisory.
• Resolution Advisories (RAs), which are avoidance maneuvers recommended to the pilot. An RA will
tell the pilot the range of vertical rates within which the aircraft should be flown to avoid the threat
aircraft. An RA can be generated against all aircraft equipped with an altitude reporting transponder
(Mode S or Mode A/C); the intruder does not need to be fitted with TCAS II. When the intruder aircraft
is also fitted with a TCAS II system, both systems coordinate their RAs through the Mode S data link,
in order to select complementary resolution senses. TCAS II does not detect non-transponder
equipped aircraft or aircraft with a non-operational transponder. A TCAS II RA takes precedence
over any Air Traffic Control instructions or clearances.
During the encounter, the RA strength is evaluated every second. Occasionally, the threat aircraft will
maneuver vertically in a manner that thwarts the effectiveness of the issued RA. In these cases, the initial
RA will be modified to either increase the strength or reverse the sense of the initial RA (when the initially
issued RA is no longer predicted to provide sufficient vertical spacing).
If the TCAS logic determines that the response to an RA has provided the sufficient vertical distance prior
to the closest point of approach (CPA) (i.e. the aircraft have become safely separated in altitude while not
yet safely separated in range), the initial RA will be weakened. This is done to minimize unnecessary
deviations from the original altitude.
Collision avoidance logic sets the minimum time limits on RA duration as follows:
• Minimum RA duration (initial RA to Clear of Conflict) – 5 seconds;

• Minimum time before a reversal RA can be issued – 5 seconds;
• Minimum time before weakening RA can be issued – 10 seconds.
When the intruder ceases to be a threat (i.e. when the range between the TCAS II aircraft and threat
aircraft increases) or when the logic considers that the horizontal distance at CPA will be sufficient, the
resolution advisory is cancelled and a “Clear of Conflict” annunciation is issued. The flight crew should
then return the aircraft to the last ATC assigned level.
For further information about airborne collision avoidance systems, it is recommended to refer to the
EUROCONTROL ACAS Guide – Airborne Collision Avoidance. This Guide has been designed to support
the understanding of the ACAS systems and the training of people involved in the operations of ACAS.
However, it is not, per se, designed for the complete training of controllers or pilots.

For a deeper knowledge, the Reader is advised to refer to documentation listed in the Reference section
of the Guide. The information contained in this Guide is based on the ICAO provisions and other
applicable regulations. This Guide, which is regularly updated, can be found at
https://www.skybrary.aero/bookshelf/books/106.pdf.
TCAS Operational Assessment
TCAS II is estimated to reduce the risk of midair collision by a factor of about 5. The most important single
factor affecting the performance of TCAS II is the response of pilots to RAs. At any time, regardless of the
level of equipage by other aircraft, the risk of midair collision for a specific aircraft can be reduced by a
factor greater than three by fitting TCAS II.
The operational evaluation of TCAS II performance using monitoring data has been demonstrated to be
an effective means for operators to improve overall flight safety. In many cases, RAs have prevented near
mid-air collisions and mid-air collisions from taking place. However, it must be stressed that TCAS II
cannot resolve every near mid-air collision and may induce a near mid-air collision if certain combinations
of events occur.
Although TCAS II significantly improves flight safety, it cannot entirely eliminate all risks of midair collision.
Expected TCAS RA Response
Nominally, RAs are triggered between 15 to 35 seconds from the CPA.
For corrective RAs (i.e. RA that instruct the pilot to deviate from current vertical rate) the response should
be initiated in the proper direction within 5 seconds of the RA being issued. The change in vertical rate
should be accomplished with an acceleration of 0.25 g.
For increase and reversal RAs, the vertical rate change should be started within 2½ seconds of the RA
being issued displayed. The change in vertical rate should be accomplished with an acceleration of 0.35 g.
In accordance to Chapter 5 of the ICAO Doc 9863 Airborne Collision Avoidance System (ACAS) Manual, a
response to RAs, should be done by disconnecting the autopilot and by using prompt, smooth control
inputs; maneuver in the direction and with the vertical rate ACAS requires.
To achieve the required vertical rate (normally 1500 ft per minute) on aircraft where the RA is displayed on
a vertical speed indicator (VSI), it is recommended that the aircraft’s pitch be changed using the
guidelines shown in the table below. Referring to the VSI or VS tape, the pilot must make further pitch
adjustments necessary to place the VS in the green area or outlined pitch guidance area:
Speed Pitch Adjustment

.80 MACH 2 degrees
250 KIAS below 10 000 ft 4 degrees
Approach below 200 KIAS 5 to 7 degrees

On aircraft with pitch guidance for ACAS RA displays, RAs shall be responded to by following pitch
commands.
Prompt and correct reaction to the weakened RA is important as it will minimize altitude deviations and
disruptions to ATC. This will also reduce the possibility of a follow up conflict.
ICAO Annex 6 recommends that the vertical rate is reduced to 1500 ft/min. or less throughout the last
1000 feet of climb or descent to the assigned altitude when the pilot is made aware of another aircraft at
or approaching an adjacent altitude or standard flight level, unless otherwise instructed by ATC.
Challenges Associated With Pilot Responses
In some cases, pilots ignore RAs. The main reasons given are presumed TCAS II malfunction,
misinterpretation of RA alert notification, giving priority to ATC instruction or performing own avoidance
maneuver (based on visual acquisition or own judgement). Inappropriate pilot responses severely impair
TCAS’s performance and increase the risk of a midair collision. It is therefore recommended that
operators ensure that their pilots receive in addition to initial and recurrent classroom training, a simulator
training covering all the RAs and their instructors select various RA scenarios on their simulators at any
time during a simulator session.
A TCAS RA takes priority over an ATC instruction and visual acquisition of traffic as it cannot be
guaranteed that the aircraft acquired visually is the same as the intruder detected by the TCAS system or
it may not be the only aircraft to which ACAS is responding.
Note: Stall warning, wind shear and Ground Proximity Warning System / Terrain Avoidance and
Warning System (GPWS/TAWS) alerts take precedence over ACAS RAs. Wind shear and
GPWS/TAWS warnings will inhibit RAs (TCAS will automatically be placed into TA-only mode).
Furthermore, analysis of STEADES data was conducted and shows that there is a decrease in the number
of TCAS RA reports in the last three years, but it shows a small increase in the number of TCAS RA Climb
reports. It is difficult to say at this stage if the trend will continue, however, IATA will continue to monitor to
check in which direction the trend will go.
Since TCAS RAs are very often triggered by the high vertical rate, STEADES data shows that the Adjust
Vertical Speed, which typically occurs due to high vertical rates, had the largest number of reports for the
last five-year period, in particular in year 2015.
Other data sources indicate that the Level off RA, which in TCAS II version 7.1 replaced the Adjust Vertical
speed RA, has a larger number and that typically also occurs due to the high vertical rates.
Some operators have confirmed that the majority of their TCAS RA alerts occurred due to high rates of
climb/descent. All of them required an action from the crew, although, in hindsight, they may be judged as
operationally unnecessary. Any RAs that appear to be generated in the absence of a credible threat or not
in the compliance with the TCAS Standards should be particularly carefully examined and investigated. If
needed, these cases should be reported to the manufacturers and/or regulators, as any false RAs may be
an indication of the problem that needs to be addressed by them. Resolving such issues will generate
greater confidence in the TCAS system and encourage flight crew compliance with TCAS RAs.
The interaction between TCAS and transponders is critical. Therefore, any TCAS monitoring program
should include provisions for monitoring the performance of transponders; as well as ensuring that
periodic testing and installation of transponders, and appropriate calibration are conducted. The data
should then be processed to look for any exceedance and deviations from the flight manual limits and

Standard Operating Procedures (SOPs). Additionally, in an event of system malfunction, it is

recommended that this event is immediately reported, by means of ASR and maintenance log.
Flight Crew Action
If an RA has been issued, the pilot is expected to act immediately as follows:

• Follow the RA as indicated, unless doing so would jeopardize the safety of the aircraft;
• Disengage the autopilot and follow the RA smoothly and promptly;
• Follow the RA even if there is a conflict between the RA and ATC instruction to maneuver;
• Immediately notify the appropriate ATC unit of any RA which requires a deviation from the current
ATC instruction or clearance;
• Whenever the RA requires a climb while the aircraft is in landing configuration, a go-around should be
executed;
• As soon as the Clear of Conflict messages is received return to the last assigned flight level;
• Notify ATC when returning to the current clearance;
• The PNF (Pilot Not Flying) should provide updates on the traffic location and monitor the response to
the RA. Proper crew resource management should be applied;
• Flight crew must adhere to company SOPs, which may differ from the above.
Reporting
Flight Data Monitoring (FDM)
One way to improve operational safety is through a pro-active non-punitive use of digital flight data from
routine operations to enhance the safety and efficiency of operation, by providing greater insight into the
total flight operations environment. The best potential source of operational data is the operators’ own
Flight Data Monitoring (FDM), Flight Data Analysis (FDA), or Flight Operations Quality Assurance (FOQA)
programs. The aim of this program is to improve safety through an analysis of information downloaded
from an aircraft’s on-board computer at the end of every flight. This information can be used by the
operator, to identify and discover underlying issues that have the potential to negatively affect aviation
safety and to allow operators to take appropriate action.
FDM programs generally involve systems that capture flight data, transform the data into an appropriate
format for analysis, and generate reports and visualization to assist in assessing the data. The following
capabilities are required for an effective FDM program:
• An on-board device to capture a wide range of in-flight parameters and record data on those
parameters using flight data recorders (FDR);
• A means to transfer the data recorded on board the aircraft to a ground-based computer system;
• A means for the ground-based computer system to analyze the data, identify deviations from
expected performance, generate reports to assist in interpreting the reports; and

• An optional software for a flight animation capability to integrate all data, presenting them as a
simulation of in-flight conditions, thereby facilitating visualization of actual events;
• On some aircraft, dedicated TCAS recorders provide accurate information which can be added or
matched to the FDR data should more detailed information be needed for a particular investigation.
FDM is an essential part of a well-functioning Safety Management System (SMS) for an aircraft operator,
and it acts as one of the main data sources for monitoring the operational safety level.
IATA encourages operators to produce set of standardized FDM safety measures and precursors related
to potential midair collisions; such as TCAS RA alert/warning (genuine, nuisance or false).
Analysis and Trends of FDM data
It should be considered how the FDM data is used and analyzed by the aircraft operator in order to
mitigate hazards and apply adequate measures for safety improvement.
It is necessary for the FDR data to be transferred to the analysis platform. Various methods are available
for downloading flight data to the analysis platform or offset for third party analysis. The data analysis
system should have the following capabilities:
• The ability to display detailed information such as where, when, and what
• The ability to display information in a logical and user friendly way.
• The ability to program a range of alert detection thresholds to generate events when parameters
exceed present values.
• The ability to enable detailed analysis of flight data. The ability to provide long term trend analysis of
events.
Most of FDM systems have the ability to record TCAS RAs. This means that data can depict an indication
of whether a TCAS RA was issued, its duration and the type of the RA (e.g. Climb, Descend, Level Off, etc.).
Alert detection thresholds are set to generate events when the value of the parameter exceeds a
predetermined level or threshold. Exceedance detection is used to identify and assess operational risks
as well as it draws the attention of the data analyst. The level at which the alert detection thresholds are
set to generate events for trending or aggregating over a period of time and to enable pilots to be alerted
to their own events. They are tailored to SOPs, aircraft type as well as specific operating scenarios. For
example, an action taken by the pilot in response to the TCAS RA. Typically, most of the operator requires
the pilots to disengage the Auto-Pilot (AP) and follow the instruction of the TCAS RA while informing the
ATC. The operator can easily cross check if the action taken by the flight crew is in compliance with their
manuals or deviates from aircraft flight manual limits and standard operating procedures.
From an analyst FDM perspective is simplified as follows:

From this example, one can notice this a/c is already experiencing a TCAS TA however between the
11:22:13 and 11:22:14 the indication switches from TA to Descend RA. Looking at the AP ENGD column,
you will notice that the crew disengage the AP within the same second and both pitch and V/S increases
indicating that the crew is following the Descend advisory.
The following elements should be taken into account while assessing pilot’s compliance with a TCAS RA:
1. Type of advisory
2. Achieved vertical rate
3. Reaction time
4. RA duration
Table 1 underneath provides a list of all TCAS II alerts (versions 7.0 and 7.1) and vertical rates that need to
be achieved while following these RAs.
Table 1: TCAS II version 7.0 and 7.1 RAs and aural annunciations.
Upward sense RAs Downward sense RAs
RA Required Aural RA Required Aural

vertical rate vertical rate
(ft/min.) (ft/min.)
Climb 1500 Climb, climb Descend – 1500 Descend,
descend
Crossing Climb 1500 Climb, crossing Crossing Descent – 1500 Descend,
climb; climb, crossing
crossing climb descend;
descend,
crossing
descend
Maintain Climb 1500 to 4400 Maintain vertical Maintain Descent – 1500 to Maintain
speed, maintain vertical speed,
– 4400
maintain
Maintain Crossing 1500 to 4400 Maintain vertical Maintain Crossing – 1500 to Maintain
Climb speed, crossing Descent vertical speed,
– 4400
maintain crossing
maintain

Upward sense RAs Downward sense RAs
RA Required Aural RA Required Aural

vertical rate vertical rate
(ft/min.) (ft/min.)
Reduce Descent 0 Adjust vertical Reduce Climb 0 Adjust vertical
speed, adjust speed, adjust
(ver. 7.0 only) – 500 (ver. 7.0 only) 500
– 1000 1000
– 2000 2000
Level Off 0 Level off, level off Level Off 0 Level off, level
off
(ver. 7.1 only) (ver. 7.1 only)
Reversal Climb 1500 Climb, climb NOW; Reversal Descent – 1500 Descend,
climb, climb NOW descend NOW;
(only subsequent (only subsequent
descend,
RA) RA)
descend NOW
Increase Climb 2500 Increase climb, Increase Descent – 2500 Increase
increase climb descent,
(only subsequent (only subsequent
increase
RA) RA)
descent
Preventive RA Prohibited Monitor vertical Preventive RA Prohibited Monitor vertical
rates above 0 speed rates below 0 speed
It should be noted that table 1 lists the vertical rates that nominally should be achieved while responding
to the corresponding RA. In reality, these rates can rarely be precisely achieved. Also, recorders show
sometimes imprecision and fluctuations of vertical rates between updates. Therefore, while assessing
pilot compliance a margin needs to be applied to rates listed above.
The margin typically was applied at ±300 feet above/below the green arc range or at 20% for RAs with
variable green arc range.
Depending of the achieved vs required vertical rates; pilot responses to RAs should be classified as
follows:
• Followed: when the required vertical rate was achieved (within a margin)
• Not Followed: when there was no change to aircraft’s vertical rate after the RA or the change was
not sufficient to meet the vertical rate required by the RA (expect for the RAs when the change in
vertical rate is not required);
• Opposite: when the achieved vertical rate was in the opposite vertical sense to the required rate;
• Excessive: when the achieved vertical rate exceeded the required rate RA (expect for the RAs
when the change in vertical rate is not required).
From the safety point of view the “Not Followed” and “Opposite” events are the most critical and call for
particular attention and investigation.
The “Excessive” events generate a risk of followed up conflicts (with a third-party aircraft) when the
vertical deviation gets too large or may generate a risk of a CFIT accident if an excessive reaction to a

downward sense RA occurs close to the ground. Moreover, the “Excessive” reactions create a potential
for higher g-forces and potential injury to the aircraft occupants.
Resolution Advisories of short duration may not give the pilot a chance to conduct the maneuver).
Therefore, it is recommended that only RAs of the total duration of 8 seconds or longer are assessing to
establish pilot compliance.
Table 2 below presents the assessment criteria for initial RAs for TCAS II versions 7.0 and 7.1.
Table 2: Assessment criteria of pilot responses to initial RAs
Resolution Advisory Followed Not followed Opposite Excessive
Climb 1201 - 2200 0 – 1200 <0 > 2200

(incl. Crossing advisories)
Descend -1201 – -2200 -1200 – 0 >0 < -2200

Upwards Maintain Vertical Speed > RARR -20% 0 – RARR -20% <0 > RARR +40%
Downwards Maintain Vertical Speed < RARR +20% RARR +20%– 0 >0 < RARR -40%
Upwards Level Off -300 – 500 --800 – -301 < -800 > 500
(version 7.1 only)
Downwards Level Off 200 – -500 201 – 500 >500 < -500
(version 7.1 only)
Reduce Climb 2000 ft/min. 1300 – 2200 2201 – 2500 > 2500 < 1300
(version 7.0 only)
(version 7.0 only)
(version 7.0 only)
Reduce Climb 0 ft/min. -500 – 200 201 – 500 > 500 < -500
(version 7.0 only)
Reduce Descent 2000 ft/min. -2200 – -1300 -2500 – -2201 < -2500 > -500
(version 7.0 only)
(version 7.0 only)
(version 7.0 only)

Resolution Advisory Followed Not followed Opposite Excessive
Reduce Descent 0 ft/min. -300 – 500 -800 – -301 < -800 > 500
(version 7.0 only)
Upwards Preventive RA >= 0 n/a <0 n/a
Downwards Preventive RA <= 0 n/a >0 n/a
For Maintain Vertical Rate RAs, the RARR (Resolution Advisory Required Rate) value is used. As for these
RAs, the required vertical rate RA will vary, depending on aircraft’s vertical rate at the time when the RA
was issued. Therefore, the compliance can only be assessed if the required rate has been recorded.
During an encounter, the RA may change. Table 3 below presents the assessment criteria for subsequent
RAs for TCAS II versions 7.0 and 7.1.
Table 3: Assessment criteria of pilot responses to subsequent RAs
Followed Not followed Opposite Excessive
Climb 1201 - 2200 0 – 1200

(incl. Reversal and Crossing <0 > 2200
advisories)
Descend -1201 – -2200 -1200 – 0
(incl. Reversal and Crossing >0 < -2200

advisories)
Increase Climb > 2200 1000 – 2200 < 1000 > 3200
Increase Descent < -2200 -2200 – 0 > -1000 < -3200
Upwards Maintain Vertical Speed

> RARR -20% 1 – RARR -20% <0 > RARR +40%
Downwards Maintain Vertical Speed

< RARR +20% RARR +20% – -1 >0 < RARR -40%
Upwards Level Off (weakening)

-500 – 200 -1000 – -501 > 1000 < -200
(version 7.1 only)
Downwards Level Off (weakening)

-500 – 200 -501 – -1000 < -1000 > 200
(version 7.1 only)
Reduce Climb 0 ft/min. (weakening)

-200 – 500 501 – 1000 > 1000 < -200
(version 7.0 only)
Reduce Descent 0 ft/min.

-500 – 200 -501 – -1000 < -1000 > 200
(weakening)

Followed Not followed Opposite Excessive
(version 7.0 only)
Upwards Preventive RA (weakening) >= 0 n/a <0 n/a
Downwards Preventive RA
<= 0 n/a >0 n/a
(weakening)
Note: in table 2 and 3, for preventative RAs no values are specified for Not Followed and
Excessive responses as these RAs do not require a positive response (change in the vertical
rate).
Following the compliance assessment, the operator should review and debrief the flight crew, using, when
applicable, effective visualization software, including instrument panel graphics, displays of relevant
aircraft systems, and graphical depiction of the aircraft and location and gather their feedback on the
situation.
Furthermore, a review of operational risks where pilots maneuvered too weak, excessive or opposite to
the issued RA should also be analyzed to understand the underlying causes. The opposite reaction to
TCAS RA (e.g. Climb instead of Descend) is of particular concern as it thwarts the effectiveness of TCAS.
However, evaluating the magnitude of the reaction is slightly more complicated as every operator and
every fleet has different sets of FDM events thresholds. To support the operators with the compliance
assessment, the range of vertical rates for each RA are provided in Tables 2 and 3.
An effective FDM tool should be able to provide trend analysis on TCAS RA including flight phases and
geographical location.
It should be noted that several Airbus aircraft are equipped with the TCAS AP/FD (Auto Pilot / Flight
Director) capability. It is a guidance mode which allows the aircraft to automatically fly the RA if the auto
pilot is engaged or the pilot to hand fly the RA by following the flight director commands. While making the
compliance assessment, it should be determined whether the capability was engaged during the event
and in case it has been disconnected, the FDAP team should seek to establish why the crew did not use
the AP/FD capability.
Furthermore, FDAP teams should review the encounter focusing on TCAS RA response or lack of
response, as well as SOP compliance (i.e. disconnecting the AP and FD).
Opposite Initial Pilot Response (OIPR)
During the monitoring of TCAS II version 7.0 performance, it was discovered that in several cases pilots
reacted to RAs in the opposite vertical sense than required, e.g. initiating a climb while a descent was
needed. In most of these cases, the pilots corrected their actions within seconds and subsequently flew
the RA in the correct vertical sense. The Initial Opposite Reactions (OIPR) were occurring across a wide
range of aircraft types and operators. The OIPR events may diminish the effectiveness of collision
avoidance advice given by TCAS or trigger excessive reactions to correct the RA sense.

Since it will not always be possible to detect the OIPR using the criteria outlined above in Tables 1 and 2, it
is recommended that the operators monitor the FDM data for OIPR events to address them accordingly.
The OIPRs may be a symptom of training or operational issue.
Cooperating with ATC Authorities (Air Navigation Service Providers -

ANSPs)
TCAS RA encounters are typically investigated by Air Navigation Service Provider (ANSPs) and in
particularly serious cases by State investigation bodies. In these cases, the airline may gain access to
radar data and other ATC information which would allow making a broader assessment of the event.
Typically, ANSPs have at their disposal radar data and Mode S downlink messages. Radar data will allow
simulating (recreating) RAs (using specialized software tools); however, this technique may not always be
accurate (due to low accuracy of radar data and infrequent updates). Moreover, ANSPs use Mode S RA
downlink messages. Using the Mode S data link, TCAS II can downlink RA Reports to Mode S ground
sensors. This information is provided in the Mode S transponder’s 1090 MHz response to an interrogation
from a Mode S ground sensor requesting information. RA downlink message will provide information
about some but not all details of the issued RA (e.g. the vertical rate limits issued in Adjust Vertical Speed
RAs are not part of the downlinked messages). In addition, the Mode S downlink is subject to latency due
to the rotating radar antenna and the exact time of the RA may not be known as the messages are not
time stamped.
Some ANSPs also intercept RA broadcast and RA coordination messages for monitoring and
investigation purposes.
ACAS Xa
ACAS Xa is a new collision avoidance system, foreseeing a “drop-in replacement” of TCAS II. Instead of
using a set of hard-coded rules, ACAS Xa alerting logic is based upon a numeric lookup table optimized
with respect to a probabilistic model of the airspace and a set of safety and operational considerations.
The system will use the same hardware (antennas and displays) as the current TCAS II system and the
same range of RAs – as in TCAS II version 7.1 – will be used. There will be no change in the way RAs are
displayed and announced to the pilot with the exception of Maintain Vertical Speed and Crossing Maintain
Vertical Speed RAs which will in ACAS Xa installation be announced using the same aural as for Descend
or Climb RAs or Reversal Descent or Reversal Climb RAs (if the Maintain Vertical Speed RA is a reversal
RA). Consequently, pilots and controllers would perceive no change with the transition to the new system,
which will be fully compatible with current TCAS II systems (versions 6.04a, 7.0 and 7.1). It is the intention
that ACAS Xa will eventually replace TCAS II.
ACAS Xa standards (RTCA DO-385 and EUROCAE ED-256) have been finalized in September 2018 and
ACAS Xa is expected to become commercially available after 2020.
Any guidance regarding the TCAS performance assessment provided in this document will be equally
applicable to the ACAS Xa system.

References
• ICAO Annex 6 – Operation of Aircraft - Part I - International Commercial Air Transport –
Aeroplanes (Tenth edition – 2016, amendment 41 – 2017).
• ICAO Annex 10 – Aeronautical Telecommunications - Volume IV - Surveillance Radar and Collision
Avoidance Systems (Fifth edition – 2014).
• ICAO Doc. 8168 – PANS-OPS – Procedures for Air Navigation Services – Aircraft Operations -
Volume I - Flight Procedures (Fifth edition – 2006, amendment 7).
• ICAO Doc. 9863 – Airborne Collision Avoidance System (ACAS) Manual (Second edition – 2012,
amendment 1 – 2017).
• EUROCONTROL ACAS Guide – Airborne Collision Avoidance. December 2017.
• RTCA DO-185B – Minimum Operational Performance Standards (MOPS) for Traffic Alert and
Collision Avoidance System (TCAS II) Airborne Equipment.
• RTCA DO-385 – Minimum Operational Performance Standards for Airborne Collision Avoidance
System X (ACAS X) (ACAS Xa and ACAS Xo).
• EUROCAE ED-143 – Minimum Operational Performance Standards (MOPS) for Traffic Alert and
Collision Avoidance System II (TCAS II).
• EUROCAE ED-256 – Minimum Operational Performance Standards for Airborne Collision
Avoidance System X (ACAS X) (ACAS Xa and ACAS Xo).
• EUROCAE ED-224 – Minimum Aviation System Performance Specification (MASPS) for Flight
Guidance System (FGS) coupled to ACAS.
• FAA AC 20-151C (Advisory Circular) – Airworthiness Approval of Traffic Alert and Collision
Avoidance Systems (TCAS II), Versions 7.0 & 7.1 and Associated Mode S Transponders.

Network Manager
nominated by
the European Commission
Appendix A -
Network
Eurocontrol nominated
Operational
Manager
bySafety Study
TCAS RA notthe
followed
European Commission
Operational Safety Study:

TCAS RA not Followed
+ 03
+ 03
+ 03
+ 03
Edition 1.0
Edition date: 01/09/2017
Status: Released Issue
Intended for: General Public EUROCONTROL
Document characteristics
Document Title Document Subtitle (optional) Edition Number Edition Validity Date
NETWORK MANAGER
TCAS RA not Followed 1.0 01/09/2017
Operational Safety Study
Abstract
As required by NM IR, EU Reg 970/2014 dated 12 September 2014, the NM identifies the Network ‘Top 5’ operational safety
risk priorities. Operational safety studies have been launched covering all of the ‘Top 5’ risk priorities. The present study aims
to serve as a reference for the Network actors in case they undertake operational safety analysis and improvement activities
covering the identified risk priority area.
Authors
Contact Person(s) Tel/email Unit

Tzvetomir Blajev tzvetomir.blajev@eurocontrol.int NMD/NOM/SAF
STATUS AND ACCESSIBILITY

Status Accessible via
Working Draft Intranet
Draft Extranet
Proposed Issue Internet (www.eurocontrol.int) n
Released Issue n
TLP STATUS
Intended for Detail
Red Highly sensitive, non-disclosable information
Amber Sensitive information with limited disclosure
Green Normal business information
White n Public information
© 2017 The European Organisation for the Safety of Air Navigation (EUROCONTROL).
This document is published by EUROCONTROL for information purposes. It may be copied in whole or in part, provided that
EUROCONTROL is mentioned as the source and the extent justified by the non-commercial use (not for sale). The information in
this document may not be modified without prior written permission from EUROCONTROL.
I
Network Manager Operational Safety Study - TCAS RA not Followed
Document approval
Authority
Signature Date
(Name and function)
Operational Safety
Coordinator,
NMD/NOM/SAF
Mr. Tzvetomir Blajev
Head of Safety Unit,

NMD/NOM
Mr. Antonio Licu
Director NM
Mr. Joe Sultana
II
Edition history
The following table records the complete history of the successive editions of the present document.
Edition No. Edition Validity Date Author Reason
0.1 17/02/2016 Creation of the working draft

0.2 16/10/2016 SISG review
0.3 29/052017 Pilot Survey Results
0.4 20/08/2017 DNM Review
1.0 01/09/2017 Release Version
III
Table of contents
Document characteristics I
Document approval II
Edition history III
Table of contents V
1. Executive summary 1
2. Introduction 3
2.1 What is the purpose of this document? 3
2.2 What are the network manager ‘top 5’ ATM operational safety priorities 3
2.2.1 Risk of operation without transponder or with a dysfunctional one 3
2.2.2 Sudden high energy runway conflict 3
2.2.3 Controller detection of potential runway conflict 3
2.2.4 “Blind spot” – inefficient conflict detection with the closest aircraft 3
2.2.5 TCAS RA not followed 3
2.3 How were the ‘top 5’ identified? 3
2.4 Why was “TCAS RA not followed” selected? 4
2.5 How big is the problem across the network? 5
3. TCAS II version 7.1 7
3.1 Improvements made in tcas v7.1 From v7.0. 7
3.1.1 New “level off level off” RA 7
3.1.2 Improved reversal logic 9
3.2 Cockpit presentation 10
3.2.1 RA display: classical instrumentation (IVSI) 12
3.2.2 RA display: electronic flight instrument system (EFIS) 12
3.3 Required pilot actions 13
3.3.1 RA and visual acquisition 13
3.3.2 Inappropriate pilot responses 13
3.3.3 RA inhibitions 13
4. The generic process 14
V
5. Operational context 15
5.1 Operational context factors 15
5.1.1 Human machine interface (HMI) 16
5.1.2 Other signals and alerts 16
5.1.3 Significant flight conditions 16
5.1.4 ATC instructions 16
5.1.5 TCAS RA credibility 16
5.1.6 Aircraft performance 17
5.1.7 Flight Deck CRM 17
5.1.8 Flight Deck mind-set 17
5.1.9 Flight Deck control inputs and handling 17
5.1.10
TCAS training 17
6. Generic scenarios 19
6.1 Analytical deconstruction of “TCAS RA not followed” 20
6.1.1 Scenario mechanisms 20
6.1.2 Scenario sources 20
6.1.3 Scenario outcomes 21
7. Barriers 23
7.1 Barriers are opportunities in some situations 23
7.2 Two types of barriers 24
7.3 Barriers preventing the initiation of a “TCAS RA not followed” event (PB) 25
7.4 Barriers mitigating the impact of a “TCAS RA not followed” event 25
8. Analysis of barriers in generic scenarios 27
8.1 Prevention barrier assessment 27
8.1.1 Colour code used in the barrier analysis 27
8.1.2 Barrier analysis tables 28
8.1.3 Effectiveness of the prevention barriers 28
8.2 Mitigation barrier assessment 29
9. Illustrative examples of actual “TCAS RA not followed” events 31
9.1 Event 1: increased rate of descent contrary to ra to follow atc instruction with visual acquisition. 31
9.2 Event 2: reversed tcas RA resolution to follow atc resolution. 33
9.3 Event 3: no response to RA on approach due misinterpreted visual acquisition of VFR traffic no nn 34
9.4 Event 4: late response to ATC resolution, ignored subsequent ras to continue with ATC resolution. 35
VI
9.5 Event 5: aircraft on go around preferred to keep departing traffic in sight rather than follow RA
in close proximity to terrain. 36
9.6 Event 6: no action taken on a “descend” RA against traffic climbing below, in belief that descending
towards climbing traffic increased risk. 37
9.7 Event 7: continued descending on approach against visual but unknown VFR traffic;
which turned and descended towards subject aircraft. 38
10. Pilot survey 41
10.1 The need for a survey 41
10.2 Survey design 41
10.3 Survey methodology 41
10.4 Survey responses 41
10.5 Significant answers that inform why some tcas ras are not followed. 42
10.5.1 Recalling a TCAS RA 42
10.5.2 Comparison of TCAS RAS that were followed with TCAS RAS that were not followed 43
10.5.3 Reported reasons not to follow TCAS RA 53
10.5.4 Demographics 56
10.5.5
TCAS training 56
10.6 Survey highlights 60
11. Industry best practice and future development 61
11.1 Autopilot / flight director (AP/FD) capability 61
11.2
TCAS alerting prevention (TCAP) capability 61
11.3 Future of collision avoidance: ACAS X 62
11.3.1
ACAS X principles 62
11.3.2
ACAS X benefits 62
11.3.3
ACAS X variants 63
12. Conclusions and recommendations 65
Annex 1 - Glossary of terms 67
Annex 2 - Abbreviations 68
Annex 3 - Bibliography 69
VII
VIII
1. Executive summary
The EUROCONTROL Safety Improvement Sub-Group (SISG), reporting to the EUROCONTROL Safety Team, is tasked with
identifying the ‘Top 5’ ATM Operational Safety Priorities. EUROCONTROL Network Manager is subsequently tasked with the
production of Operational Safety Studies for each of the agreed ‘Top 5’ ATM Operational Safety Priorities.
SISG performed a series of reviews during summer 2015 and involved a series of dedicated workshops with 14 ANSPs, rep-
resenting a large part of European air traffic. The two areas of specific interest selected by EUROCONTROL Safety Team are:
n Preventing mid-air collision en-route.

n Preventing runway collision.
With regards to en-route safety, 90 separation minima infringements were analysed. This represents a sample size of 36%
of all Severity A and B events reported via the EUROCONTROL Annual Summary Template (AST). Seven (7) of the 90 events
were classified as Severity A, 2 of which involved a failure to follow a TCAS Resolution Advisory (RA). Both of these events
crossed through all managed barriers and were, therefore resolved only by Providence. As a result of this analysis, “TCAS RA
Not Followed” was selected as being one of the ‘Top 5’ priority areas. The complete list of the current ‘Top 5’ priority areas is
shown below for reference.
n Risk of operation without transponder or with a dysfunctional one.

n Sudden High Energy Runway Conflict.
n Controller Detection of potential runway conflict.
n “Blind spot” – inefficient conflict detection with the closest aircraft.
n TCAS RA Not Followed.
The methodology employed to create the Operational Safety Study was as follows:
n Generate a set of generic scenarios from possible scenario sources, mechanisms and outcomes.
n Consider what barriers exist that if implemented and deployed correctly could prevent a “TCAS RA Not Followed” event.
n Consider what barriers exist that if implemented and deployed correctly could mitigate the impact of a “TCAS RA Not
Followed” event.
n Analysis of each generic scenario against the potential barriers to establish which of these barriers could be most effec-
tive over the whole range of scenarios.
n Review a set of actual events to confirm that the barriers suggested by the generic analysis validate that the same
barriers should be the most effective in the live environment.
n Carry out an online survey of pilots that elicit responses to aid our understanding of why pilots sometimes do not follow
a TCAS RA.
n Consider current industry best practice and known future developments.
n Draw conclusions from the study and make recommendations to stakeholders.
1
This study concluded that:
n The study is principally validated by an online Pilot Survey, supported by IATA, that elicited 3800 responses from 90
countries.
n The generic study identified only 3 barriers available to Flight Deck crew that could potentially prevent TCAS RAs from
not being followed correctly. All of these barriers are only effective for a small number of generic scenarios. However,
since one type of scenario is prevalent in the actual operation i.e. TCAS not followed due to visual acquisition; training
and promotion of expected pilot response could be effective.
n This study identified only one barrier available to Flight Deck crew that could potentially mitigate the impact of pilots
not following TCAS RAs correctly and that is Autopilot/Flight Director capability to fly the RA.
n According to the Pilot Survey, around 36% of pilots reported experiencing at least one TCAS RA situation within the 5
year period covered by the survey
n According to the Pilot Survey, around 15% of the pilots that reported encountering at least one TCAS RA within the 5
year period covered by the survey, reported not following an RA for various reasons.
n The percentage of TCAS RAs that are not followed is likely to be around 11%.
n According to the Pilot Survey, a decision not to follow an RA due to visual acquisition of the apparent intruder accounts
for more than 70% of all RAs not followed without a valid reason.
n According to the Pilot Survey, neither having a Traffic Advisory (TA) prior to the RA, nor the type of RA was reported to
make any significant difference to the likelihood of an RA being followed.
Recommendations are made that:
n IATA, Pilot Associations, Aircraft Operators and Regulators review the findings of this study and consider undertaking
operational safety analysis and improvement activities for “TCAS RA Not Followed”.
n IATA, Pilot Associations, Aircraft Operators and Regulators consider actions to support an increased active use of FDM in
the monitoring of TCAS RA compliance and the provision of feedback to training organisations and flight crew involved.
n European ANSPs and the EUROCONTROL Safety Improvement Sub-Group (SISG) monitor occurrences involving “TCAS
RA Not Followed” to determine changes in frequency and severity.
n All European stakeholders monitor and support the development of tools and procedures that may assist in the preven-
tion and/or mitigation of “TCAS Not Followed” events.
n All European stakeholders promote and emphasise the requirement and importance of following TCAS RA commands
despite an apparent intruder being visually identified and monitored (subject to the overriding safety of the aircraft).
2
2. Introduction
2.1 What is the purpose of this document?
The purpose of this report is twofold:
n To document the operational safety study on one of the ‘Top 5’ Network Manager operational safety priorities – “TCAS
RA Not Followed”.
n To serve as a reference for the Network actors in case they undertake operational safety analysis and improvement
activities for “TCAS RA Not Followed” events.
2.2 What are the Network Manager ‘Top 5’ ATM Operational Safety Priorities
2.2.1 Risk of operation without transponder or with a dysfunctional one
Operations without transponder or with a dysfunctional one constitute a single threat with a potential of
“passing” through all the existing safety barriers up to “see and avoid”.
2.2.2 Sudden High Energy Runway Conflict

The scenario typically involves rapidly developing situation of runway entry in front of a high energy landing or
taking-of aircraft at position where the available reaction time is close or less than the needed reaction time for
detection, communication and collision avoidance manoeuvre.

2.2.3 Controller detection of potential runway conflict
Some Runway Incursion incidents could be prevented if controllers had better means to detect that the runway
was occupied.

2.2.4 “Blind spot” – inefficient conflict detection with the closest aircraft
Loss of separation “Blind Spot” events are typically characterised by the controller not detecting
a conflict with the closest aircraft. They usually occur when a controller is focussed on a “future situation” and has
filtered out the most immediate aircraft.
2.2.5 TCAS RA Not Followed

Losses of Separation in the En-Route environment sometimes involve “TCAS RA not followed by one
or more flight crews”. Coordinated RA generation and response is an essential safety barrier; however, some events
include a failure to follow the RA correctly or not following the RA at all.
2.3 How were the ‘Top 5’ identified?

The Network Manager identifies Network safety issues to enable aviation stakeholders to identify existing hazards and
anticipate new operational risks.
Our ultimate goal is to keep the Network safe and able to increase its capacity and efficiency.
The EUROCONTROL Safety Improvement Sub-Group (SISG), reporting to the EUROCONTROL Safety Team, is tasked to
identify the ‘Top 5’ ATM Operational Safety Priorities. The SISG followed a structured two-step process of operational safety
prioritisation. Firstly SISG identified a list of priority areas.
The first step was to define broad priority areas for further prioritisation.
3
Based on the availability of reliable safety information, two risk areas were selected for detailed review:
n Runway Incursion
n Loss of Separation En-Route.
SISG performed a series of reviews during 2015 which involved a series of dedicated workshops with 14 ANSPs, representing
a large part of European air traffic. It can be concluded that the analysed sample of incidents is sufficiently representative
for the overall population of incidents in Europe.
Comprehensive barrier models – Safety Functions Maps (SAFMAPs) - were developed and populated with representative
data from the participating ANSPs. The incident data is for high severity (classified as ‘A’ and ‘B’) events, which are on one
side thoroughly investigated and on the other side – highly informative because the incident scenarios ‘test’ the majority of
the available safety barriers.
As a result of the SAFMAP analysis, the ‘Top 5’ priority areas were suggested, agreed by SISG and endorsed by the Safety
Team:
n Risk of operation without transponder or with a dysfunctional one.

n Sudden High-Energy Runway Conflict.
n Controller detection of occupied runway.
n “Blind spot” – inefficient conflict detection with the closest aircraft.
n TCAS RA Not Followed.
2.4 Why was “TCAS RA Not Followed” selected?

The sample of events analysed by EUROCONTROL includes 90 significant separation minima infringements for the En-Route
phase of flight, of which 7 classified as severity A and 83 are classified as severity B.
The EUROCONTROL data shows a total of 248 Severity A or B events in the same period. Thus, the sample includes 36% of
all reported events.
49 / 90 loss of separation events were resolved in the ATC domain, leaving 41 events to be resolved in the Pilot domain.
37/41 events were resolved either by the event being outside the range of TCAS intervention or by actual TCAS pilot
response.
4 events remained. 1 event was resolved by late visual collision avoidance. Collision was only avoided in the remaining
3 events by Providence (luck).
Review of the 7 Severity A events:
n 3 incidents involved the ATCO not preventing a potentially conflicting aircraft to be overlooked when clearing or
instructing another one – “blind spot”.
n 2 incidents involved flight crew not following the flight profile commanded by a TCAS RA.
n 1 incident involved an operation without a functional transponder.
n 1 incident involved air-ground communication misunderstanding.
n So, whilst “TCAS Not Followed” events are infrequent, nearly 30% of the most severe events in the sample involve this
phenomenon.
It was therefore decided to select “TCAS RA Not Followed” as a subject for an Operational Safety Study.
4
2.5 How big is the problem across the Network?
A detailed study provided by one major European short-haul operator, suggests that a TCAS RA is received once in every
6500 flight hours or 3300 sectors. Data from IATA puts the figure at around one every 5500 flight hours. This figure may
however be slightly skewed by a prevalence of TCAS RAs on Approach in USA involving a different traffic mix than Europe.
How many of these TCAS RAs were not followed? This depends on the definition used to scope “not followed”.
IATA studies of global TCAS RAs not followed put the percentage at around 4% of all RAs. An IATA analysis of European only
data placed the percentage at only 1%; however, both of these studies were based on searching for key words in databases
and not analysis of individual events.
Data collected in 2011 indicates that, taking all types of “not followed”, the rate of non-conformity is closer to 35% in En-Route
airspace. This was broken down into 21% inappropriate vertical speed, 11% no reaction and 3% opposite sense reaction.
Thus, conflicting data sources suggest that the scale of TCAS RAs not being followed correctly could be anywhere between
1% and 35%.
It was concluded in the planning stage of this study that the best source of information is the pilots themselves. A voluntary
online survey was proposed and was supported by IATA and a number of European aircraft operators.
5
6
3. TCAS II Version 7.1
TCAS II Version 7.0 has been available since 1999. Based on an extensive analysis of version 7.0 performance, two changes
were identified to improve the TCAS logic in Version 7.1.
3.1 Improvements made in TCAS v7.1 from v7.0.

3.1.1 New “Level Off Level Off” RA
Many cases were found in which pilots did not respond correctly to the “Adjust Vertical Speed, Adjust” (AVSA) RAs.
Pilots increased their vertical rate instead of reducing it, consequently causing a deterioration of the situation (see
Figure 1). The Adjust Vertical Speed RA was the RA whose aural annunciation did not clearly communicate what
exact manoeuvre was required. It was also the most common RA, representing up to two-thirds of total RAs, all of
which increased the potential for incorrect pilot response.
Figure 1: Unintentional opposite response to Adjust Vertical Speed RA in version 7.0.
Additionally, there have been numerous cases of level bust when pilots following the Adjust Vertical Speed RA
went through their cleared level, often causing a follow up RA for the other aircraft above or below (see Figure 2).
Figure 2: Level bust resulting from the response to Adjust Vertical Speed RA in version 7.0.
7
To address these issues, in version 7.1 the AVSA RA has been replaced with a new “Level off, level off” RA. The AVSA RA
in version 7.0 required a reduction of the vertical rate to 0, 500, 1000, or 2000 ft/min depending on the geometry of
the situation. The “Level Off, Level Off” RA requires a reduction of vertical rate to 0 ft/min. The level off is to be achieved
promptly, not at the next standard flight level. The aural message “Level off, level off” has the benefit of being intuitive and
the associated manoeuvre corresponds to the standard levelling off manoeuvre.
Figure 3: Comparison of Adjust Vertical Speed (version 7.0) and Level Off (version 7.1) RAs.
Additionally, replacing the multiple climb/descent rates of the AVSA RA, the “Level Off, level Off” RA should minimise the
altitude deviations induced by TCAS (level busts while “flying the green arc” – see Figure 4).
Figure 4: Comparison of Adjust Vertical Speed (version 7.0) and

Level Off (version 7.1) RAs: minimised altitude deviations.
8
3.1.2 Improved reversal logic
The design of TCAS II version 7.0 allowed for reversal RAs (i.e. “Climb, climb NOW” and “Descend, descend NOW”)
to be issued in coordinated encounters (i.e. both aircraft TCAS II equipped) when the current RA is no longer
predicted to provide sufficient vertical spacing.
After version 7.0 was introduced a weakness in the sense reversal logic was discovered in “vertical chase with low
vertical miss distance” geometries: version 7.0 failed to reverse an RA if two aircraft converging in altitude remained
within 100 feet (see Figure 5). This scenario could occur when one aircraft was not following the RA or was not
TCAS II equipped, and followed an ATC instruction or performed an avoidance manoeuvre based on visual
acquisition.
Figure 5: Geometry in which version 7.0 does not reverse an RA.

When version 7.1 detects that an aircraft is not responding correctly to an RA, it will issue a reversal RA to the aircraft
which manoeuvres in accordance with the RA (i.e. “Climb, climb NOW” or “Descend, descend NOW” RA) and
will change the sense of RA issued to the aircraft that is not responding correctly to be compatible with the reversal,
e.g. “Maintain vertical speed, maintain” RA (see Figure 6). The feature will be activated only if at least 4 seconds
remain before CPA, because a reversal RA triggered in the last 4 seconds gives little chance for correct pilot’s
response, and only if at least 10 seconds have elapsed since the initial RA, because a reversal RA triggered too early
does not give the pilot enough time to comply with the initial RA.
Figure 6: Improvement of reversal logic in version 7.1 (both aircraft equipped).
9
In single equipage encounters, version 7.1 recognises the situation and will issue a reversal if the unequipped threat aircraft
moves in the same vertical direction as the TCAS II equipped aircraft.
Figure 7: Improvement of reversal logic in version 7.1 (only one aircraft equipped).
3.2 Cockpit presentation

It may be useful for the non-flying community e.g. air traffic controllers, to see and understand what the pilot sees and hears
during a TCAS RA event.
The traffic display depicts the position of nearby traffic, relative to own aircraft. It indicates the relative horizontal and vertical
position of other aircraft based on the replies from their transponders.
Displayed traffic information also indicates Proximate, TA, and RA status. The primary purpose of the traffic display is to aid
the flight crew in the visual acquisition of transponder equipped aircraft. The secondary purpose of the traffic display is to
provide the flight crew with time to prepare to manoeuvre the aircraft in the event an RA is issued.
A majority of the traffic displays also provide the pilot with the capability to select multiple ranges and to select the altitude
band for displayed traffic. These capabilities allow the pilot to display traffic at longer ranges and with greater altitude
separation while in cruise flight, while retaining the capability to select lower display ranges in terminal areas to reduce the
amount of display clutter.
Examples of traffic displays are shown in Figures 8, 9 and 10 below.
Figure 8: TCAS traffic display example - dedicated display.
10
Figure 9: TCAS traffic display example – Figure 10: TCAS traffic display example –
IVSI combined with TCAS traffic display. Electronic Flight Instrument System (EFIS).
Figure 11: Standardised traffic
On the TCAS traffic display both colour and shape are used to assist the pilot in interpreting the displayed information.
Own-aircraft is depicted as a white or cyan (light blue) aircraft-like symbol. The location of own aircraft symbol on the
display is dependent on the display implementation.
Targets are displayed by different symbols, according to their threat status:
n Hollow cyan (light blue) or white diamond – for other traffic.

n Solid cyan (light blue) or white diamond – for proximate traffic.
n Solid yellow or amber circle – for intruders (i.e. aircraft which trigger a TA).
n Solid red square – for threats (i.e. aircraft which trigger an RA).
11
Traffic within 6 NM and 1200 ft of own aircraft, are called Proximate traffic and are differentiated from other traffic by a solid
white or cyan (light blue) diamond. In the event of an advisory, this symbol indicates that the aircraft is not the intruder
generating the advisory, when the closest traffic may not necessarily be the most threatening. Each symbol is displayed
according to its relative position to own aircraft. The display accuracy depends on the selected scale. When the 10 NM scale
is in use the positional accuracy is approximately ±1 NM in range and approximately ±10 degrees in bearing.
Vertical data is also shown next to the relevant symbol (when the intruder is reporting altitude). The relative altitude is
displayed in hundreds of feet, above the symbol if the intruder is above own aircraft and below the symbol in the opposite
case. In some aircraft, the flight level of the intruder can be displayed instead of its relative altitude. Additionally, an “up” or
“down” trend arrow is shown when the target aircraft is climbing or descending, respectively, at more than or equal to 500
ft/min.
3.2.1 RA display: classical instrumentation (IVSI)

The traffic display is incorporated into the centre of the Instantaneous Vertical Speed Indicator (IVSI) – see Figure 9
and Figure 10. A 2-NM radius circle is shown by dots or lines around the own aircraft symbol.
An RA is shown by the display of a red arc, which indicates the range of vertical rates, which are to be avoided.
When appropriate, a green arc, shown next to the red arc, indicates to the pilots that they should manoeuvre the
aircraft to reach the required vertical rate, shown in the green arc. If there is more than one threat, two red arcs may
flank the range of the required vertical rates.
3.2.2 RA Display: Electronic Flight Instrument System (EFIS)

There are two Primary Flight Display (PFD) concepts:
n Display on the artificial horizon – a resolution advisory is shown by a red or orange isosceles trapezoid
delineating an area showing the flight attitude values which are to be avoided. This provides direct guidance
on the pitch angle to be achieved by the pilots. This form of display does not include any green fly-to area.
n Display on the Vertical Speed Indicator – the RA is shown in the same way as in “classic” cockpits. A red area
marks the range of vertical rates to be avoided; a green area indicates to the pilots the required vertical rate.
Figure 12 shows examples of TCAS II advisories as shown on IVSI and Electronic Flight Instrument System (EFIS)
instrumentation. A full list of TCAS II advisories can be found in the EUROCONTROL ACAS Guide.

Advisory and aural
IVSI EFIS Manoeuvre
annunciation
Climb
“Climb, climb”
Level Off
(upward sense initial RA)
“Level off,
level off”

Figure 12: Examples of TCAS II advisories
12
3.3 Required pilot actions
In the event of an RA, pilots shall respond immediately by following the RA as indicated unless doing so would jeopardise
safety of the aircraft. The aural annunciation depends on the RA issued. The threat aircraft symbol on the traffic displays
changes to a red solid square and the ranges of the vertical rates to be avoid and the required vertical rate are displayed on
appropriate instruments (cf. ICAO PANS OPS, Doc 8168, vol. I, Chapter 3, 3.2).
Pilots are required to respond to the first RA within 5 seconds and any subsequent RAs within 2.5 seconds. Pilots must not
manoeuvre contrary to the RA. If a TCAS RA manoeuvre is contrary to other critical cockpit warnings, pilots should respect
those other critical warnings – responses to stall warning, wind shear, and GPWS/TAWS take precedence over an TCAS RA,
particularly when the aircraft is less than 2,500 feet above ground level (AGL).
3.3.1 RA and visual acquisition

Pilots sometimes do not follow an RA as they believe they have the threat aircraft in sight and judge there will
be sufficient separation. In this respect, ICAO provisions are quite clear that in the event of an RA, the pilot must
respond immediately by following the RA unless doing so would jeopardise the safety of the aircraft. That provision
applies in all airspace classes and all meteorological conditions. In real-time the pilot has little chance to assess
whether the traffic acquired visually is in fact the one against which the RA has been generated or its trajectory.
3.3.2 Inappropriate pilot responses

In some instances pilots ignore RAs or respond in the opposite sense. The main causes are performing their
own avoidance manoeuvre (based on visual acquisition or own judgement); misinterpretation of RA display or RA
aural annunciation; or giving priority to ATC instruction. Pilots must respond to all RAs a timely manner, applying
the vertical rate required by the RA as accurately as possible in the circumstances and must never manoeuvre in an
opposite sense to a RA.
3.3.3 RA inhibitions
If a TCAS RA occurs when the aircraft is at the maximum altitude for its current weight, the pilots should not assume
that they cannot comply with a climb RA because of that. In these cases it is expected that speed will be traded for
height. Some aircraft types have built-in inhibits which will preclude Climb RAs at maximum altitudes.
Whether or not inhibitors apply, it is still possible in some cases for an RA to exceed the capabilities of the aircraft.
If a stall warning is generated, a response to stall warning takes precedence over a TCAS RA.
For all aircraft, pre-defined limitations apply at lower altitudes to prevent RAs in proximity to the ground
(see Table 1). RAs are inhibited based on radar (radio) altimeter reported heights. Hysteresis values of +100 feet (for
climbing aircraft) and –100 feet (for descending aircraft) ensure that the inhibition state does not oscillate rapidly
should the aircraft be flying close to the nominal altitude boundary but periodically passing above and below that
boundary (e.g. when flying over hilly terrain).
Alert type Alert inhibited below
Increase Descent RA 1550 ft (±100 ft) AGL
Descend RA 1100 ft (±100 ft) AGL
All RAs 1000 ft (±100 ft) AGL
All TCAS aural alerts 500 ft (±100 ft) AGL
Table 1: TCAS alert generation inhibitions.
13
4. The Generic Process
BARRIERS
SCENARIOS ANALYSIS CONCLUSIONS
OPERATIONAL
CONTEXT
Figure 13 above provides an overview of the generic steps in the Operational Safety Study.
A generic process was designed to analyse ATM Operational Safety Priorities (the ‘Top 5’) as to provide a common method-
ology for assessment and evaluation. The process starts with three preparatory steps:
n Identification of the operational context pertaining to the operational area considered.

n Definition of the operational scenarios.
n Identification of safety barriers (both preventing and mitigating the effect of the event).
Once all those data are collated an analysis of effectiveness of barriers against the identified operational scenarios will be
performed and correlated with analysis of real life occurrences. Once the analysis is complete the study will provide the
conclusions.
14
5. Operational Context
BARRIERS
OPERATIONAL
CONTEXT
• Human Machine Interface (HMI) • Aircraft performance

• Other signals and alerts • Flight Deck CRM
• Significant flight conditions • Flight Deck mind set
• ATC instructions • Flight Deck control inputs and handling
• TCAS RA credibility • TCAS training
Figure 14: Operational Context in the Analytical Framework
The Operational Context is the total situational scenario present at the time of a TCAS RA that may impact on the decisions
and actions of the flight crew.
5.1 Operational Context Factors

The operational context may include any or a combination of the following factors:
n Human Machine Interface (HMI).

n Other signals and alerts.
n Significant flight conditions.
n ATC Instructions.
n TCAS RA credibility.
n Aircraft Performance.
n Flight Deck CRM.
n Flight Deck mind set.
n Flight Deck control inputs and handling.
n TCAS Training.
15
5.1.1 Human Machine Interface (HMI)

This includes any factor concerning the perception and interpretation of the TCAS HMI and annunciations.
The appearance and functionality of the TCAS HMI may differ depending on aircraft fit.
The look and method of presentation of the TCAS HMI must meet the requirements of TCAS MOPS but may vary in
detail according to the aircraft manufacturer and associated supplier.
5.1.2 Other signals and alerts

The presence of other Flight Deck signals and alerts simultaneous to the TCAS RA may affect Flight Deck response.
Any factor associated with other warnings which upon recognition of either, TCAS will automatically operate in
TA-Only mode and all TCAS aural annunciations will be supressed:
n TAWS (EGPWS)/GPWS.
n Wind-shear warning.
5.1.3 Significant Flight Conditions

The timing and assessed urgency of significant flight conditions may lead flight crew to react to an RA differently.
This includes any factor concerning concurrent significant flight conditions. These are professional judgement
calls balancing the requirement to follow the TCAS RA against significant conditions:
n Stall warning or Buffet.

n Negative wind-shear without warning.
n Proximity to Significant CB activity and currents.
n Severe weather.
n Proximity to terrain.
5.1.4 ATC Instructions

The presence of ATC instructions that contradict the TCAS RA could influence flight crew action.
Prior ATC instruction to turn being actioned when the RA occurs.
This includes any factor concerning conflicting ATC instructions.
Receipt of an ATC instruction either just before or during the response to the TCAS RA may instil uncertainty or
result in the flight crew not following the flight profile demanded by TCAS or by a delayed response:
n ATC instruction to climb/descent in the opposite sense to that required by the TCAS RA.
n ATC instruction to turn whilst flight crew are executing the RA.
n ATC instruction to ignore the RA.
n Prior ATC instructions to turn being actioned when the RA occurs.
5.1.5 TCAS RA Credibility

The frequency and the credibility of TCAS alerts could influence flight crew action.
This includes any factors concerning a flight crew assessment of TCAS RA credibility:
n RA perceived as false (self-tracking, corrupted or incorrect data).

n RA assessed as a Nuisance Alert (real but deemed to be irrelevant).
n Rapid succession of different RAs (as in the case of encounter with military fast jet).
16
5.1.6 Aircraft Performance
Assessments by the flight crew of the available performance of the aircraft when an RA is triggered may constrain
their response.
This includes any factor concerning actual or perceived aircraft performance limitations.
n High climb/descent rate.

n Low climb/descent rate.
n Limited options at high level (coffin corner, fuel temperature, heavy or not clean).
5.1.7 Flight Deck CRM

Seniority issues and who is actually flying the aircraft and from which seat could influence flight crew output.
This includes any factor concerning flight crew interactions, hierarchy and relations.
It is possible that there may be different or delayed responses to the depending upon seniority issues, physical
position of the crew member acting as pilot flying (PF) and the working relationship of the crew.
n Significant authority gradient on the Flight Deck.

n Physical position of crew member acting as pilot flying (PF).
n Temporary absence of one pilot from the Flight Deck (or taking in-seat controlled rest) at the time of the RA.
5.1.8 Flight Deck mind-set

The mind-set of the flight crew immediately prior to and during the TCAS RA may influence their actions.
This includes any factor concerning the mind-set or expectations of the crew immediately beforehand. A crew may
change or limit its response accordingly:
n Expectation of an RA, following a Traffic Advisory or ATC advice.

n Expectation of an RA based on previous experience.
n Perceived pressure not to delay flight e.g. RA during approach that would result in a go-around.
n Limited response or no action upon visual acquisition.
n Limited response during Initial climb or final approach.
n Autopilot (AP) engaged or not at the time
n Requirement/No requirement to disconnect AP to fly the RA
5.1.9 Flight Deck control inputs and handling

This includes any factor concerning control inputs and handling of the aircraft, such as an inappropriate vertical
speed approaching the selected level. A crew may change or limit its response accordingly:
n Inappropriate high vertical speed.

n Control inputs in an inappropriate order.
n Over-reliance on automation e.g. being behind the aeroplane.
5.1.10 TCAS Training

The scope, completeness and recency in TCAS training could influence the response to a TCAS RA.
This includes any factor concerning the training of TCAS.
TCAS should be included in recurrent training sessions and flight crews should be tested to ensure they are fully
conversant with TCAS procedures, capabilities and limitations, and know how to respond correctly to RAs:
n The type of training e.g. simulator, CBT.

n Scope of TCAS training e.g. number of contacts, expectation.
n Completeness of TCAS training i.e. coverage of all types of RA.
n Frequency of training scenario updates.
n Frequency of pilot training in TCAS manoeuvres.
17
18
6. Generic Scenarios
BARRIERS
Scenario
mechanisms:
• Detection SCENARIOS ANALYSIS CONCLUSIONS
• Interpretation
• Decision
• Execution
OPERATIONAL
CONTEXT
Figure 15: Scenario Mechanisms in the Analytical Framework
Generic operational scenarios are needed to deconstruct the complexity of analysis. Scenario definition is specific to help
decide the efficiency of the safety barriers whilst generic enough to keep their number relatively small. The definition of
generic operational scenarios takes the form of a synthesis of two sources (top-down and bottom-up) of information:
n A systematic analytical break-down of the operational scenario into sub-scenarios. This is based on all theoretically pos-
sible combinations of the scenario (1) mechanisms, (2) sources and (3) outcomes.
n A review of the publicly available information from investigation reports of accidents and serious incidents investigated
following the provisions of ICAO Annex 13 and confidentially provided data in respect of less significant incidents.
19
6.1 Analytical deconstruction of “TCAS RA not followed

6.1.1 Scenario Mechanisms
A failure to follow an TCAS RA alert correctly could be triggered at any one of four Scenario Mechanisms:
1. DETECTION – the flight crew do not detect the TCAS RA alert correctly or react in the belief that an TCAS RA
has been triggered when it has not.
.
2. INTERPRETATION – the flight crew have detected the TCAS RA alert but do not interpret the TCAS HMI correctly
and therefore do not fly the flight profile demanded.
3. DECISION – the flight crew have detected the TCAS RA alert and understand the flight profile demand,
however they may choose not to carry out the required actions.
4. EXECUTION – the flight crew have detected the TCAS RA alert, understand the flight profile demanded and
decide to follow the TCAS RA but do not complete the action required correctly.
6.1.2 Scenario Sources

The OPERATIONAL CONTEXT list produces the SCENARIO SOURCES:

1. Detection
a) HMI
b) Other signals and alerts
c) Flight Deck Mind-set
2. Interpretation
a) HMI
b) Flight Deck CRM
c) Crew TCAS training
3. Decision
a) Other signals and alerts
b) ATC Instructions
c) TCAS credibility
d) Aircraft Performance
e) Flight Deck mind-set
f ) Significant Flight Conditions
g) Crew TCAS training
4. Execution
a) HMI
b) Delayed or Inadequate Response
c) Unintentional Response
d) Aircraft Performance effects
e) Flight Deck CRM
f ) Crew TCAS training
g) Flight Deck mind-set
h) No or limited response possible due short duration of RA
i) Flight Deck control inputs and handling
20
6.1.3 Scenario Outcomes
The Scenario Outcome may be any of four possible outcome types.
1. No action taken in response to TCAS RA
2. Flight profile change is less than that required by TCAS RA
3. Flight profile change is in excess of that required by TCAS RA
4. Flight profile change is in the opposite sense than that required by TCAS RA
The Scenario Outcomes are elucidated by their Mechanisms and Sources as tabulated as Table 2 below:
Table 2: TCAS RA Not Followed Scenarios
1. TCAS RA not followed

1a Lack of Detection of TCAS RA due HMI issues
due lack of Detection
1b Lack of Detection of TCAS RA due presence of other signals and alerts
1c Lack of Detection of TCAS RA due Flight Deck Mind-set

2a Misinterpretation of TCAS RA due HMI issues
due Misinterpretation
2b Misinterpretation of TCAS RA due Flight Deck CRM
2c Misinterpretation of TCAS RA due crew TCAS training

3a Decision not to follow RA due to other signals and alerts
due pilot decision
3b Decision not to follow RA due to contradictory ATC Instructions
3c Decision not to follow RA due TCAS credibility issues
3d Decision not to follow RA due aircraft performance issues
3e Decision not to follow RA due to Flight Deck mind-set
3f Decision not to follow RA due to Significant Flight Conditions
3g Decision not to follow RA due to inadequate TCAS Training

4a TCAS RA not executed as demanded due to HMI Issues
due pilot execution
4b TCAS RA not executed as demanded due to late or inadequate response
4c TCAS RA not executed as demanded due to unintentional response
4d TCAS RA not executed as demanded due to aircraft performance issues
4e TCAS RA not executed as demanded due to Flight Deck CRM
4f TCAS RA not executed as demanded due to inadequate TCAS training
4g TCAS RA not executed as demanded due to Flight Deck mind-set
4h TCAS RA not possible to execute due duration of RA less than reaction time
4i TCAS RA not executed as demanded due to Flight Deck control inputs and handling
21
22
7. Barriers
• Preventing the initiation of a “TCAS RA Not Followed” event

• Mitigating the impact of a “TCAS RA Not Followed” event
BARRIERS
OPERATIONAL
CONTEXT
Figure 16: Barriers in the Analytical Framework
7.1 Barriers are opportunities in some situations

The barriers are not recommendations. The Barriers included in this risk review have been identified as possible ways that
detection of a potential “TCAS RA not followed” event could be employed and/or the consequences mitigated.
Their inclusion does not imply that they are relevant to all situations and neither does it imply that promotion of their
adoption by aircraft operators would necessarily be appropriate in all circumstances. It may be possible to identify more
potentially useful barriers than are included here. In order to define the barrier there is a need first to define the generic
barrier groups for reducing the risk of detection of potential “TCAS RA Not Followed” events. The figure below represents a
generalised SAFMAP for Mid-Air Collision En-Route.
This generalised SAFMAP is derived from the Level 0 Mid-Air Collision SAFMAP V2016 and is the most generic barrier model
for preventing Mid-air collisions because of situations of “TCAS RA not followed”.
23
PROVIDENCE
Potential collision unresolved by visual warning
PILOT COLLISION AVOIDANCE - VISUAL
Potential collision unresolved by TCAS
PILOT COLLISION AVOIDANCE - TCAS
Potential collision unresolved by ATC
ATC COLLISION AVOIDANCE
Separation infrigement / inadequate separation
ATC TACTICAL SEPARATION ASSURANCE
Airborne tactical conflict
TACTICAL CONFLICT PREVENTING
Figure 17: A Barrier Model for Mid-air Collision Avoidance
7.2 Two types of barriers

There are two major sets of barriers which can reduce the risk associated with “TCAS RA not followed” events. These barriers
are identified based on a wide literature search and consultation.
n Prevention of “TCAS RA Not Followed”: These barriers, when deployed and employed correctly, are capable of prevent-
ing a “TCAS RA not being followed”.
n Mitigation of the outcome of “TCAS RA Not Followed”: These barriers, when deployed and employed correctly,
are capable of alerting Pilots to a potential conflict during the initial stages of an event involving not following the com-
mands of an activated TCAS RA, in sufficient time to act in order to prevent a collision.
24
7.3 Barriers preventing the initiation of a “TCAS RA Not Followed” event (PB)
Table 3: Preventing Barriers
Barriers that could

prevent a “TCAS RA Not Following a TCAS TA: TCAS RA Prevention Functionality by Auto Pilot reducing rates
PB1
Followed” event from of climb/descent approaching selected level.
initiation
PB2 Enhanced TCAS Training to include multiple tracks, crossing and reversal RAs,
Training and Enforcement of SOPs requiring TCAS RAs to be followed unless the
PB3
safety of the aircraft would be compromised.
7.4 Barriers mitigating the impact of a “TCAS RA Not Followed” event

Table 4: Mitigating Barriers
Barriers mitigating the

safety risk of a “TCAS RA MB1 Autopilot/Flight Director capability to fly the RA
Not Followed” event
25
26
8. Analysis of Barriers
in Generic Scenarios
BARRIERS
OPERATIONAL
CONTEXT
Figure 18: Analysis in the Analytical Framework
8.1 Prevention Barrier Assessment

The first step of the analysis consists of assessing the potential effectiveness of the prevention barriers in the defined opera-
tional scenarios. The high level assessment considers the operational scenarios and is based on expert judgement. The barri-
ers are assessed individually; the analysis does not consider the interactions of more than one barrier acting in combination.
8.1.1 Colour Code Used in the Barrier Analysis
Table 5: Colour Coding for Barrier Efficiency
Red Barrier that is either inefficient or is not intended for the operational scenario
Yellow Barrier that is partially effective for the operational scenario OR only effective under certain conditions
Green Barrier that is effective and efficient for the operational scenario.
27
8.1.2 Barrier Analysis Tables
Table 6: Preventing Barrier Analysis (1)
1a 1b 1c 2a 2b 2c 3a 3b 3c 3d 3e 3f 3g
PB1: Following a TCAS TA:
TCAS RA Prevention Functionality
by Autopilot reducing rates of
climb/descent approaching
selected level.
PB2: Enhanced TCAS Training
to include multiple tracks,
crossing and reversal RAs.
PB3: Training and Enforcement
of SOPs requiring TCAS RAs
to be followed.
Table 6: Preventing Barrier Analysis (2)
4a 4b 4c 4d 4e 4f 4g 4h 4i
PB1: Following a TCAS TA: TCAS RA
Prevention Functionality by
Autopilot reducing rates of climb/de-
scent approaching selected level.
PB2: Enhanced TCAS Training to include

multiple tracks, crossing and reversal
RAs
PB3: Training and Enforcement of SOPs
requiring TCAS RAs to be followed.
8.1.3 Effectiveness of the Prevention Barriers

In order to organise the results, a scoring system was considered. The main purpose is to give a comparison scale
and an indication on how effective a barrier can be across all considered scenarios and not to provide an
absolute ranking (the higher the score doesn’t necessarily equate to a more effective barrier).
The scoring system utilised to rank the applications is as follows: zero points for an ineffective barrier (red), one
point for a partially effective barrier (yellow) and three points for and effective barrier. Effectiveness is rated on a
scale of:
n 76% - 100% Very High

n 51% - 75% High
n 26% - 50% Medium
n 1% - 25% Low
Table 8: The ranking for the Preventing Barriers; this ranking indicates which are the barriers that
are more effective in most operational scenarios.
Barriers Barrier Description Score Effectiveness

Following a TCAS TA: TCAS RA Prevention Functionality by
PB1 Autopilot reducing rates of climb/descent approaching 9 LOW
selected level.
Enhanced TCAS Training to include multiple tracks,

PB2 4 LOW
crossing and reversal RAs.
Training and Enforcement of SOPs requiring TCAS RAs to

PB3 12 LOW
be followed
28
Table 6 shows that little is available to prevent a potential “TCAS RA not followed” event from being initiated.
Training and enforcement of SOPs requiring TCAS RAs to be followed in all circumstances (PB3) could prevent the events
whereby pilots detect and interpret the RA but make a decision not to follow it. This could be backed up by increased
active use of FDM in the monitoring of TCAS events.
8.2 Mitigation Barrier Assessment

Table 9: Mitigating Barrier Analysis (1)
1a 1b 1c 2a 2b 2c 3a 3b 3c 3d 3e 3f 3g
MB1: Autopilot/Flight Director

Capability to fly the RA
Table 10: Mitigating Barrier Analysis (2)
4a 4b 4c 4d 4e 4f 4g 4h 4i
MB1: Autopilot/Flight Director Capability

to fly the RA
Table 11: Mitigating Barrier Effectiveness
Barriers Barrier Description Score Effectiveness
MB1 Autopilot/Flight Director Capability to fly the RA. 49 VERY HIGH (78%)
The analysis of the only barrier identified that may mitigate the outcome of a “TCAS RA not followed” event shows that it
could, if deployed and allowed to operate to design, have a significant impact on the outcome of TCAS RA events.
At present, this capability is only available on new models from one manufacturer, with a retrofit being available for
older fleet aircraft. Currently, it is unlikely that this capability will become available from any other manufacturer due to
patenting issues.
29
30
9. Illustrative examples of actual
“TCAS RA not followed” events
The actual safety events described are either in the public domain or have been supplied with the permission of the relevant
authorities. In order to dis-identify all stakeholders whilst maintaining the safety lessons, the following editorial actions have
been taken:
n No location, Aircraft Operator, ATCC or ANSP is specified.

n The aircraft involved in each event are denoted solely by the aircraft type.
n No airway, reporting point or routing is specified. In cases where a description of operational situation is helpful, such
details may be fictitious whilst maintaining the general relationship that needs to be described.
Because the reports used in this section came from various sources, they varied in the level of detail provided (e.g. note very
report provided details on timing of Traffic Advisories or Clear of Conflict messages) which is reflected in the description of
events below.
Figures depicting the events below are not to scale.
9.1 Event 1: Increased rate of descent contrary to RA to follow

ATC instruction with visual acquisition.
Table 12: Event 1 Factsheet
Aircraft: B757 (A) Aircraft: B757 (B)
Type of Initial Adjust Vertical Type of Initial Descend

TCAS RA: Speed to zero TCAS RA:
(Level Off in V7.1)
Pilot Action: Increased rate of descent Pilot Action: Correct - Descend
Why: Decision (3e): Why:

Visual acquisition.
Obeyed ATC instruction
Type of Subsequent Maintain Vertical Speed Type of Subsequent Climb

TCAS RA (descent) TCAS RA
Pilot Action: Correct – Continued Descent Pilot Action: Correct - Climb
B757 (A) was at FL390 in the vicinity of B757 (B) aircraft at The controller reported that at this point, the radar display
FL370. Both aircraft were heading southwest with B757 (A) label for B757 (B) was showing FL400 (not FL370) and the
a mile to the west of B757 (B). The ATC controller gave B757 one for B757 (A) was showing FL390. The Flight Progress
(A) a direct routing and following a request for immediate Strips were, however, correctly marked.
descent, clearance to FL250. (Note: a request for immediate
descent is a common precursor to “Controller Blind Spot The B757 (B) crew noted from its TCAS display that the
error”). aircraft above was beginning a descent and TCAS activation
followed, first a TA and then a “Descend, descend” RA and
then soon afterwards a reversal “Climb, climb NOW” RA. The
two RAs were followed and clear of conflict reported to ATC.
31
The B757 (A) crew, descending at 3300 ft/min., would have received an RA to “Adjust vertical speed, adjust” to 0 ft/min. at
the same time as the B757 (B) aircraft received their TA.
Figure 19: Event 1 Reconstruction
Note that with TCAS v7.1 today this RA would have been “Level off, level off”.
One second later, coordinated RAs were annunciated annunciated six seconds later. Minimum separation was 0.8
requiring a climb by B757 (A) and a descent by B757 (B), both NM as B757 (A) descended through the level of B757 (B).
at a rate of 1500 ft/min. The B757 (B) aircraft began descent
but the B757 (A) aircraft, still above the other aircraft, Investigation of the radar system found that it fitted the
ignored the RA and continued descent at an increased rate characteristics of garbling seen in non-Mode S radars
of 4200 ft/min. with the other aircraft in sight. when the oblique distances between each aircraft and the
respective radars are very similar. The label for the B757 (B)
The effect of this had been to trigger the reversal “Climb, aircraft temporarily disappearing from the radar screen and
climb NOW” RA for the B757 (B) aircraft simultaneously with being replaced by two labels, one showing the correct level
a “Maintain vertical speed, maintain” RA (see 8.8.8 below) of FL370 and another showing FL405, with the B757 (A)
for the B757 (A) aircraft which had by now passed through aircraft still showing FL390.
the level of the B757 (B) aircraft. “Clear of conflict” was
32
9.2 Event 2: Reversed TCAS RA resolution to follow ATC resolution.
Table 13 Event 2 Factsheet
Aircraft: A 319 Aircraft: PRM1
Type of Initial Crossing Maintain vertical Type of Initial Crossing Climb

TCAS RA: speed TCAS RA:
Pilot Action: Correct - Descend Pilot Action: Correct Climb initially

but reversed action to Descend
Why: Why: Decision (3b)

Interrupted TCAS resolution to
follow ATC resolution
Type of Subsequent Increase descent Type of Subsequent

TCAS RA TCAS RA
Pilot Action: Correct – Continued Descent Pilot Action:

An A319 lost separation against a Raytheon 390 (PRM1). Late ATC attempts to resolve the conflict were pre-empted by
coordinated TCAS RAs but during them, the PRM1 crew elected to reverse their initial TCAS RA climb whilst it was still an-
nunciated, in favour of an ATC instruction to descend. This action very significantly reduced the separation which would
otherwise have prevailed.
The A319 was tracking 220° and in the descent to FL250. ACC ACC B in the descent to FL250 and gave this cleared level,
A controller had then entered FL280 into the system before neither of the two controllers involved noticed that the level
handing the aircraft over to ACC B with whom there was a stated was different to the one on the system. Meanwhile,
standing agreement for silent transfers at FL 280 without the PRM1 was tracking 060° in the climb to FL270 on a
explicit coordination. Unfortunately ACC A had mistakenly trajectory that would cross less than a mile to the south of
cleared the A319 to FL250. Neither he nor his colleagues the A319. As the A319 descended below FL280 towards the
on the sector noticed the discrepancy between the flight climbing PRM1, STCAs were activated successively in both
level clearance given by radio and the one entered into the ACCs.
system. When the A319 crew subsequently checked in with
33
The A319 received a TCAS RA to “Maintain vertical speed, The ATC controller, after giving the PRM1 the instruction
crossing maintain”. This was followed and resulted in de- to descend rather than climb, subsequently realised that
scent at 1500-2000 ft/min. contrary to an ATC instruction a coordinated TCAS RA must be in progress and instructed
which had just been received to stop descent at FL270. the aircraft to “follow TCAS, opposite traffic one mile, follow
TCAS”.
A TCAS RA was annunciated for the PRM1 two seconds lat-
er to “Climb, crossing climb” which was followed and re- Six seconds later, at Closest Point of Approach the A319 was
quired a slight increase in the prior rate of climb to achieve 50 feet below and 0.6 NM horizontally from the PRM1 with
1500-2000 ft/min. both aircraft still descending. As the cross took place the RA
on the A319 changed to a Reversal “Climb, climb NOW” re-
However, 9 seconds after this annunciation and with it still quiring a climb rate of 1500-2000 ft/min. Thereafter, separa-
displayed, ATC instructed the aircraft to descend to FL 260 tion increased as their tracks diverged and “Clear of conflict”
and the crew disregarded the RA - which continued un- was annunciated onboard both aircraft 8 seconds later.
changed for a further 23 seconds - and began a descent at
a higher rate than the A319 which then received an RA to During the time when a TCAS RA was active, the PRM1 crew
“Increase descent, increase descent”. had twice read back ATC instructions which were both con-
trary to the RA but did not inform ATC that a TCAS RA had
been received. Despite flying in VMC, neither aircraft crew
reported having acquired the other aircraft visually at any
point.
9.3 Event 3: No Response to RA on approach due misinterpreted

visual acquisition of VFR traffic No NN
Aircraft: RJ100 Aircraft: Ultralight
Type of Initial Climb Type of Initial Not equipped

TCAS RA: TCAS RA:
Pilot Action: No Action Pilot Action:
Why: Decision (3e) Why:

Misinterpretation of Visual
Information. Believed that VFR
traffic was already vertically
clear
Type of Subsequent Reverse descent Type of Subsequent

TCAS RA TCAS RA
Pilot Action: Correct – Continued Descent Pilot Action:
An ultra-light single engine aircraft on a VFR flight obtained

a clearance from TWR to cross the control zone at 3000 feet.
Less than 2 minutes later, an Avro RJ100 called the tower
reporting established on ILS. At this time, the RJ100 was ap-
proximately 800 feet above the ultra-light.
34
The TWR controller gave traffic information to the RJ100 tra-light. Based on visual acquisition, the RJ100 pilot judged
crew about the VFR aircraft in their 11 o’clock position at a that the ultra-light was already above. Therefore, he decid-
distance of 3 NM, moving left to right. The RJ100 crew re- ed to descend to fly below the ultra-light rather than climb
sponded that they had the traffic in sight. Twenty seconds as advised by the RA. While the RJ100 was descending, the
later, TWR gave traffic information to the ultra-light pilot on RA reversed to “Descend, descend NOW”.
the RJ100 now in his 2 o’clock position, 2 NM, descending
600 feet above. The pilot responded that he could not see At the time of the RA, the ultra-light was in fact 200 feet
the RJ100. below the RJ100 at the distance 1.2 NM. The RA changed to
Descend when both aircraft were at the same altitude at a
When the RJ100 was passing through 3450 feet, it got a distance of 0.64 NM. Shortly thereafter, the two flight paths
TA. After 11 seconds, when the RJ100 was passing through crossed. At the Closest Point of Approach, the spacing be-
3200 feet it received a “Climb, climb” RA against the ul- tween the aircraft was just 0.07 NM and 200 feet.
9.4 Event 4: Late response to ATC resolution, ignored subsequent RAs

to continue with ATC resolution.
Aircraft: B777 Aircraft: Not stated in the report
Type of Initial Crossing Climb Type of Initial n/a

TCAS RA: TCAS RA:
Why: Decision (3g) Why:

Following ATC resolution. Did not
respond to contrary TCAS RA
Type of Subsequent Adjust Vertical Speed Type of Subsequent

TCAS RA (Level Off in v7.1) TCAS RA
A B777 came within 200 feet vertically and 0.61 NM laterally of another aircraft after climbing significantly above the SID
stop altitude of 4000 feet believing clearance was to Flight Plan level (FL310). The crew responded to ATC avoiding action to
descend and then disregarded “Climb, climb” RA and subsequently “Adjust vertical speed, adjust” RA (which would be
“Level off, level off” RA in v7.1) which were subsequently announced.
The level information given on the SID plate included a

requirement to “MAINTAIN 4000’ or as assigned by ATC”.
However, this information had been missed. The PF stated
that he had scrolled “to the pictorial portion of the chart
and zoomed in to check on the track and distances”. He
missed the separate text box. He had concluded that there
were no altitude restrictions and set the initial FPL cruising
altitude of FL310 on the MCP.
35
Once airborne, initial contact with ATC was followed by a When ATC instructed the aircraft to descend to 5000 feet,
radar heading to take up after passing 2500 feet; the read this altitude was not set on the MCP so that when the air-
back of this clearance did not include the altitude that the craft reached 5000 feet, another unauthorised climb began
aircraft was climbing to. The climb was continued through and it took another query from ATC before the crew realised
the 4000 feet stop altitude and passing 5600 feet, a TCAS TA this error.
was annunciated followed by an ATC instruction delivered
“in an urgent tone” to descend to 5000 feet. The PF discon- During post-event interviews, it was apparent that all crew
nected the AP and initiated the descent and soon after this members expected that there would be some intermediate
a TCAS “Climb crossing, climb” RA was annunciated. The altitude to level off at before being cleared to FL310. The PF
traffic was passing 6000 feet descending to its cleared level said that when he did not see any intermediate altitude on
of 5000 feet. The B777 crew did not follow its RA and after 19 the SID chart, he thought that it would be given to him by
seconds, with the aircraft still descending, the RA changed ATC later. The Departure Clearance given by ATC concluded
to an “Adjust vertical speed, adjust” RA. This type of TCAS with the phrase “climb via SID”. The R/T phrase “climb via” is
RA would be “Level off, Level off” in v7.1. deemed to include the stop altitude as well as the route and
neither are repeated when including a SID in the Departure
The AVSA RA continued until “Clear of conflict” followed 8 Clearance.
seconds later and the aircraft started to level off at 5000
feet. This was the first time the PF had experienced a Crossing
Climb RA in actual flight and the first time the other two
pilots had experienced any RA in actual flight.
9.5 Event 5: Aircraft on Go Around preferred to keep departing traffic

in sight rather than follow RA in close proximity to terrain.
Aircraft: CRJ200 Aircraft: A320
Type of Initial Descend Type of Initial Climb

TCAS RA: TCAS RA:
Pilot Action: No Action Pilot Action: Correct Action
Why: Decision (3f ) Why:

Maintaining visual contact
with traffic. Proximity to terrain
deemed to carry more risk.
Type of Subsequent Type of Subsequent
TCAS RA TCAS RA
Pilot Action: Pilot Action:
The crew of a Bombardier CRJ200 on a visual go around took proach landing assured”. When the CRJ was at about 1.5 NM
visual avoiding action overhead the aerodrome to ensure from the threshold, the crew decided that separation from
separation from an Airbus A320 which had just been cleared the rolling A320 was insufficient and initiated a go around.
to line up and take-off by ATC. Both aircraft received TCAS Almost immediately after this, the TWR controller instruct-
RAs. Minimum achieved separation from radar was 0.17 NM ed the crew to make a go around.
metres and 260 feet vertically.
30 seconds after issuing the go around instruction to the
The TWR controller had, after checking with the crew of the CRJ, the TWR controller had instructed the CRJ crew to
departing A320 that they were ready for an immediate take commence an early left turn “when safe and able”. By this
off with “no stopping on the runway”, cleared the A320 to time, the A320 was passing 1300 feet and the CRJ had al-
enter the runway and take-off. The CRJ, at approximately 3 ready reached the Missed Approach altitude of 1800 feet
NM from the runway threshold, was told “continue with ap- and levelled off.
36
The CRJ crew deviated slightly to the right to keep the A320
climbing ahead and below in sight. As the A320 climbed
towards the same level, a coordinated TCAS RA was trig-
gered. The A320 received a “Climb, climb” RA, whilst the CRJ
received a “Descend, descend” RA. The CRJ crew reported
later that, at this point, the A320 was approximately 5° to
10° to the left of them and slightly below them.
The PF of the CRJ elected to continue with a right turn and Figure 23: Event 5 Reconstruction
not to follow the RA, which would have put the aircraft in
close proximity to terrain with the possibility of Enhanced
Ground Proximity Warning System (EGPWS) activation.
Having the traffic visual, he considered that the safer option
was not to follow the RA. As the A320 passed through the
level of the CRJ a few seconds later, the CRJ turned right 0.5
NM behind the A320.
9.6 Event 6: No action taken on a “Descend” RA against traffic climbing below,

in belief that descending towards climbing traffic increased risk.
Aircraft: B777 Aircraft: C550
Type of Initial Crossing Descend Type of Initial Not equipped with TCAS II
TCAS RA: TCAS RA:
Why: Decision (3g) Why:

Believed that a “Descend
instruction against a target
below was dangerous.
Lack of TCAS training
Type of Subsequent Increase Descent Type of Subsequent

TCAS RA TCAS RA
Pilot Action: As above Pilot Action:
Type of 2nd Subsequent Climb Type of 2nd Subsequent

TCAS RA: TCAS RA:
Pilot Action: Correct response began but Pilot Action:

CoC followed immediately
A Cessna Citation C550 on the ground requested start up was not equipped with TCAS II. Once airborne the aircraft
and departure clearance. The SID has an ultimate altitude started to climb to 4000 feet at a rate of 3300 ft/min.
limit of 4000 feet, however there is an intermediate stop at
3000 feet until it digresses from the track of inbound air- As the Citation got airborne off runway 27 turning north,
craft descending to 4000 feet for a neighbouring airport. a B777 was vectored on a southerly heading for an ILS ap-
The TWR controller cleared the Citation to climb on the SID proach to its destination. The B777 was cleared to descend
to 3000 feet after take-off. The pilot read back “4000 feet” to 4000 feet. The B777 and the C550 were also on different
(having not seen the 3000 feet restriction on the SID plate) radio frequencies.
and this incorrect read-back was not detected. The Citation
37
As the C550 was passing through 3000 feet, the B777

crew got a TCAS TA. Although not required, the B777 pilot
reported the TA to ATC, “We have a traffic alert”. During this
radio transmission a “Descend, crossing descend” RA was
generated.
The controller handling the B777 controller saw the conflict

developing and, suspecting that an RA had already have
been issued, he asked the B777 crew if they could climb
back to 5000 feet, rather than instructing them to climb.
Soon after that the RA on board the B777 strengthened to
“Increase descent, increase descent”. The controller due
to handle the C550 did not receive the aircraft on frequency
until it had already passed 3000 feet and STCA was flashing.
Just seconds before they passed each other, the B777 re- Figure 24: Event 6 Reconstruction
ceives a reversal “Climb, climb NOW” RA. The B777 crew be-
gan to respond to this RA but by then the Citation is already
behind. The B777 crew did not respond to either of the pre-
vious RAs and levelled off at 4000 feet. The Captain of the Later, one crew member reported seeing the C550 briefly in
B777 stated that seeing the other aircraft on the TCAS dis- and out of the clouds. The C550 pilot did not see the B777.
play he was concerned that the “crossing descent” demand The aircraft passed each other on a reciprocal headings just
would put his aircraft closer to the C550. 0.5 NM apart horizontally and 100-200 feet vertically.
9.7 Event 7: Continued descending on approach against visual but unknown

VFR traffic; which turned and descended towards subject aircraft.
Aircraft: BAe125 Aircraft: C152
Type of Initial Climb Type of Initial Not equipped

TCAS RA: TCAS RA:
Why: Decision (3f ) Why:

Continued descending on
approach against visual but
unknown VFR traffic
Type of Subsequent Reverse Descent Type of Subsequent
TCAS RA TCAS RA
Pilot Action: No specific response as Pilot Action:
descending anyway.
A British Aerospace 125 business jet approached its

destination (located outside controlled airspace), main-
taining 2000 feet on the extended runway centreline on a
visual approach. A Cessna 152 on a VFR flight on a track 90°
crossing track was also maintaining 2000 feet.
38
descending through 1700 feet, the BAe125 got a “Climb,
climb” RA. The BAe125 crew did not respond to the RA
continuing the descent as they had seen the C152 but did
not appreciate that it was also descending. After 13 sec-
onds, as the aircraft continued to descend head-on, the
RA reverses to “Descend, descend NOW”. Soon after, they
pass each other with a horizontal spacing of 0.2 NM at the
same level.
Figure 25: Event 7 Reconstruction Had the initial RA been followed, the vertical miss distance
between the aircraft would have increased to over 250
When the C152 neared the extended runway centreline, it feet.
made a turn away from the airport, thus putting both air-
craft on a head-on track at the same altitude. The RA was not followed and that led to the reduction of
the vertical miss distance between the aircraft compared
Soon after, the BAe125 started its descent for landing. Al- to what following the RA would have achieved. The inten-
most simultaneously, the Cessna also started to descend. tions of the intruder were not known. The C152 pilot was
When both were passing 1900 feet, the BAe125 got a TA unaware of the BAe125 and did not see it until the last
against the C152. Some 10 seconds later, when both were moment.
39
40
10. Pilot Survey
10.1 The Need for a Survey 10.4 Survey Responses
It was concluded in the planning stage of this study that a A total of 3800 pilots from 95 countries responded to the
pilot survey was needed to: survey call. The invitation was distributed through multiple
channels and because of that the invitation reach is con-
n Characterise the reasons for “TCAS RA Not Followed”. sidered representative. There is no information available to
n Outline any difference in the context of “TCAS RA fol- argue about the significance of the sample of those that
lowed” and “TCAS RA Not Followed” situations. answered the survey from those that were invited and did
n Scope the magnitude of the problem without aiming at not. EUROCONTROL Network Manager confidential bench-
precise frequency of occurrence calculations. marking with some major European aircraft operators con-
firmed that there is a reasonable assurance that the results
A voluntary online survey was proposed and was support- can be considered as valid.
ed by IATA and a number of European aircraft operators
that elicited 3800 responses from 90 countries. From the 3800 pilots that responded to the Survey, 1387
(36.5%) reported to have “encountered a TCAS RA situation
in the past 5 years”. 194 (5.1%) reported experiencing a
10.2 Survey Design TCAS RA that was not followed either at all or not as com-
manded. This equates to 14.7% of all pilots that reported
The design of the survey framework and the promotion of it experiencing any TCAS RA.
were performed in collaboration between EUROCONTROL
Network Manager and IATA and were supported by many The results based on the respondents’ answers to each
organisations and individuals. question are presented as ratios from those that answered
the particular question. The number of respondents to each
The pilots were asked to recall if they had experienced a question is not one and the same as, because of the large
TCAS RA event in the past 5 years. A total of 55 questions scale of the survey, some respondents abandoned at some
were encompassed in a set of 5 Sections: point when answering the survey questions.
Section 1: Recalling a TCAS RA (in the past 5 years). For example, 3800 pilots answered Section 1, Question 1,
Section 2: Detail a TCAS RA possible to follow and but only 1321 from the 1387 that reported to have “en-
followed. countered a TCAS RA situation in the past 5 years” reached
Section 3: Detail a TCAS RA not possible to follow or was the question about the number of TCAS RA situations en-
not precisely followed. countered during the last 5 years. Similarly, the valid rate
Section 4: Demographic information. for those indicating that they have been involved in TCAS
Section 5: TCAS training. RA situation when it was “not possible to follow RA or RA
was not precisely followed” is calculated based on those
that answered the question – 1321 - and not on those that
10.3 Survey Methodology responded in the beginning – 1387 – to have encountered
TCAS RA.
In cases where pilots answered “no” to whether they had
experienced a TCAS RA event in the past 5 years, they were The drop-off can be explained by several factors: unwill-
automatically redirected to Section 4 and 5. If pilots report- ingness to continue because of perceived sensitivity of the
ed they had encountered a TCAS RA in the past 5 years, de- information asked; unwillingness to continue because of
pending on whether they had reported to have “followed the length of the survey; problems with internet connec-
the TCAS RA” or not, they were directed accordingly to tivity; and technical problems with navigating further on
either Section 2 or 3. Once Sections 2 or 3 had been ad- the survey platform (e.g. browser compatibility issues). The
dressed, the pilots were directed to Section 4 and 5, which drop-off rate is considered to be within the expected and
applied to all respondents independent of their TCAS RA acceptable range for this type of survey.
experience. As the survey was anonymous, not all ques-
tions in Section 4 were mandatory.
41
10.5 Significant Answers that inform why some TCAS RAs are not followed.
In the Survey Analysis below, only those questions and answers are detailed that are considered to be of significance in
terms of the number of responses and in the interpretation of the overall study aim to inform why some TCAS RA events are
not followed.
10.5.1 Recalling a TCAS RA

The following questions were asked.
n Within the last 5 years, have you experienced a TCAS RA situation?

n Approximately how many TCAS RA situations have you experienced during the last 5 years?
n Within the last 5 years, have you been involved in a TCAS RA situation when it was not possible
to follow an RA or one was not precisely followed?
Table 19: Recalling a TCAS RA situation
Within the last 5 years, have you experienced a TCAS RA situation?

Answer Options Response Percent Response Count
Yes 36.5% 1387
No 63.5% 2413
Total 3800
Out of 3800 pilot responses, 1387 (36.5%) reported to have encountered a TCAS RA situation in the past 5 years, as detailed
in Table 19. 1321 of the 1387 respondents continued and answered question 2 about the number of TCAS RA situations
encountered during the last 5 years.
Table 20: Recalling frequency of TCAS RAs
Approximately how many TCAS RAs have you experienced during the last 5 years?
1 57.5% 760
2 30.0% 395
3 or more 12.5% 166
Total 1321
It follows that at least 2048 RAs were experienced by the 1321 respondents. (760 x 1, 395x2, 166 x3).
The exact number was not asked.
Table 21: Recalling not following TCAS RA
Within the last 5 years, have you been involved in a TCAS RA situation when it was not possible
to follow an RA or one was not precisely followed?
Yes 14.7% 194
No 85.3% 1127
Total 1321
14.7% of the pilots that reported to have encountered a TCAS RA indicated that they had been involved in TCAS RA situation
when it was “not possible to follow RA or RA was not precisely followed”.
42
When asked how many RAs were not followed, 28% of the previous respondents decided not to answer, leaving only 139
responses.
Table 22: Recalling frequency of TCAS RAs not followed
Approximately how many TCAS RA situations “not possible to follow” or “not precisely followed”
have you experienced in over the last 5 years?
1 85.6% 119
2 10.8% 15
3 or more 3.6% 5
Total 139
From Table 20 we know that at least 2048 RAs were reported. projected number of RAs not followed is 85.6% x 194 x 1 +
10.8% x 194 x 2 + 3.6% x194 x 3 = 166 + 42 + 21 = 229
Table 21 shows us that 194 respondents reported having at
least one RA not followed. Only 139 of these 194 respond- The percentage of RAs that are not followed is therefore
ents answered the question about how many RAs were not likely to be around 11% (229/2048).
followed. From Table 13 and 4 we can extrapolate that at
least 219 RAs were not followed. (194 + 15 + 10) Conversely, it should be remembered that the Survey an-
swers are all based on Voluntary Reporting. The respond-
The exact number of RAs Not Followed was not captured ents are reporting events which in many cases are in con-
in the survey, as only 72% (139) of the cohort (194) an- tradiction to ICAO and company regulations and standard
swered so the data has a degree of uncertainty. There is practices. It may be therefore that the actual number of
no reason to believe that the relative spread of respons- TCAS RAs are higher than those reported to this study.
es is not the same for the remaining 28%. In this way the
10.5.2 Comparison of TCAS RAs that were followed with TCAS RAs that were not followed
Section 2 asked questions about TCAS RAs that were followed. Section 3 asked the same questions about TCAS RAs
that were not followed.
n What was your position at the time of the RA?

n Were you the Pilot Flying at the time of TCAS RA?
n How was the aircraft being flown up to the time of the TCAS RA?
n What was the TCAS type when event occurred?
n Did you have a TCAS TA before the RA?
n Was the TA briefed with the other crew members?
n What was the TCAS RA type?
n Was the TCAS RA communicated to ATC?
n Did the RA change?
n If so, what was the subsequent TCAS RA type?
n Who made the decision to follow/not follow the RA?
n How was the RA flown?
n If not flown, what was the action after the RA has been announced?
Many of the answers did not show any significant difference between RAs that were followed and RAs that were not fol-
lowed and, therefore, do not aid our understanding of reasons why TCAS RAs are not followed. These questions and answers
are detailed in the Appendix.
It is most useful to compare the answers given in Section 2 about TCAS RAs that were followed, directly with the equivalent
answers given in Section 3 about TCAS Ras that were not followed.
43
SECTION 2: TCAS RA possible to follow and followed SECTION 3: TCAS RA not possible to follow or RA was not precisely followed
Exhibit 1 Exhibit 2
What was your position at the time? What was your position at the time?
Answer Options Response Percent Response Count Answer Options Response Percent Response Count
Captain 57.9% 578 Captain 51.6% 82
First Officer 41.2% 411 First Officer 47.2% 75
Relief Pilot 0.9% 9 Relief Pilot 1.2% 2
Total 998 Total 159
44
Exhibits 1 and 2 show a small change in the position of the reporter. First Officers reported more “TCAS RAs Not Followed”.
This could be explained by the willingness of junior officers to report the actions of other Flight Deck members.
Exhibit 3 Exhibit 4
What was the aircraft type when the event occurred? What was the aircraft type when the event occurred?
Airbus 42.0% 417 Airbus 33,.% 52
Boeing 27.5% 273 Boeing 21.% 34
Embraer 11.1% 110 Embraer 15.9% 25
Other 19.4% 193 Other 29.3% 46
Total 993 Total 157
45
Exhibits 3 and 4 show an interesting reduction in the number of both Boeing and Airbus aircraft involved in “TCAS RAs Not Followed”.
It was hypothesised that one or the other may be a factor, however this is shown not to be the case. There is an increase from 30% to 45% for events involving other aircraft types.
Exhibit 5 Exhibit 6
How was the aircraft flown up to the time of the TCAS RA? How was the aircraft flown up to the time of the TCAS RA?
Manually 25.0% 249 Manually 30.2% 48
Automated 70.1% 700 Automated 59.1% 94
Combination of both 4.9% 49 Combination of both 10.7% 17
Total 998 Total 159
46
Exhibits 5 and 6 show a small increase from 30% to 40% of “TCAS RAs not followed”, to involve aircraft that are being flown manually either completely or partially before the RA.
Exhibit 7 Exhibit 8
Were you the Pilot Flying at the time of TCAS RA? Were you the Pilot Flying at the time of TCAS RA?
Yes 64.2% 641 Yes 56.6% 90
No 35.8% 357 No 43.4% 69
Total 998 Total 159
47
Exhibits 7 and 8 together showing an increase in reports by FOs who were not the Pilot Flying at the time of the TCAS RA.
Exhibit 9 Exhibit 10
Did you have a TCAS TA before the RA? Did you have a TCAS TA before the RA?
Yes 80.3% 801 Yes 76.1% 121
No 19.7% 197 No 23.9% 38
Total 998 Total 159
48
It was hypothesised that the presence or absence of a pre-warning Traffic Alert may influence the pilot response to a subsequent Resolution Advisory.
Exhibits 9 and 10 however show no significant difference.
What was the TCAS RA type? What was the TCAS RA type?
Adjust vertical speed 18.1% 178 Adjust vertical speed 19.7% 31
Level off 9.8% 96 Level off 8.3% 13
Climb 33.7% 332 Climb 33.8% 53
Crossing climb 0.9% 9 Crossing climb 2.5% 4
Descend 18.6% 183 Descend 18.5% 29
Crossing descend 0.4% 4 Crossing descend 0.0% 0
Maintain vertical speed 3.7% 36 Maintain vertical speed 1.3% 2
Crossing maintain vertical speed 0.0% 0 Crossing maintain vertical speed 0.0% 0
Monitor vertical speed 6.1% 60 Monitor vertical speed 6.4% 10
Other/ I don’t remember 8.7% 86 Other/ I don’t remember 9.6% 15
49
Total 984 Total 157
Exhibits 11 and 12 show that the Type of RA has no impact on the likelihood of it being followed.
Did the RA change? Did the RA change?

Yes 20.0% 197 Yes 26.8% 42
No 68.1% 670 No 58.6% 92
I don’t remember 11.9% 117 I don’t remember 14.6% 23
Total 984 Total 157
50
Exhibits 13 and 14 show small increase in the percentage of RAs not followed if the RA changes.
What was the subsequent TCAS RA type? What was the subsequent TCAS RA type?
Adjust vertical speed 26.9% 52 Adjust vertical speed 20.6% 7
Level off 15.5% 30 Level off 8.8% 3
Climb 8.8% 17 Climb 14.7% 5
Crossing climb 1.0% 2 Crossing climb 2.9% 1
Increase climb 3.1% 6 Increase climb 0.0% 0
Descend 10.4% 20 Descend 14.7% 5
Crossing descend 0.5% 1 Crossing descend 0.0% 0
Increase descend 4.2% 8 Increase descend 0.0% 0
Maintain vertical speed 4.7% 9 Maintain vertical speed 2.9% 1
Crossing maintain vertical speed 0.5% 1 Crossing maintain vertical speed 2.9% 1
Monitor Vertical Speed 5.7% 11 Monitor Vertical Speed 5.9% 2
51
Reversal Climb 0.0% 0 Reversal Climb 2.9% 1
Reversal descend 0.0% 0 Reversal descend 0.0% 0
Other/ I don`t remember 18.7% 36 Other/ I don`t remember 23.5% 8
Total 193 Total 34
Exhibits 15 and 16 show that the likelihood of a TCAS RA not being followed increases if the subsequent command is for Climb or Descend.
At what flight level or radio altimeter height (for low level events), At what flight level or radio altimeter height (for low level events),
was the TCAS RA? was the TCAS RA?
0-3000 feet AGL 13.5% 124 0-3000 feet AGL 46.4% 64
3000 feet AGL- FL200 49.3% 453 3000 feet AGL- FL200 34.8% 48
Above FL200 37.2% 342 Above FL200 18.8% 26
Total 919 Total 138
52
Exhibit 17 and 18 show a substantial shift in the altitude bands for RAs Not Followed v RAs followed. Only 13% of RAs that are followed occur below 3000ft.
However, 46% of RAs Not Followed are in this lowest altitude band. This aligns with Table 21 and begins to show a clear relationship of RAs not being followed in the low level
phase of flight due to visual acquisition of the intruder aircraft and deciding to continue with visual separation.
10.5.3 Reported reasons not to follow TCAS RA

Table 23
Within the last 5 years, have you experienced a TCAS RA situation?

Response Response
Answer Options
Percent Count
TCAS RA not detected due to Human Machine Interface issues 0.8% 1
TCAS RA not detected due to other signals and alert 0.8% 1
TCAS RA misinterpreted due to HMI issues 0.0% 0
Decision not to follow due to other signal and alerts 1.5% 2
Decision not to follow due to contradictory ATC instructions 3.8% 5
I did not trust the TCAS RA system 4.6% 6
Decision not to follow due to aircraft performance issues (e.g. maximum level) 1.5% 2
Decision not to follow due to visual acquisition and/or avoidance of the
45.0% 59
conflicting traffic
Proximity to the ground 10.7% 14
Proximity to severe weather 0.0% 0
Short duration RA (RA terminated before a response could be taken) 15.2% 20
This type of RA was not covered in my training 1.5% 2
Because I heard on the frequency that the pilot of the intruder is reacting 3.0% 4
Parallel approaches or closed spaced runways 4.6% 6

Already reacted to TA manually flying in correct direction. less or
0.8% 1
no action to RA
Multiple RAs 0.8% 1
Aircraft on autopilot already capturing level. Less or no action to RA 1.5% 2
Late or incorrect response by PF 2.3% 3
Distraction looking for traffic 0.8% 1
Incorrect controls inputs 0.8% 1
Total 131
Table 23 clearly shows that the most common reason given for not following a TCAS RA is visual acquisition. This
is three times more prevalent than the second highest type, too short duration; and four times more prevalent
than the third highest type, proximity to terrain. Between them, these three types of event account for over 70% of
TCAS RAs Not Followed.
ICAO documents state that all RAs should be followed unless the safety of the aircraft is at risk.
Three categories above would come under that derogation:
n Proximity to terrain.
n Decision to prioritise other signals and alerts.
n Aircraft performance issues e.g. high altitude.
53
These 3 types of RAs equate to 13.7% of all RAs Not Followed.

Three other types of RA may be deemed to reasonable not to follow or not possible to follow:
n Short Duration
n Deemed to be False/Incorrect RAs
n Parallel approaches or closely spaced runways
These 3 types of RAs equate to 24.4 % of all RAs Not Followed.
Therefore, some 38% of RAs Not Followed might be for legitimate reasons.
Table 24: Reasons not to follow TCAS RA
What was the reason not to follow the TCAS RA? (Potentially no valid reasons)
Response Response
Answer Options
Percent Count
Decision not to follow RA due to visual acquisition and/or avoidance
72.8% 59
of the conflicting traffic
Decision not to follow RA due to contradictory ATC instructions 6.2% 5
Heard and/or observed the pilot of the intruder resolving encounter 4.9% 4
Late or incorrect response by PF 3.7% 3
Aircraft on autopilot already capturing level. Less or no action to RA 2.5% 2
This type of RA was not covered in my training 2.5% 2
TCAS RA not detected due to other signals and alert 1.2% 1
TCAS RA not detected due to Human Machine Interface issues 1.2% 1

PF already reacted to TA manually flying in correct direction
1.2% 1
less or no action to RA
Multiple conflicting RAs 1.2% 1
Distraction looking for traffic 1.2% 1
Incorrect controls inputs 1.2% 1
Total 81
Table 24 shows that a decision not to follow an RA due to visual acquisition of the apparent intruder accounts
for more than 70% of all RAs not followed without a valid reason.
54
Table 25: Types of TCAS RA per level band
TCAS RA not followed: At what flight level or radio altimeter height (for low events), was the TCAS RA?
Answer Options 0-3000 ft. AGL 3000 ft. AGL- FL200 Above FL 200
Climb, % 24 – (48.0%) 14 - (28.0%) 12 – (24.0%) 50
Descend, % 16 – (64.0%) 6 – (24.0%) 3 – (12.0%) 25
Adjust vertical speed, % 7 - (29.2%) 12 – (50.0%) 5 – (20.8%) 24
Level off, % 4 - (36.4%) 5 – (45.4%) 2 – (18.2%) 11
Monitor vertical speed, % 6 – (66.7%) 1 – (11.1%) 2 – (22.2%) 9
Crossing climb, % - 3 – (75.0%) 1 – (25.0%) 4
Maintain vertical speed, % 1 – (100.0%) - - 1
Total 58 41 25 124
Table 25 gives the additional information that 69% of RAs not followed below 3000ft are either to climb or de-
scend. The response for “Descend below 3000ft” matches the answer given in Table 24 for reason not to follow the
RA as “proximity to ground”.
Table 26: Reasons not to follow TCAS RA per level band
Reason not to follow RA/ flight level or radio altimeter height:

Answer Options 0-3000 ft. AGL 3000 ft. AGL- FL200 Above FL 200
Decision not to follow RA due to visual
acquisition and/or avoidance of the 32 20 6
conflicting traffic
Short duration RA (RA terminated

7 7 6
before a response could be taken)
Proximity to terrain 12 1 0
Table 26 confirms the prevalence of not following TCAS RAs due to visual acquisition at all Flight levels. The level
of inability to follow RAs due to their short duration is constant across all altitude bands.
Table 27: Actions after TCAS RA has been announced
What was the reason not to follow the TCAS RA? (Potentially no valid reasons)
No manoeuvre 46.0% 64
Less than the required RA rate 35.2% 49
More than the required RA vertical rate 13.0% 18
In opposite sense to the RA 5.8% 8
Total 139
Table 27 confirms the actions expected for findings that the majority of TCAS not followed events involve either
visual acquisition, short duration or proximity to terrain. More than 80% of TCAS RA not followed events result in
either no action being taken or less than the required action.
55
10.5.4 Demographics
Section 4 asked questions about demographics:
n Name, organisation, email (all optional)

n Type of operator
n Base Country of operation
n Total Hours flown
n Hours flown on type
n How does the aircraft represent TCAS commands
Only this last question (that was not a pure demographical characteristic) on Flight Deck equipage is significant.

Table 28: Representation of TCAS RA
TCAS RA not followed: How does your aircraft visually represent the TCAS RA commands?
Answer Options Response Percentage Response Count
On both the Attitude Indicators and
45.5% 61
Vertical Speed Indicators
On the Vertical Speed Indicators 43.3% 58
On the Attitude Indicators 11.2% 15
In opposite sense to the RA 5.8% 8
Total 134
10.5.5 TCAS Training

Section 5 asked questions about TCAS training.

n How often are you tested on your response to TCAS RA?
n What is the type of training?
n In simulation, is more than one aircraft shown on the TCAS display that may become the target aircraft?
n In simulation, are you pre-warned that the exercise will include a TCAS RA?
n What type of TCAS RAs were included in your training?
56
How often are you tested on your response to TCAS RA? How often are you tested on your response to TCAS RA?
once a year 89.9% 834 once a year 84.3% 113
once per 2 years 6.8% 63 once per 2 years 9.7% 13
more than once per 5 years 1.0% 9 more than once per 5 years 1.5% 2
never 0.9% 8 never 2.2% 3
Total 928 Total 134
57
Exhibits 19 and 20 show a small difference in training recency. More frequent training accords with less TCAS RA events not followed.
Almost 5% of pilots involved in such events report either no training or not within the last 5 years.
In simulation, is more than one aircraft shown on the TCAS display that may In simulation, is more than one aircraft shown on the TCAS display that may
become the target aircraft? become the target aircraft?
Yes 78.3% 713 Yes 74.6% 97
No 21.7% 198 No 25.4% 33
Total 911 Total 130
58
Exhibits 21 and 22 show a small difference, but given the considerable difference in response numbers there may be no significance.
In simulation, are you pre-warned that the exercise will include a TCAS RA? In simulation, are you pre-warned that the exercise will include a TCAS RA?
Yes 46.2% 421 Yes 58.5% 76
No 53.8% 490 No 41.5% 54
Total 911 Total 130
59
Exhibits 23 and 24 suggest that pilots who, in training, are not pre-warned that a TCAS RA is coming, react better in live situations.
Table 29
TCAS RA not followed: How does your aircraft visually represent the TCAS RA commands?
Answer Options Response Percentage
Adjust vertical speed 2309
Level off 1579
Climb 2333
Increase climb 1843
Crossing climb 930
Descend 2254
Increase descend 1769
Crossing descend 835
Table 29 shows that no one type of TCAS RA is trained more than 80% each time. The less common, but perhaps
more demanding RAs such as Crossing Climb/Descend are only trained around 30% of the time. There were no
responses that reported Reversal of RAs being trained.
10.6 Survey Highlights

There are several important observations that can be made from the survey:
n 3800 pilots responded. 1387 (36.5%) reported n The Type of RA has no impact on the likelihood of it be-
having experienced a TCAS RA situation in the ing followed.
past 5 years. n More than 80% of TCAS RA Not Followed events result
n At least 2048 RAs were experienced by the in either no action being taken or less than the required
1321 respondents. (The exact number was not action.
asked). n The most common reason given for not following a TCAS
n 14.7% of the pilots that reported encounter- RA is visual acquisition. This is three times more preva-
ing a TCAS RA indicated that they had been in- lent than the second highest type, too short duration;
volved in at least one TCAS RA situation when it and four times more prevalent than the third highest
was “not possible to follow the RA or the RA was type, proximity to terrain. Between them, these three
not precisely followed”. types of event account for over 70% of TCAS RAs Not Fol-
n At least 229 RAs were not followed. (The exact lowed.
number was not asked). n A decision not to follow an RA due to visual acquisition
n The percentage of RAs that are not followed is of the apparent intruder accounts for more than 70% of
likely to be around 11%. all RAs not followed without a valid reason.
n There is a reduction in the number of both Boe- n 46% of RAs not followed are below 3000ft. Most of these
ing and Airbus aircraft involved in TCAS RAs Not involve a decision not the follow the RA after visual ac-
Followed. There is a matching increase from quisition of the apparent intruder.
30% to 45% for events involving other aircraft n Pilots who, in training, are not pre-warned that a TCAS
types. RA is coming, are disposed to react better in live situa-
n There is a small increase for RAs not being fol- tions.
lowed, from 30% to 40%, in events where the n No one type of TCAS RA is trained more than 80% each
aircraft are being flown manually prior to the time. The less common, but perhaps more demanding
RA. RAs such as Crossing Climb/Descend are only trained
n The presence or absence of a pre-warning Traf- around 30% of the time.
fic Alert had no influence on the pilot response n 5% of pilots involved in TCAS RA Not Followed events re-
to a subsequent Resolution Advisory. port no TCAS training within the previous 5 years.
60
11. Industry Best Practice and
Future Development
11.1 Autopilot / Flight Director (AP/FD) Capability
An Airbus AP/FD TCAS capability is a guidance mode which RA response), unless the RA prescribes a null vertical speed
allows the aircraft to automatically fly the RA if the autopilot (e.g. a Level Off RA), in which case the aircraft will change
is engaged or the pilot to manually fly the RA by following its vertical speed to 0 ft/min. The manoeuvre starts without
the flight director commands. The AP/FD capability is delay and has been demonstrated to be equivalent to
currently available for all Airbus aircraft (except A340s). standard response delays (3 – 5 seconds) and accelerations
(0.15 – 0.25 g).
Broadly speaking, when an AP/FD aircraft receives an RA,
the aircraft will automatically change its vertical speed to When the RA is terminated, the AP/FD function will guide
the vertical speed prescribed by the current RA plus 200 the aircraft to the selected altitude. All low level TCAS
ft/min. (in order to help the flight crew monitoring of the inhibitions are applicable.
11.2 TCAS alerting prevention (TCAP) capability

A TCAS Alert Prevention (TCAP) functionality has been the selected altitude, once a TA has been generated and the
introduced by Airbus to prevent the generation of RAs in auto-pilot and/or flight director are engaged (see Figure
1000-foot level-off geometries. The functionality which 26).
is currently available for the Airbus wide body aircraft
(and expected to become available on the A320 family of The TCAP functionality is complementary to the flight
aircraft) uses a new altitude capture law for flight guidance guidance computer’s conventional altitude capture
computers, which decreases aircraft’s vertical rate towards function.
Figure 26: Illustration of an encounter without and with TCAP functionality.
61
11.3 Future of Collision Avoidance: ACAS X
The Federal Aviation Administration (FAA) has funded research and development of a new approach to airborne collision
avoidance (known as ACAS X) since 2008. This approach uses ‘dynamic programming’ and other computer science techniques
(which were not available when TCAS II was first developed) to generate alerts using an off-line optimisation of resolution
advisories.
11.3.1 ACAS X Principles 11.3.2 ACAS X Benefits

Instead of using a set of hard-coded rules, ACAS The following benefits are foreseen through the
X alerting logic is based upon a numeric lookup introduction of ACAS X:
table optimised with respect to a probabilistic
model of the airspace and a set of safety and n Reduction of ‘unnecessary’ advisories: TCAS II is
operational considerations. an effective system operating as designed, but
it can issue alerts in situations where aircraft
The ACAS X probabilistic model provides a will remain safely separated.
statistical representation of the aircraft position
in the future. It also takes into account the safety n Adaptability to future operational concepts:
and operational objectives of the system enabling Both SESAR and NextGen plan to implement
the logic to be tailored to particular procedures or new operational concepts which will reduce
airspace configurations. the spacing between aircraft. TCAS II in its cur-
rent form is not compatible with such concepts
This is fed into an optimisation process called and would alert too frequently to be useful.
“dynamic programming” to determine the best
course of action to follow according to the context n Extending collision avoidance to other classes
of the conflict. This employs a rewards versus of aircraft: To ensure advisories can be followed,
costs system to determine which action would TCAS II is restricted to categories of aircraft ca-
generate the greatest benefits (i.e. maintain pable of achieving specified performance cri-
a safe separation while implementing a cost- teria (e.g. aircraft must be able to achieve a rate
effective avoidance manoeuvre). Key metrics for of climb of 2500 ft/min.), which excludes many
operational suitability and pilot acceptability General Aviation (GA) and Unmanned Aircraft
include minimizing the frequency of alerts that Systems (UAS) or Remotely Piloted Aircraft Sys-
result in reversals/intentional intruder altitude tems (RPAS).
crossings or disruptive advisories in non-critical
encounters. n Use of future surveillance environment: Both
SESAR and NextGen make extensive use of new
The look-up table is used in real-time on-board surveillance sources, especially satellite-based
the aircraft to resolve conflicts. ACAS X collects navigation and advanced ADS-B functionality.
surveillance measurements from an array of TCAS II however relies solely on transponders
sources (approximately every second). Various on-board aircraft which will limit its flexibility
models are used (e.g. a probabilistic sensor model to incorporate these advances.
accounting for sensor error characteristics) to
estimate a state distribution, which is a probability n Safety improvement: It is envisaged that ACAS
distribution over the current positions and X will provide an improvement in safety while
velocities of the aircraft. The state distribution reducing the unnecessary alert rate.
determines where to look in the numeric lookup
table to determine the best action to take (which
includes the option ‘do nothing’). If deemed
necessary, resolution advisories are then issued to
the pilots.
62
11.3.3 ACAS X Variants
n ACAS Xa – The general purpose ACAS X that
makes active interrogations to detect intrud-

ers. ACAS Xa is the baseline system, the suc-
cessor to TCAS II. The Standards are expected
to be ready by 2018 and ACAS X may become
available by 2020.
n ACAS Xo – an ACAS X extension designed for

particular operations, like closely spaced paral-
lel approaches, for which ACAS Xa is less suita-
ble because it might generate a large number
of nuisance alerts. The Standards are also ex-
pected to ready by 2018.
n ACAS Xu – an ACAS X extension designed for

Remotely Piloted Aircraft Systems (RPAS), in-
corporating horizontal resolution manoeuvres.
Work on Standards has started in 2016 and is
expected to be finished in 2020.
n ACAS Xp – A version of ACAS X that relies solely

on passive ADS-B to track intruders and does
not make active interrogations. It is intended
for general aviation aircraft (that are not cur-
rently required to fit TCAS II).
63
64
12 . Conclusions and
Recommendations
BARRIERS
OPERATIONAL
CONTEXT
Figure 27: Conclusions in the Analytical Framework
Conclusion 1 This study is principally validated by an online Pilot Survey supported by IATA that elicited
3800 responses from 90 countries.
Conclusion 2 The generic study identified only 3 barriers available to Flight Deck crew that could potentially prevent
TCAS RAs from not being followed correctly. All of these barriers are only effective for a small number of
generic scenarios. However, since one type of scenario is prevalent in the actual operation i.e. TCAS RA
Not Followed due visual acquisition; training and promotion of expected pilot response could be
effective.
Conclusion 3 This study identified only one barrier available to Flight Deck crew that could potentially mitigate the
impact of pilots not following TCAS RAs correctly and that is Autopilot/Flight Director capability to fly the
RA.
Conclusion 4 The main findings of the Pilot Survey suggest that:

a. According to the Pilot Survey, around 36% of pilots reported experiencing at least one TCAS RA situation
within the 5 year period covered by the survey.
b. According to the Pilot Survey, around 15% of the pilots that reported encountering at least one TCAS RA
within the 5 year period covered by the survey, reported not following an RA for various reasons.
c. The percentage of TCAS RAs that are not followed is likely to be around 11%
d. According to the Pilot Survey, a decision not to follow an RA due to visual acquisition of the apparent
intruder accounts for more than 70% of all RAs not followed without a valid reason.
e. According to the Pilot Survey, neither having a Traffic Advisory (TA) prior to the RA, nor the type of RA
was reported to make any significant difference to the likelihood of an RA being followed..
65
Recommendation 1 It is recommended that IATA, Pilot Associations, Aircraft Operators and Regulators review the
findings of this study and consider undertaking operational safety analysis and improvement
activities for “TCAS RA Not Followed”
Recommendation 2 It is recommended that IATA, Pilot Associations, Aircraft Operators and Regulators
consider actions to support an increased active use of FDM in the monitoring of TCAS RA
compliance and the provision of feedback to training organisations and flight crew involved.
Recommendation 3 It is recommended that European ANSPs and the EUROCONTROL Safety Improvement Sub-Group
(SISG) monitor occurrences involving “TCAS RA Not Followed” to determine changes in frequency
and severity.
Recommendation 4 It is recommended that all European stakeholders monitor and support the development of
tools and procedures that may assist in the prevention and/or mitigation of “TCAS RA Not
Followed” events.
Recommendation 5 It is recommended, in particular, that all European stakeholders promote and emphasise the
requirement and importance of following TCAS RA commands despite an apparent intruder
being visually identified and monitored (subject to the overriding safety of the aircraft).
66
Annex 1 - Glossary of Terms
ACAS II – (“ay-cas two”) – Provides vertical resolution advisories (RAs) in addition to traffic advisories (TAs).
ACAS X – (“ay-cas eks”) – A family of new collision avoidance systems currently under development. It takes advantage of
recent advances in ‘dynamic programming’ and other computer science techniques.
Altitude crossing RA – A resolution advisory is altitude crossing if own ACAS aircraft is currently at least 100 feet below or
above the threat aircraft for upward or downward sense advisories, respectively.
Closest Point of Approach (CPA) – The occurrence of minimum (slant) range between own ACAS aircraft and the intruder.
Range at CPA is the smallest range between the two aircraft and time at CPA is the time at which it occurs.
Increased rate RA – A resolution advisory with a strength that recommends increasing the altitude rate to a value exceeding
that recommended by a previous climb or descend RA.
Resolution Advisory (RA) – An indication given to the flight crew recommending a manoeuvre intended to provide
separation from all threats; or a manoeuvre restriction intended to maintain existing separation.
RA sense – The sense of an ACAS II RA is “upward” if it requires climb or limitation of descent rate and “downward” if it
requires descent or limitation of climb rate. It can be both upward and downward simultaneously if it requires limitation of
the vertical rate to a specified range.
Reversed sense RA – A resolution advisory that has had its sense reversed.
Short Term Conflict Alert (STCA) – A ground-based safety net intended to assist the controller in preventing collision
between aircraft by generating, in a timely manner, an alert of a potential or actual infringement of separation minima.
Strengthening RA – A change in RA to another RA that is more restrictive or requires a greater vertical rate but is in the
same sense as the previous RA.
TCAS II – Traffic alert and Collision Avoidance System II – an aircraft equipment that is an implementation of the ACAS II
standards.
Traffic Advisory (TA) – An indication given to the flight crew that a certain intruder is a potential threat.
This indication contains no suggested manoeuvre
Weakening RA – A resolution advisory with a strength that recommends decrease the altitude rate to a value below that
recommended by a previous RA, when the initially issued RA is predicted to provide sufficient vertical spacing.
67
Annex 2 - Abbreviations
ACAS Airborne Collision Avoidance System
AGL Above Ground Level
ANSP Air Navigation Service Provider
AP/FD Autopilot/Flight Director
ATC Air Traffic Control
CBT Computer Based Training
CoC Clear of conflict
CPA Closest Point of Approach
EFIS Electronic Flight Instrument System
FAA Federal Aviation Administration (USA)
FL Flight Level
ft Feet
ft/min. Feet per minute
GPWS Ground Proximity Warning System
HMI Human Machine Interface
IATA International Air Transport Association
ICAO International Civil Aviation Organization
IVSI Instantaneous Vertical Speed Indicator
NM Nautical Mile(s)
RA Resolution Advisory
SAFMAP Safety Functions Maps
sec Second(s)
SESAR Single European Sky ATM Research Programme
STCA Short Term Conflict Alert
TA Traffic Advisory
TAWS Terrain Avoidance and Warning System
TCAP TCAS Alert Prevention
TCAS Traffic alert and Collision Avoidance System
68
Annex 3 - Bibliography
ICAO Publications:
n ICAO Annex 10 – Aeronautical Telecommunications - Volume IV - Surveillance Radar and Collision Avoidance Systems
(Fifth edition – 2014).
n ICAO Doc. 4444 – PANS-ATM – Procedures for Air Navigation Services – Rules of the Air and Air Traffic Services
(Sixteenth edition – 2016).
n ICAO Doc. 8168 – PANS-OPS – Procedures for Air Navigation Services – Aircraft Operations - Volume I - Flight Procedures
(Fifth edition – 2006, amendment 7).
n ICAO Doc. 9863 – Airborne Collision Avoidance System (ACAS) Manual
(Second edition – 2012).
EUROCONTROL Publications:
n EUROCONTROL ACAS Guide
n EUROCONTROL ACAS Bulletins 1-21 (published since 2002)
69
70
© September 2017 - European Organisation for the Safety of Air Navigation (EUROCONTROL)
This document is published by EUROCONTROL for information purposes. It may be copied in

whole or in part, provided that EUROCONTROL is mentioned as the source and it is not used for
commercial purposes (i.e. for financial gain). The information in this document may not be modi-
fied without prior written permission from EUROCONTROL.
www.eurocontrol.int

IATA Guidance Assessment of Pilot Compliance To TCAS

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IATA Guidance Assessment of Pilot Compliance To TCAS

Uploaded by

Copyright:

Available Formats

Guidance Material

Performance assessment of pilot

Edition 1 January 2019 1

2 Edition 1 January 2019

3 Edition 1 January 2019

Abbreviations and Acronyms

4 Edition 1 January 2019

Note: TCAS II does not provide RAs in the lateral direction.

Currently, there are three versions of TCAS II in operation:

5 Edition 1 January 2019

Monitoring Operational Safety Issues (TCAS RA Event)

1. Was an ASR/MOR filed by the flight crew?

6 Edition 1 January 2019

FOQA/FDM Review Group

Key Performance Indicators

7 Edition 1 January 2019

8 Edition 1 January 2019

• Minimum RA duration (initial RA to Clear of Conflict) – 5 seconds;

9 Edition 1 January 2019

TCAS Operational Assessment

Expected TCAS RA Response

Nominally, RAs are triggered between 15 to 35 seconds from the CPA.

Speed Pitch Adjustment

10 Edition 1 January 2019

Challenges Associated With Pilot Responses

11 Edition 1 January 2019

Standard Operating Procedures (SOPs). Additionally, in an event of system malfunction, it is

Flight Crew Action

If an RA has been issued, the pilot is expected to act immediately as follows:

12 Edition 1 January 2019

Analysis and Trends of FDM data

13 Edition 1 January 2019

Upward sense RAs Downward sense RAs

RA Required Aural RA Required Aural

14 Edition 1 January 2019

Upward sense RAs Downward sense RAs

RA Required Aural RA Required Aural

15 Edition 1 January 2019

Table 2: Assessment criteria of pilot responses to initial RAs

Resolution Advisory Followed Not followed Opposite Excessive

Climb 1201 - 2200 0 – 1200 <0 > 2200

Descend -1201 – -2200 -1200 – 0 >0 < -2200

(version 7.0 only)

(version 7.0 only)

16 Edition 1 January 2019

Resolution Advisory Followed Not followed Opposite Excessive

(version 7.0 only)

Upwards Preventive RA >= 0 n/a <0 n/a

Downwards Preventive RA <= 0 n/a >0 n/a

Followed Not followed Opposite Excessive

Climb 1201 - 2200 0 – 1200

Descend -1201 – -2200 -1200 – 0

(incl. Reversal and Crossing >0 < -2200

Increase Descent < -2200 -2200 – 0 > -1000 < -3200

Upwards Maintain Vertical Speed

Downwards Maintain Vertical Speed

Upwards Level Off (weakening)

Downwards Level Off (weakening)

Reduce Climb 0 ft/min. (weakening)

Reduce Descent 0 ft/min.

17 Edition 1 January 2019

Followed Not followed Opposite Excessive

(version 7.0 only)