Professional Documents
Culture Documents
No 7
EUROCONTROL Voluntary ATM Incident Reporting
2008-2010
EUROCONTROL
© Skyguide
contents
Annexes 31
PERSPECTIVE
The EVAIR team is pleased to bring you EVAIR Safety Bulletin only aim of collecting, analysing and identifying areas of safety
No7. The issue covers the period 2008-2010. It is a great pleasure concerns is to improve flight safety. It is this and only this that
to state that ATM occurrence reporting (by airlines and ANSPs) drives the agreement and relationships that EVAIR has with the
and the number of reports received continues to rise. data providers.
Data collection Verification of the robustness and the quality of EVAIR data
The graph below shows this constant increase in reporting for As a part of the verification process of the robustness and quality
period 2008-2010. Indeed, in 2010 the increase versus 2009 was of EVAIR data, EVAIR established a process to make a cross-check
57% in relative figures. This follows a similar large increase of 73% with the best reporters among the airlines and the best feedback
we already had in 2009 against 2008. It is very encouraging to see providers amongst the ANSPs.
this rising trend and we hope it will continue in the future.
During 2010, EVAIR verified its trend data with several ANSPs
many of whom are responsible for areas of high traffic density.
7
6 Dialogue with key stakeholders indicates that within ATM there
5 are no major differences between EVAIR and ANSPs data relating
4 to identified ATM trends. Moreover, there is a very strong match
3 in the identified areas of concern and their trends in terms of the
2
type of the events, and their increase or decrease, and also in
1
terms of causal factors.
0
2008 2009 2010
Figure 1 – Incident data collection 2008-2010 In addition, EVAIR compares trends of some specific events (e.g.
Go-around, Runway incursion, ACAS RA, Level Bust and Callsign
For the period 2008-2010, more than 110 airlines provided their confusion) with IATA STEADES data. In these areas, there is also no
ATM incidents to EVAIR; of these around 80 are reporting regularly. marked difference between European and global trends, in terms
The reporting airlines fly across the whole of Europe and account of an increase or decrease in the number of incidents.
for 72% of the overall European air traffic.
Feedback – Support to quick fixes
The high increase in reporting is not related to the quality of the The feedback mechanism is the driving factor for the increased
performance of the ATC or airlines’. Instead there are a number reporting. Our data shows that during the period 2008-2010,
of reasons for the rise. One of the most important of these is the EVAIR addressed requests for feedback to more than 50 different
feedback mechanism, which within a short period provides a reply
3000 ANSPs and more than 40 different airlines within and outside
to the incident reporter. The short time frame and information
2500 Europe. Regarding the airlines, most requests for the feedback
about the problem which was behind the incident as well as any
2000 were related to Callsign similarity/confusion issues. EVAIR is the
actions taken, motivates reporters to report more since they see
1500 monitoring mechanism for the EUROCONTROL Callsign simi-
very quickly the result of their reporting. EVAIR is the facilitator
1000 larity project and through the feedback process the airlines
of500
the feedback mechanism. The second important reason for are taking the necessary steps to change conflicting callsigns
the 0high reporting increase is the trust that all players have in the within the same airline or sometimes to make the agreement
2006 2007 2008 2009 2010
reporting and feedback mechanism. Specifically, EVAIR takes great between two different airlines to change the callsign of one
care with the sensitive parts of the reports like the names of the of them. In all 88% of feedback requests are successful. The
reporters or airlines and ANSPs as well as locations of the incidents usual feedback response time varies from a few days to a few
so that they will not be disclosed in public and not abused. The weeks, but sometimes this stretches to several months. This
Throughout the report, relative figures and reporting rates are For this Safety Bulletin, IATA STEADES prepared global information
lower in the EVAIR data set than in STEADES. Whilst several reasons based on the ASRs submitted from over 130 participating airlines
for this may exist, it is noteworthy to mention two specifically. throughout the world. The analysis included research of ASRs for
the period January 2008 - December 2010, during which 189,684
Firstly, the impetus for reporting. Airlines that cooperate with reports were submitted.
EVAIR do not provide all reports, they provide mainly reports
for issues where they seek feedback from the specific ANSP and Security and Confidentiality
which is facilitated by EVAIR. Whereas the airlines participating In collecting and processing data, EVAIR follows strict security and
in IATA’s STEADES program provide all their air safety reports to confidentiality arrangements. Safety data provided are properly
STEADES as part of their agreement with STEADES. safeguarded and de-identified and the information is only used
for the promotion and enhancement of aviation safety.
Secondly, the higher reporting rates in STEADES may also be
attributable to the fact that STEADES coverage is global. This
means that regional differences and/or deficiencies in ATC perfor- EVAIR Suggestions/Improvements
mance, ATM/CNS infrastructure, applied standards, level of the EVAIR is constantly looking at ways to improve its services and
performance, etc, when compared with Europe, may result in products. Suggestions and proposals are more than welcome.
more incidents reported per number of flights. Please forward any thoughts, ideas and comments to:
Ms Dragica Stankovic
Regarding taxonomy and level of analysis, in EVAIR the ATM data EVAIR Function Manager
are analysed by ATM experts and coded following ICAO ADREP dragica.stankovic@eurocontrol.int
taxonomy. STEADES data on the other hand is comprised of pilot
reports [Air Safety Reports, (ASR)] collected and coded by the
airline as part of an airline’s Safety Management System. In EVAIR
multiple reports of the same event are correlated and counted
once as a single incident, for example a single ACAS event reported
The statistics presented in EVAIR Safety Bulletins are based on inci- EVAIR Safety Bulletins are issued twice per year. One covers the
dents received from commercial aircraft operators and feedback whole year period such as this one, whilst the other only the
data provided on some of them by ANSPs. Incident reports are summer season (April - September). Data are provided by 108
very much based on the subjectivity of those who were involved airlines. When solicited, all ANSPs have provided feedback to
and in the first place the pilots who filed the report and described airline reports.
the occurrence. However, improved feedback processes through
which EVAIR is provided with a reply either by Air Navigation Notes:
Service Providers (ANSPs) or airlines - the two main stakeholders 1. In this EVAIR Safety Bulletin within the manual part of
participating in the incident - improves significantly the view and reporting, only relative figures are presented - i.e. the
the knowledge about the causals of the analysed events. EVAIR number of reported occurrences per 10,000 flights of
statistics do not contain severity analysis, since the analysed the airlines participating in the reporting. Within the
reports are not officially investigated or the official (airline/state) automated ACAS RA data collection sections, the data
investigation is still waiting to be closed. Nonetheless, the statis- comprises absolute values.
tics do provide a general view and show some main trends of the 2. The graphs which show the drill down through the data-
current operational safety acceptability. base identifying causal factors could count the same
incident more than once. The reason for that is that one
EVAIR activity covers the whole ECAC airspace as well as some of incident could be associated with more than one causal
the airspaces neighbouring the ECAC region like the Eastern part factor.
of the ICAO EUR region, Middle East, Africa etc. In the same way,
the airlines and ANSPs who participate in EVAIR come from across Definitions:
Europe and, indeed, some airlines based outside of Europe but Definitions for each element contained in the graphs can be found
who regularly fly through European airspace. in the Annex 2..
In this chapter EVAIR presents some of the events which are already addressed in the European Action Plans or projects like Air-ground
communication, Level Bust, Runway Incursion, ACAS, and Callsign Similarity/Confusion. Analysis of EVAIR data helps to monitor the situa-
tion and draws attention to the negative trends which require corrective actions. It also assists in prioritising actions which could be taken
on a pan-European level. ‘Missed approach’/ ‘Go-around’ type events were included in this section for the first time in EVAIR Safety Bulletin
No 6. The reason for this inclusion is the high growth of the events reported to EVAIR during 2010, but also an increase at the global level
which is seen from the IATA STEADES information. The inclusion of this area in the statistics aims to raise awareness amongst the experts
who read EVAIR Safety Bulletins to continue the work in identifying the spectrum of contributory factors for this type of the event and find
the best solutions for their mitigation. We want to emphasise that we are fully aware that missed approach/Go-around is a part of the ap-
proach procedures and that if the approach is unstable, pilots are trained to assess and to make the decision either to continue or to make
a Go-around to avoid e.g a runway excursion. However as it was said, the increased number of Go-arounds indicates that more in-depth
analysis are needed to discover the reasons behind the events. EVAIR has started with this analysis and we hope that more information will
be available for EVAIR Safety Bulletin No 8.
The identified and presented contributors are spread across all phases of flights and linked with the five selected events mentioned
previously. The contributors have the highest levels within our data base. Traditionally, ‘Mistakes’ and ‘Air-ground communication’
(Operational and Spoken) are the most common contributors through 2008 – 2010. However, ‘Traffic’ and ‘Airspace problems’, which
encompass traffic load and traffic complexity, airspace design, pilot problems with the airspace familiarity and lack of traffic informa-
tion, showed a very high increase in 2010 versus 2009. For the majority of them especially for ‘Traffic information’, and ‘Air-Ground
communication’ – language issues - better training is one mitigation factor. However, for our next Bulletin, we will continue to monitor
the situation and try to identify all potential problems behind these contributing factors.
Mistakes
Lapses
Traffic information
2010
Coordination issues 2009
ATC Clearance/Instructions 2008
GO-AROUND
First-time statistics about Go-around were presented in EVAIR Safety Bulletin No 6. It was noted that the number of occurrences in com-
parison with the other events was much higher and that the trend was consistently upwards. The EVAIR database showed a 196% jump in
2010 against 2009 and this confirmed our suspicions that there was an inexorable rise in this type of events. Consequently, we decided
to verify our information with IATA STEADES global data and several ANSPs that have areas of high traffic density. The IATA STEADES data
shows a 78% increases in the rate of reported Go-arounds for the period 2008-2010; furthermore ANSPs who were contacted also show a
similar steady growth.
The increasing number of Missed approach/Go-around events indicates that further work is needed to identify the main contributory
factors and define remedial actions. In this regard EUROCONTROL together with a number of other stakeholders (ANSPs, airlines and their
associations, manufacturers etc) has initiated a study which aims to capture the issues and identify potential solutions to mitigate the
situation. EVAIR data will be used as one of the main sources of information. Furthermore, EVAIR will make additional analysis to support
this initiative.
Finally, by extrapolating EVAIR data, it suggests that, each day, there are at least 4 Go-around events due to various reasons. Considering
that in many states/ANSPs the reporting culture is still too low, the likelihood is that in reality this figure is much higher.
Go-around reports
Go-around reports per 10,000 flights
STEADES
12 Linear (Go-around rpts per 10,000 flights) 4500 © International Air Transport
4000 Association (IATA). All rights
Reports per 10.000 flights
In 2010, from the ATM point of view, the Oper. Comm. Issue
largest number of Go-Arounds occurred
Spoken comm.
due to Mistakes either on the ATM or pilot
Mistakes
side. Lack of, or late essential traffic infor-
Lapses
mation is the 2nd highest contributor with
a strong rise in 2010. Another contributor Traffic information
The EVAIR database shows that runway incursions reported by pilots had the lowest rate among five major categories of events presented
in this Safety Bulletin. The IATA STEADES database shows a similar pattern (Figures 3 and 3a). For the period 2008-2010, both EVAIR and
IATA STEADES data repositories show an increase in the number of RWY incursions. Relative figures for Runway Incursion as well as for
other type of the events within the STEADES database are much higher than within EVAIR. The explanation for that has been given within
the chapter, ‘Stakeholders corner – IATA’.
0,07
0,060
0,06
0,05
0,04
0,039
0,03
0,033
0,02
Tendency of Runway incursions per 10,000 operations
0,01 Yearly development per 10,000 operations
0,00
2008 2009 2010
Incidents rate per 10.000 flights
Figure 7 - RWY incursion 2008-2010
Number of reports
recast, reformatted or trans- 0,5
mitted in any form by any
0,4 150
means, elctronic or mechan-
ical, including photocopying,
recording or any informa- 0,3 100
tion storage and retrieval
system, without written 0,2
consent of IATA, Senior Vice 50
President, Safety, Operations 0,1
& Infrastructure.
0,0 0
2008 2009 2010
Figure 7a - Runway incursion Reports 2008-2010
Traffic information
which is a very similar pattern to that seen in the other
monitored events. Once again this points towards
0,000 0,005 0,010 0,015 0,020
training as one of the possible solutions for improving
Incidents rate per 10.000 flights the situation.
Figure 8 - Runway incursions main contributors 2008 – 2010
According to EVAIR data, the Level Bust rate is lower than Go-Around
0,20 and ACAS RA but higher than Runway Incursions and Callsign
similarity, in terms of the number of incidents per 10,000 flights. It is interesting that the yearly trends during 2008-2010 in EVAIR
0,15
and IATA STEADES databases are very similar. 2008 had a higher level than 2009 and then in 2010 at the European and global level
0,10
the number of Level Busts increased. Thus the overall trend in both databases shows an increase for this three-year period. At the
European level, the EVAIR data for 2010, within manual reporting, 0,05
shows that there is almost one Level Bust daily during the summer
season. The largest number of Level Busts occurred within the En-route phase. However, data collected automatically from one service
0,00
provider, shows that the Level Bust situation is much more problematic in TMAs, especially in busy
2008 and complex areas.
2009 2010Therefore for
better understanding and potential prevention activities, the introduction of the automatic data collection mechanisms such as ASMT
is recommended, at least in areas of high traffic density.
0,25 1,2
Reports per 10.000 flights
400
Number of reports
0,20 1,0 350
0,8 300
0,15 250
0,6 200
0,10
0,4 150
0,05 100
0,2
50
0,00 0,0 0
2008 2009 2010 2008 2009 2010
Incidents rate per 10.000 flights © International Air Transport Association (IATA). All rights reserved. No part of this graph
may be reproduced, recast, reformatted or transmitted in any form by any means,
Figure 9 - Level Bust 2008-2010 elctronic or mechanical, including photocopying, recording or any information storage
and retrieval system, without written consent of IATA, Senior Vice President, Safety,
Operations & Infrastructure.
Previous EVAIR Safety Bulletins have featured the monitoring activity involving the EUROCONTROL Callsign Similarity (CSS) Project. This
project aims to reduce the incidence of Callsign Similarity/Confusion by establishing pan-European CSS solutions centred on a coordi-
nated Callsign Similarity Service and Tool provided by the Callsign Management Cell (CSMC) established in EUROCONTROL’s Directorate
Network Management (formerly CFMU).
Call Sign Similarity Tool Development of the EUROCONTROL Callsign Similarity Tool (CSS Tool) is ongoing. At the time of writing, the
manual and semi-manual version is being tested by a number of aircraft operators. More AOs are encouraged to join the testing regime
and those that are interested should contact the CSMC at cfmu.csmc@eurocontrol.int for more details.
The aim is for the evaluation of the initial CSS Tool to be completed using the IATA Winter Season 2011/12 schedules. Currently, the test
version is only available to AOs via a CFMU NOP Portal Test Platform; the intention is that it should be available for all AOs on the full
CFMU NOP Portal in Spring 2012 when Service Level 1 operations will begin.
Note: CSS Service Level 1 is the detection and deconfliction of callsign similarities within a single aircraft operator’s schedule.
Thereafter, as shown in Fig 11, a second, fully-automatic version of the Tool is planned to be released for testing in September 2012.
Unfortunately this is a year later than originally planned but it remains a medium-term aim of the Project. In the interim, the manual
and semi-manual version of the Tool will still offer AOs a useful and credible means to deconflict their flight schedules.
CSS Tool Development Work Packages: WP1 (Manual and Semi-Manual Deconfliction)
WP2 (Automatic Deconfliction)
2009 CSMC Service Level 0 (including ex DSNA task) until Spring 2013
* AO Involvement
Figure 11 - CSS Service Level/Tool - Timeline
These rules were also used as the baseline set to be configured in the initial version of the CSS Tool. However, it soon became apparent
that when applied to aircraft operators’ schedules (in particular the larger ones) certain rules were too broad in nature and resulted in the
detection of too many similarity conflicts.
Thanks to data collected by EVAIR, the application of some rules has now been refined. By analysing actual callsign similarities and
confusions occurrences reported to EVAIR, it was possible to identify the most common patterns/ combinations of numbers and letters
used in the callsign suffixes that occurred in real-life operations. Consequently, a number of rules were changed and many of the more
obscure similarity conflicts are no longer detected. This has significantly improved the overall performance of the CSS Tool and increased
its credibility amongst the testing AOs.
Callsign Similarity Monitoring As the CSS Project progresses, it will be necessary to monitor the effectiveness of the CSS Tool by those
aircraft operators using it in ‘live’ operations. It is hoped that a number of the AOs engaged in the testing will come forward and volunteer
to use the output from their work detecting and deconflicting their IATA winter season 2011/12 schedules for this purpose.
The CSS UG has agreed that the existing EVAIR communications channels should be used for the transmission of Callsign Similarity/Confusion
reports to EUROCONTROL. Namely, as with other air safety reports, these can be sent to mailbox, Dragica.stankovic@eurocontrol.int.
Pleasingly, some EVAIR members are playing an active role in the CSS UG. Furthermore, many EVAIR colleagues are already contributing
CSS reports to EVAIR. In particular, a growing number of ANSPs that are monitoring callsign similarities in their areas of operation are
providing callsign similarity data to EVAIR. This is excellent news and will become even more important in the future when the ANSPs’
monitoring role really kicks-in as more and more AOs start to use the Tool.
As an example, the Czech ANS provider has recently forwarded a significant amount of CSS data gathered from recorded radar and FPS
data to capture the number of conflicts when two aircraft with similar callsigns (according to the rules) crossed through their airspace in
the same sector and on the same frequency. This data reinforces information received from other ANSPs and also confirms previous asser-
tions that detecting and deconflicting callsign similarities between aircraft from different airlines should be a long-term goal of the CSS
Project. Similar contribution come regularly to EVAIR from ENAV, the Italian ANSP, as well as from France’s DSNA and Aena, Spain’s ANSP.
As described previously, the data already supplied by ANSPs has played a very useful role in determining a more representative and refined
application of the similarity rules inside the Tool. Moreover, the growth in reporting has led to increased contact with aircraft operators
who have been asked to conduct ad-hoc deconfliction on reported similarity events. This contact has also provided a useful recruitment
opportunity and a number of new airlines have joined the CSS UG/Project. More importantly though, the changes the AOs subsequently
make to their callsign schedules based on this feedback are reducing the potential for the same similarities to occur again. This
should reduce the risk and have a positive impact on operational flight safety which is the ultimate aim of the CSS Project.
For those AOs and ANSPs who are already submitting CSS data to EVAIR, we thank you. For those of you who are not, there is an open
invitation to begin sending us such data now.
Important Reporting Note: It is requested that reports sent to EVAIR use the ATC Call Sign or operational Flight Identification rather
than the Commercial Flight Number (CFN) (unless they are the same) as this helps the EVAIR staff process the data more easily and efficiently.
0,06
If you are interested in learning more about the CSS Project then please contact the CSS Project Manager and co-chair of the CSS User
Group, Mr Richard Lawrence, at: richard.lawrence@eurocontrol.int or0,04
via callsign.similarity@eurocontrol.int
0,026
0,02
The latest data reported to EVAIR is shown below.
0,00
2008 2009 2010
160
0,085
Number of reports
0,08 0,4 140
120
0,06 0,3 100
80
0,04 0,2 60
0,026 40
0,02 0,1
20
0,00 0,0 0
2008 2009 2010 2008 2009 2010
Incidents rate per 10.000 operations © International Air Transport Association (IATA). All rights reserved. No part of this graph
may be reproduced, recast, reformatted or transmitted in any form by any means,
Figure 12 -call sign confusion 2008-2010 elctronic or mechanical, including photocopying, recording or any information storage
and retrieval system, without written consent of IATA, Senior Vice President, Safety,
Operations & Infrastructure.
Callsign confusion reports Figure 12a -Call Sign Confusion 2001-2010
Callsign conf. reports per 10,000 flights
STEADES
0,6 Linear (Callsign conf. rpts per 10,000 flights) 200
180
0,5 It is very interesting to see that European and global Callsign similarity trends through 2008-2010 are very similar. In both areas, European
160
and global, 2009 had a higher number of occurrences than140 in 2008 and 2010. As already explained, in general the global STEADES rate is
0,4
much higher than EVAIR. 120
0,3 100
80
0,2 60
0,1 40
20 Mistakes
0,0 Spoken Communication through the monitored period 0
2008 2009 2010
had the highest rate. After an increase in 2009 it went
Spoken comm.
slightly down in 2010. As said above, the reason could be
the influence of the European Callsign similarity project. 2010
2009
Oper. Comm. Issue 2008
In-depth analysis of the operational Use of the English Language in ATM around Europe
Aviation represents a domain which arguably more than any other, requires world-wide standardised and harmonised regulations and
practices. As in other activities, common language and communication facilities represent the main means for exercising a job. But it
seems that in no other industry is the use of language and communications more critical, and where misunderstandings in pilot-air traffic
controller communication is so potentially disastrous, than in air traffic control.
The reason to dedicate attention to the language issues in this issue of EVAIR is that in our database the language (operational and spoken)
is the most frequent incident contributor to different type of events. This is also EVAIR’s contribution to the ongoing language discussions
across Europe and a trial to make the best use of the collected data.
Most EVAIR incidents come from pilots through their airlines’ SMSs. Flying across the world every day, pilots have a clear perspective of the
knowledge of the English language in air traffic control (ATC). The Air Traffic Controllers (ATCOs), on the other hand, have a direct line to the
different airlines and their pilots’ knowledge of the English language. EVAIR reports show that language problems are pan-European issues.
The graph below which represents the cumulative figures of the main incident contributors for the period 2008-2010, shows that
Air-Ground communication (Operational and Spoken) make 32% of the main contributors.
ATC clearance/
Traffic & instruction related item
airspace problems 2%
15%
Oper. & spoken comm.
Coordinations 31%
Issues
6%
Lapses Mistakes
6% 29%
The most frequent Air-ground communication problems are shown in the graphs below (Figures 15 and 16)
R/T monitoring
sector
6%
Language/accent
Poor/no coordination 7%
16%
Misunderstanding/
Pilot breach of R/T interpretation
Phraseology 7%
30% 1%
Hearback ommited
29% Other pilot problems Noise interference
5% 2%
Figure 15– Drill down through Figure 16– Drill down through
Operational communication 2006-2010 Spoken communication 2006-2010
The most frequent events that have language as one of the causes are: Level Bust, Callsign Confusion, Go-around and Runway and Taxiway
Incursion/Excursion.
Below, we give two incident examples related to the English language proficiency and the use of English and the national language in a
single environment on the same operational radio frequency.
EXAMPLE: Short summary of an incident with language proficiency as the contributory factor
After take off, the landing gear lever got stuck. In addition, an engine problem appeared. The pilot requested to be allowed to make a holding
pattern due to technical problems. Quality of ATC radio telephone (R/T) communications and English proficiency was quite poor. R/T communi-
cation consumed half of the entire time required to solve the technical problem! Assistance came from the company pilot on the jump seat, who
spoke the local language.
EXAMPLE: Short summary of an incident with the use of two languages in a single environment and the problem of pilot’s situational
awareness
The incident occurred with three aircraft; one just landed, the second one was ready for take off, and the third one was in the approach phase
approaching the final. The communication with the landing aircraft was in the national non-English language. After landing the a/c stayed on
the runway longer but that was not known to the non-national language speaking pilots. The departing traffic communicated with the ATC in
English and got clearance for take off, which was ATC’s mistake. Due to lack of situational awareness, the pilot of the departing traffic accepted
the clearance. Take off was aborted at a speed of 60kts. The third a/c made a Go Around since there were two a/c on the runway. The pilot of the
departing traffic stated that if he had been aware of the communication with the landing traffic he would have been in a position to warn ATC
and correct the mistake. The use of the national language reduced general awareness and possible correction of the mistake.
Worldwide common practice and operational and safety reasons have made English the Number One language in aviation, but not the
only one. On the operational frequency, as can be seen from the above incident, it is not uncommon to hear two different languages,
which could create a problem to those who do not know/understand both of them. The use of two different languages accounts for 10% of
the spoken communication problems across Europe according to the EVAIR data. The need to have only one language on the operational
frequency, when pilots and air traffic controllers communicate, opens a lot of issues amongst which the most important are: regulation,
safety and efficiency but very often followed by historical, cultural and political discussions.
Political and cultural reasons could create obstacles for the use of English as the single language in airspace designated for international
use. In a number of states across the world, the use of English in aviation communication is not regulated by national regulation or if it is,
then the national language has the same priority as English. This is the origin of various problems.
The best solution from a safety and efficiency point of view, supported very much by the airline associations, is the use of the English
language as the only one at airports and in the airspace designated for international use. Unfortunately, as previously mentioned, currently
this is not always possible due to political, cultural and historical reasons.
Improvements in the training field, refresher courses and periodic checks flexible enough to take into consideration areas of concern
identified through the reporting process, could mitigate the situation and bring positive results and ensure an appropriate level of English
language knowledge. Besides R/T phraseology, attention has to be given also to proficiency in plain language, as phraseology alone
cannot satisfy all communication requirements. This should concern both those to whom English is a foreign language but also those to
for whom English is the mother tongue.
The previous EVAIR Safety Bulletins have highlighted the growing menace to aviation caused by the misuse of hand-held laser devices
which are being used to ‘target’ aircraft and sometimes ATC facilities. In the EVAIR Safety Bulletin No 6, which covered the summer seasons
there were laser interferences at 74 different locations within 24 states. However, in this issues which covers the full years from 2008-2010,
the trend increased not only in the number of interferences but also in the number of locations. Now we have identified 100 different loca-
tions within 27 states from the reports we have received from 53 AOs.
The dramatic growth in numbers prompted the decision for EUROCONTROL to host a multi-disciplinary Laser Seminar (10-11 Oct 2011). The
aim was to gather experts from different fields to begin the search for potential answers and possible proposals to mitigate the situation at
national and international levels. It is recognised that if the fight is to be won then it will need concerted and coordinated effort by aviation
stakeholders working in conjunction with local authorities, laser manufactures, medical experts, police, prosecutors and law makers.
Landing Standing
0% 0%
Tendency of Laser interferences per 10,000 operations
Yearly development per 10,000 operations En-route Take-off
1,2 2% 6% Taxiing
EUROCONTROL
0,098 0%
1,0
0,8 0,085
0,6
0,4
0,2
0,026
0,0
-0,2
2008 2009 2010
EVAIR ACAS monitoring aims to ensure the continued safe and effective operation of ACAS by identifying and measuring issues associated
with RAs and their trends and taking preventive measures where necessary.
ACAS is the generic term for Airborne Collision Avoidance Systems, of which TCAS II is the only example of implementation so far. ACAS
is intended to improve air safety by acting as a ‘last-resort’ method of preventing mid-air collisions or near collisions between aircraft.
Although ACAS II implementation was completed in 2005, ACAS monitoring continues to improve safety by identifying technical and
procedural deficiencies.
ACAS data have been collected either automatically via the Automated Safety Monitoring Tool (ASMT) developed by EUROCONTROL, or
manually thanks to airlines and ANSPs reporting.
It should be noted that ACAS / TCAS statistics from manual reporting rely on pilots’ and controllers’ perceptions and memories of
the events rather than measured or calculated values. Therefore care is needed when comparing manually collected data and data
that are captured automatically. Messages about typical performance should generally be taken from the Automatic recording
of events. Manual reporting tends to give emphasis to more significant events and insights into perception of the ACAS II system.
In 2010 through the automatic data collection from twelve radars a total of 8152 events with valid RAs were recorded in the EVAIR data-
base. Among them there were 203 complex events that had more than two different RAs. To ensure consistent continuity of the statistical
analysis, data from only one of the radars covering a busy airspace is compared with previous years. The reason to compare the previous
period with data from only one radar is because the other 11 were not previously included in the system. It is unfortunate that monitoring
for that one radar was not continuous for the whole period in 2010 that has approximately one month worth of missing data.
l Useful RA - The ACAS II system generated an advisory in accordance with its technical specifications in a situation where there
was or would have been a risk of collision between the aircraft.
l Unnecessary (Nuisance) RA - The ACAS II system generated an advisory in accordance with its technical specification in a situa-
tion where there was not, or would have not been, a risk of collision between the aircraft.
l Unclassifiable RA - The ACAS II system generated an advisory that cannot be classified because the manual report contained
insufficient data.
RA INSTRUCTIONS 2008-2010
Within EVAIR manually collected data, ACAS reports have decreased in 2010 and have lower levels than in 2009 and 2008. We believe
that the reasons behind this are the number of initiatives taken by the aviation community to improve ACAS operational performance.
EUROCONTROL contributions amongst the others in this regard, are very wide and well accepted across the aviation community. ‘Reduce/
Adjust’ RA is in general the RA that occurs with the greatest frequency. In 2010 the ‘Reduce/Adjust’ RA decreased by 11% versus 2009.
The 2nd most frequent RA instruction is ‘Climb’ RA, which decreased by 6% for the whole 2010 versus 2009. However during the summer
season this reduction in 2010 versus 2009
was 28%; with the available information it is
difficult to explain the reasons behind. One Climb RA
of the areas for the improvement is within
training. Training regimes have the flexibility Crossing RA
Reduce/Adjust RA
140
120 2010
2009
100 2008
80
60
40
20
0
50
60
70
80
90
98
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
310
320
330
340
350
360
370
380
390
410
Rate of incidents - absolute figures
Figure 22 - ACAS Flight Level Distribution 2008-2010
Statistics related to the FL distribution of the manually collected ACAS incidents show that the ACAS incidents are located in clusters
of FLs. For the period 2008 – 2010 absolute figures, the ACASAS RAs are grouped in 3 big areas: FLs between 50-170; 180 – 250; and
330-370. In general, the largest number of RAs occurred between FL 50 – 130. However, in 2010 there was an overall decrease in the
number of RAs and the most affected levels were between FL230 and FL 250 and FL 310 and FL 350. Geographical location of incidents
and comparison with the route network structure as well as sectorisation could give more information whether some of these incidents
have been generated by the airspace design and sectorisation structure or some other reasons. In the next issue, more attention will
be dedicated to this type of analysis.
Tim Baldwin
Analysis of Automatically provided Garfield Dean
ACAS RAs ACAS Expert
Tim.baldwin@eurocontrol.int Garfield.Dean@eurocontrol.int
Automatic monitoring of ACAS RA downlink messages has been expanded in the last 12 months to cover larger numbers of radars in
Europe. The procedures for monitoring have been changed by using the Automated Safety Monitoring Tool (ASMT) to monitor Radar LAN
recordings rather than being connected on line. Standard radar LAN recordings are transferred to ASMT on a daily basis and the detection
process in ASMT is run once for each ModeS radar source storing detected events in a dedicated radar community database. This process
is now used by two ASNP’s.
The total numbers of RA downlink events recorded by ASMT attached to 12 radars in Europe is shown below:
All radars
Events with valid RA messages = RA downlink messages which a valid RA downlink code.
Events with erroneous RA messages = RA downlink messages which are empty or contain only a stop bit.
Complex RA events = Events where more than two different RAs are detected for one aircraft during the encounter.
Note: many of the events are seen by up to 7 radars and so result in multiple recordings.
ASMT is being used to record and analyze all ACAS downlink messages. The set of data presented below covers one radar head, to enable
comparison of statistics through 2007-2010. The data has been collected by ASMT from the same ModeS radar in European airspace.
During 2007 the monitoring was not continuously available due to radar maintenance and a change to the ASMT monitoring machine.
In 2008 automatic filtering of erroneous messages from selected airframes was also introduced. In 2009 additional filtering of erroneous
messages was introduced to reject all continuous downlink transmissions of messages with only the stop bit set hence a large reduction
in numbers of erroneous downlink messages in 2009 and 2010 (downlink anomalies).
Although there are many erronous downlink messages these appear to be generated by a small number of aircraft fitted with Honeywell
TCAS units and Honeywell Enhanced ModeS transponders that continuously report RAs. With the addition of a new ANSP providing data
a new group of airframes has been detected exhibiting the same phenomena. There is a much larger group of aircraft that generate a
downlink message with only the RA stop bit set. With these aircraft it is a set of three erroneous messages sent once or twice per flight and
we have not yet found a common factor.
Detection and correction of RA downlink anomalies is now being handled by the ModeS monitoring team.
Downlink anomalies
In summary, a small number of aircraft are non- compliant with the SARPS and ‘pollute’ the RF environment. Corrective action is underway
with the operators and manufacturers concerned.
For the 10 radars the graph below indicates the number of aircraft reporting RA downlink to one or more radars. Some military actually
flew through the coverage of more than one radar and reported RAs in different geographic locations. The numbers here are filtered using
ModeS code and therefore represent a clearer picture of total events but still include duplicate counts for coordinated events;
200
186
180
164 167
160
143 144
140 131
120
100 97
80
60
42
40
20 11
1
0
1 2 3 4 5 6 7 8 9 10
Data from the year was analysed and some aircraft were detected as reporting a valid RA on more than one occasion, the Modes S code
was looked at for those reporting on multiple days and all reporting on 4 or more days were identified as military. It is very unlikely that
commercial airframes will get more than one RA in a year.
900
793
800
700
600
500
400
300
200
157
100
49
28 18 12
0 5 5 1 1 1 1 1 1 0 0 4 0 1 2 1 1 0 1 2 0 0 1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Figure 24 - Days reporting by airframe
Figure 25 below shows the number of RAs by Flight Level and compares 2007, 2008, 2009 and 2010. All recorded RAs with a valid down-
link message (not blank and not stop only) have been counted even if the intruding aircraft has not been identified. FLs in this report are
rounded to the nearest FL. There is no major trend difference between the 2007, 2008, 2009 and 2010 distributions except for a reduction
in higher level RAs.
500
400 2010
2009
300 2008
200
100
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
310
320
330
340
350
360
370
380
390
400
410
420
The diagram below shows the number of RAs events recorded on both aircraft and the number of events where only one aircraft reported
an RA.
Events shown as an RA with a Mode S intruder are to a very large extent ACAS equipped. ACAS does not symmetrically generate RAs.
There are many events where only one aircraft receives an RA, even though both are ACAS equipped. In 2010 both aircraft reported an RA
in 11% of the recorded events. RA with intruder Mode C are confirmed as Mode C intruders by the Own RA downlink. In the lower levels
there is an increased percentage of aircraft identified as Mode S intruders and a reduction in Mode C intruders, which may be due to the
increase in the number of General Aviation aircraft fitting Mode S transponders.
RA by FL and YEAR
Figure 26 - TCAS equipage in Encounters
Note: The number of events at each level is lower than the number of RAs because more than one RA can exist for each event.
At lower flight levels, the majority of RAs are against Mode C intruders, whereas at higher levels, most RAs are against intruders with Mode
S or are coordinated ACAS encounters. A significant reduction of level off RAs in upper flight levels is possibly due to the increased use of
25ft reporting Mode S transponders.
Plotting XY distributions of RAs at different altitude ranges allows “RA hotspots” to be identified. Frequently these are related to airspace
design and airspace classification issues, e.g. where climbing aircraft level off just below a flight level where descending aircraft level off
too. As an example, a hot spot diagram is shown below. Better conclusions would have been given if the route network had been shown.
200
150
100
50
0
-200 -150 -100 -50 0 50 100 150 200
-50
-100
-150
-200
250
200
150
100
50
0
Head-on Acute Crossing Obtuse Chasing Changing
Figure 28 - Horizontal Geometry (2010 only)
As expected this shows that the majority of RAs occur with acute and crossing tracks with 60% in this area. The low number of head on
geometry could be due to the fact that head on traffic is not normally climbing or descending to within 1000ft in close proximity, when on
the same airway in opposite directions. However it is usual to have aircraft aiming for 1000ft separation at airway crossing points.
40%
30%
20%
69% 27% 4%
10% 58% 29% 13%
0%
Normal High Level
For an RA that does not change for at least 2 radar cycles, i.e. 10 seconds ASMT automatically calculates compliance with the RA. We see that
81% of pilots have at least adequately responded to the RA. There were 73% achieving the requested vertical rate and 8% exceeding the
requested rate. 19% of responses were not satisfactory from a safety perspective with 12% either slow or failing to achieve the requested rate
and 7% were calculated to have made opposite responses. These are similar numbers to 2009. The numbers here are for all RA’s from a single
900
radar and include those from Military training and transport aircraft. Most of the military will fall in the short RA sample in the diagrams below.
811
800
731
Opposite Excessive
700 Followed
655 7% 8% 2008
27%
600 2007 Below
Partial required rate
Followed 25%
500 followed 73%
468
12%
419
400
351
300
236
200 192
100
33 27
Excessive
13 10 13 13 20
0 0 9 8 1 9 5 6 10%
MVS KVS AVS CL ICL CCL RCL DE IDE CDE RDE
Opposite
38%
For short RAs (353 events - 39%), where the RA is observed to change after one radar cycle (changes in less than 10 seconds), compli-
ance with the RA is not as good, even though ACAS expects a quicker response from pilots. The ACAS logic expects an initial response
to an RA in 5 seconds and only if the RA exists for less than 6 seconds do the responses here appear to be reasonable.
The high level of opposite responses to changing RAs is a cause for concern and re-emphasizes the need for pilots to be trained to follow
the full sequence of RAs that they receive.
900
“Pilots should use appropriate procedures by which an aero-
800
plane climbing or descending to an assigned altitude or flight
700
level, especially with an autopilot engaged, may do so at a
600
rate less than 8 m/s (or 1 500 ft/min) throughout the last 300
500
m (or 1 000 ft) of climb or descent to the assigned altitude or
400 flight level when the pilot is made aware of another aircraft
300 at or approaching an adjacent altitude or flight level, unless
5 6
200 otherwise instructed by ATC. These procedures are intended to
RDE
100 Normal avoid unnecessary airborne collision avoidance system (ACAS
High
0 Level II) resolution advisories in aircraft at or approaching adjacent
2007 2008 2009 2010 altitudes or flight levels. For commercial operations, these
ADVISORIES ISSUED
The number of RAs is greater than the number of events for two reasons: more than one RA can be issued to an aircraft during an event;
and both own aircraft and intruder can issue an RA. The table below shows the breakdown of advisories issued in the collected data.
The reduction in the number and percentage of MVS (monitor vertical speed) RAs is probably due to the increased use of ModeS by
general aviation aircraft at lower flight levels.
The ACAS advisories AVS and MVS do not require deviation from ATC clearance unless ATC requests a specific vertical rate. They correspond to
62% of RAs in 2010 similar to 2008 and 2009 but a drop of 4% from 2007.
900
811
800
731
700 2010
655
2009
600 2008
2007
486
500
468
420
419
400
351
282
278
300
242
236
205
192
200
147
130
100
33
27
17
19
13
10
17
13
13
12
20
17
11
12
9
8
8
8
11
9
11
0
6
0
5
5
6
7
0
0
1
0
MVS KVS AVS CL ICL CCL RCL DE IDE CDE RDE MUL
To obtain the statistics presented below all RA downlink messages from all Radars were merged into a single table and sorted on time.
These were then filtered to eliminate duplication by selecting only the first message from a series collected from the same ModeS address.
350
300
250
200
150
100
50
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
310
320
330
340
350
360
370
380
390
400
Figure 33 - Events by Flight Level
Due to the method of extracting the statistics, the results here represent the number of events rather than the number of RA’s as shown
in figure 4 above covering four years.
Times were extracted for one of the radars to get an estimate of warning time that a controller might expect from RA downlink. The
measurement was taken from receipt of first downlink to calculated Horizontal Closest Point of Approach (CPA) and is shown at one
second intervals.
200
180
160
140
120
100
80
60
40
20
0
1
13
17
21
25
29
33
37
41
45
49
53
57
Time in seconds
Figure 34 - RA downlink to horizontal CPA
The ASMT database was interrogated with a Query that selected all events with a horizontal CPA of 0.2nM and vertical separation of less
than 700ft. This set was individually assessed using InCAS to check what the separation might have been if the aircraft had continued
with its vertical rate rather than responding to ACAS advisory.
One of the events between GA aircraft is described below showing that following the RA increased vertical separation.
AC1 was a CESSNA C560 Simulation from radar data indicates 240 knots at FL 21.5 and AC2 was a PA32on a reciprocal heading at 170 knots FL 21.
Simulation suggests that AC 1 received a Traffic Advisory at 08:32:01 when AC2 was 2.8nm head-on with a closing speed of 410mph. Simu-
lation suggests that at 08:32:11 AC1 received a Resolution Advisory to climb when AC 2 was now 1.7nm ahead, still on reciprocal heading.
At 08:32:14 the radar received a downlink message from AC1 reporting a climb advisory. The simulation suggests that AC1 started to climb
at 1500fpm as requested by ACAS at 08:32:18 and the intruder AC 2 passed 223ft below with 0.04nm horizontal separation.
Figure 35
Reducing Level busts is one of EUROCONTROL’s highest priorities. EUROCONTROL began raising awareness of the Level Bust issue in 2001,
organised series of workshops, and established a Level Bust Task Force to define the recommendations and to formulate an action plan to
reduce level busts.
The level bust action plan is the product of work carried out by EUROCONTROL’s cross-industry Level Bust Task Force, which was set up
in 2003. The Task Force reviewed the evidence available, identified the principal causal factors, and listened to the Air Navigation Service
Providers and aircraft operators with experience in reducing level busts.
The Action Plan contains recommendations for Air Traffic Management, Air Traffic Controllers, and Aircraft Operators. It designed to reduce
the frequency of level busts and reduce the risks associated with level busts. Implementation of the Action Plan will be monitored by the
Task Force monitoring group reporting to the EUROCONTROL Safety Improvement Sub Group (SISG).
http://skybrary.aero/index.php/Solutions:Level_Bust
THE EUROPEAN ACTION PLAN FOR THE PREVENTION OF RUNWAY INCURSIONS (EAPRI)
Although runway safety includes issues such as foreign objects, debris and wildlife straying onto the runway and other technical deficien-
cies, this action plan specifically addresses the subject of runway incursion prevention.
EAPRI is the result of the combined efforts of organisations representing all areas of aerodrome operations that are totally committed
to enhancing the safety of runway operations by advocating the implementation of the recommendations that it contains in the ECAC
area. The ICAO secretariat has lent its strong support to the work of this group and urges all States to fully implement the ICAO provisions
relevant to runway safety.
The version 2.0 of EAPRI came to light n April 2011 and contains more than 56 recommendations. When implemented, it will enhance
runway safety by the consistent and harmonised application of existing ICAO provisions, improving controller - pilot - vehicle driver
communications and working procedures at the aerodrome, and by the subsequent increase in situational awareness.
http://www.eurocontrol.int/runwaysafety/public/standard_page/EuropeanAction.html
The European Action Plan for Air Ground Communication Safety (conceived inter alia by EUROCONTROL, aircraft operators (AOs) and the
Flight Safety Foundation) identified Call Sign Similarity (CSS) as a significant contributor to air-ground communication issues. Analysis of
ATC reported events shows that 5% involve incidences where CSS is involved. Some aircraft operators are trying to find solutions; the only
known ANSP actively operating a service to de-conflict call signs is France’s DSNA.
Research and CBA studies show that the most cost efficient way of providing a long-lasting, Europe-wide solution is to create a central
management service to de-conflict ATC call signs. This strategy provides economies of scale and rapid pay back of investment (3 years).
More importantly, it is calculated that it will eliminate over 80% of the CSS incidences and thus improve safety.
http://www.eurocontrol.int/safety/public/standard_page/Callsign_Similarity_project.html
Following definitions are extracted from the HEIDI and/or HERA Taxonomies.
HEIDI (Harmonisation of European Incident Definitions Initiative for ATM) intends to finalise a harmonised set of definitions (taxonomy)
for ATM related occurrences.
HERA (Human Error in European Air Traffic Management) develops a detailed methodology for analysing human errors in ATM, including
all error forms and their causal, contributory and compounding factors.
ATC clearance/instruction (HEIDI): Related to incorrect or wrong aircraft action. Authorisation for an aircraft to proceed under conditions
specified by an air traffic control unit and deviations from the clearance which cause runway incursions, taxiway incursions, apron incur-
sions, level bust, unauthorised penetration of airspace etc.
Coordination (HEIDI): Internal coordination encompassing coordination with sectors within the same unit, and sectors within the ATC
suite; external coordination, civil/civil and civil/military; and special coordination, covering expedite clearance, prior permission required,
revision and other special coordination.
Call sign confusion: The major cause for aircraft taking a clearance not intended for them.
Contributory factors (HEIDI): A part of the chain of events or combination of events which has played a role in the occurrence (either
by easing its emergence or by aggravating the consequences thereof ) but for which it cannot be determined whether its non existence
would have changed the course of events.
Lapses (HEIDI): Psychological issues encompassing: Reception of information, Identification of information, Perception of information,
Detection, Misunderstanding, Monitoring, Timing, Distraction, Forgetting and Loss of awareness.
Level bust (HEIDI): Any unauthorised vertical deviation of more than 300 feet from an ATC flight clearance Departing from a previously
maintained FL, overshooting, undershooting, levelling-off at a different level than cleared level.
Mistakes (HEIDI): Psychological issues encompassing: Information wrongly associated, Workload issues, Information not detected,
Failure to monitor, Recall of information, Misunderstanding or insufficiently learned information, Judgement, Planning, Decision mak-
ing, Assumptions and Mindset.
Missed approach/Go-around (HEIDI): The operating phase where the aircraft flies the procedure to be followed if the approach cannot
be continued.
Operational communication (HEIDI): Air-Ground, Ground-Ground and Use of equipment verification testing. Air-Ground communica-
tion encompasses hear back omitted, pilots’ read back, standard phraseology, message construction, R/T monitoring including sector
frequency monitoring and emergency frequency monitoring, handling of radio communication failure, unlawful radio communications
transmission. Ground-Ground communication refers to the standard phraseology, speech techniques, message construction, standard
use of equipment like, radio frequency, telephones, intercoms, etc.
Crossing Crossing
Limit Limit
Crossing
Chasing -Head-On
Obtuse Acute
Limit Limit
Chasing Head-On
Chasing -Head-On
Obtuse Acute
Limit Limit
Crossing
Crossing Crossing
Limit Limit
Spoken communication (HEIDI): Human/human communication encompassing air-ground and ground-ground communications but
also call sign confusion, noise interference and other spoken information provided in plain language. Air-ground communication refers to
language/accent, situation not conveyed by pilots, pilot’s breach of radio telephony (R/T), workload, misunderstanding/misinterpretation,
and other pilot problems. Ground-ground communication refers to misunderstanding/misinterpretation, poor/no coordination.
Taxiway Incursion (HEIDI): Any occurrence unauthorized presence on a taxiway of an aircraft, vehicle, person or object that creates a
collision hazard or results in a potential loss of separation.
Traffic & Airspace problems (HEIDI): There are four set of causal factors under this element.
l Traffic load & complexity, encompassing excessive and fluctuating load, unexpected traffic demand, complex mix of traffic,
unusual situations (emergency, high risk, other), Abnormal time pressure, underload and call signs confusion;
l Airspace problems composed of flights in non controlled and controlled air space, Airspace design characteristics(complexity,
changes, other) and temporary sector activities(military, parachuting, volcanic activity, training);
l Weather problems such as poor or unpredictable(snow, slush, ice, fog, law cloud, thunderstorm, wind shear);
l Pilot problems concerning language, culture and experience aspects.
Traffic Information (HEIDI): Essential and local traffic information provided by an air traffic controller to the pilot. Essential information is
related to the provision of traffic information containing:
Local traffic in this context consists of any aircraft, vehicle or personnel on or near the runway to be used, or traffic in the take-off and
climb-out area or the final approach area, which may constitute a collision hazard to the other aircraft and about which the information
has to be provided.
Call Sign Confusion Two or more aircraft on frequency with similar Call Signs creating confusion
Other Aircraft - Runway/Taxiway Another aircraft made an unauthorised entry into a runway or taxiway
Incursion
Runway Incursion by Ground Vehicle A ground vehicle entering an active runway or taxiway without ATC permission
creating an unsafe situation
Runway/Taxiway Incursion The aircraft made an unauthorised entry into a runway or taxiway
October 2011 - © European Organisation for the Safety of Air Navigation (EUROCONTROL)
www.eurocontrol.int