Overrun Accidents
How about this database: https://aviation-safety.net/database/
Unfortunately a cross-referenced search cannot be run from the internet, so at some point the Mk. 1 human eyeball/brain has to be used to additionally filter for location (specific country, not EASA area).
Does provide thumbnail "reports" - and links to complete official reports if available.
Running the online filters for:
Contributory/cause > Result > Runway excursion
....produces a worldwide 15-page list dating to 1935. The last four pages cover 2006-present day. Mk. 1 eyeball can filter that down for EASA locations by city/airport (or national flag icon).
If it were me, I'd probably also filter for and skim through some side-related "cause" categories, just to make sure I'd caught all the excursions. Such as:
Cause > Airplane - Engines - reverse thrust/prop ground
Cause > Airplane - Flight control surface - rudder
Cause > Airplane - Undercarriage - brakes
Cause > Airplane - Undercarriage - gear collapse
Cause > Airplane - Undercarriage - tire failure
Cause > Cargo Occurences - Overloaded/cargo shift
Cause > External factors - FOD/Wake vortex/wind/hail
Cause > Landing/TO > Landing - bounced/fast/heavy/late/unstabilized
Cause > Landing/TO > TO - aborted/failed to become airborne/locked controls/wrong runway/wrong config
Cause > Weather - crosswind/windshear
....or any others you find "suggestive" of risking or resulting in an excursion.
You could also try contacting them (Aviation Safety Network) directly to see if there is a way for industry professionals to get access to the whole database to run their own filter set (location PLUS outcome).
https://aviation-safety.net/about/contact.php
Unfortunately a cross-referenced search cannot be run from the internet, so at some point the Mk. 1 human eyeball/brain has to be used to additionally filter for location (specific country, not EASA area).
Does provide thumbnail "reports" - and links to complete official reports if available.
Running the online filters for:
Contributory/cause > Result > Runway excursion
....produces a worldwide 15-page list dating to 1935. The last four pages cover 2006-present day. Mk. 1 eyeball can filter that down for EASA locations by city/airport (or national flag icon).
If it were me, I'd probably also filter for and skim through some side-related "cause" categories, just to make sure I'd caught all the excursions. Such as:
Cause > Airplane - Engines - reverse thrust/prop ground
Cause > Airplane - Flight control surface - rudder
Cause > Airplane - Undercarriage - brakes
Cause > Airplane - Undercarriage - gear collapse
Cause > Airplane - Undercarriage - tire failure
Cause > Cargo Occurences - Overloaded/cargo shift
Cause > External factors - FOD/Wake vortex/wind/hail
Cause > Landing/TO > Landing - bounced/fast/heavy/late/unstabilized
Cause > Landing/TO > TO - aborted/failed to become airborne/locked controls/wrong runway/wrong config
Cause > Weather - crosswind/windshear
....or any others you find "suggestive" of risking or resulting in an excursion.
You could also try contacting them (Aviation Safety Network) directly to see if there is a way for industry professionals to get access to the whole database to run their own filter set (location PLUS outcome).
https://aviation-safety.net/about/contact.php
Join Date: Oct 2015
Location: With Wonko, outside the asylum
Posts: 69
Likes: 0
Received 0 Likes
on
0 Posts
Neither did I say it did, merely that the gear collapsed, but in his post#5 PEI said he was looking for accident/incident reports, and if you are trying to build up a total picture then near misses can be as relevant as full accidents. That is why I included gear failure incidents.
If we are discussing overruns, as I believe we are, it is not of relevance. That is why I made the point. If you're seeking to build a 'total picture' of everything that goes wrong during landing, it would be relevant, but it was an extraordinarily benign event in terms of risk to life, and the insurance market takes care of the rest.
Finally, the global, industry, viewpoint is that near misses are not as relevant. The two classes of events are given completely different treatment.
Hi PEI,
This may not be a lot of use to you but I have been compiling data on approach and landing accidents and incidents for my website at picma.info . Where I have been able to provide a commentary (which is on a different aspect than you are looking at) I have provided a link to the original reports. There MAY be a few in there which you are not already familiar with. List accessed at Accident & Incident Reports Analysed | PicMA
This may not be a lot of use to you but I have been compiling data on approach and landing accidents and incidents for my website at picma.info . Where I have been able to provide a commentary (which is on a different aspect than you are looking at) I have provided a link to the original reports. There MAY be a few in there which you are not already familiar with. List accessed at Accident & Incident Reports Analysed | PicMA
Thread Starter
Steve thanks for the link, I am familiar with the principles in your site (we have corresponded), but was unaware of the event data.
The analysis appears to be more towards what I have been looking for, however the problem of validity remains.
Acknowledging your expert interpretations, these are still founded on the originating data from investigation reports or a range of public data bases often no better to first hand reports.
I believe that some form of speculative analysis will help, particularly when made by those with relevant experience and balanced thought on the subject, unfortunately I have yet to find any regulatory authority to heed such analysis.
The analysis appears to be more towards what I have been looking for, however the problem of validity remains.
Acknowledging your expert interpretations, these are still founded on the originating data from investigation reports or a range of public data bases often no better to first hand reports.
I believe that some form of speculative analysis will help, particularly when made by those with relevant experience and balanced thought on the subject, unfortunately I have yet to find any regulatory authority to heed such analysis.
Thread Starter
Near miss
I started this thread under the wide umbrella of safety statistics and their relation to safety activity. The lack of overruns data stands out in the EASA report, but the principles apply to most, if not all, safety issues.
The shortage of investigative data would be an asset if the number of incidents were few, yet for overruns it is the opposite which is of concern; many hull loss events, fortunately few fatalities, but this does not equate to being safe, nor managing the risk of a fatality from any overrun.
There is a chronic shortage of data suitable for meaningful analysis to justifying safety activity.
An increasingly popular view is that the industry needs to look at what goes right opposed to measuring what goes wrong. So far this thread would support the need for a wider view, ‘what happens every day’, because most of this is ‘right’, but will this data (even if collectible) enable adequate meaning.
The point about ‘near miss’ triggers thoughts about what is a near miss; the gear collapse above is a ‘negative’ event, it's just not put into the box labelled overrun. Thus the choice and definition of safety boxes biases data and thoughts.
Thus what should a definition of ’near miss’ include.
All landings should be based on having a safety margin in the landing distance, but how often are these margins compromised, is this a near miss. One view is that only when the safety of a landing is assured then the remaining runway can be used for roll out; but this is a judgement - proactive.
Daily operations might suggest otherwise, that pilots only use what is perceived to be required to stop in the full length; then change the action as the landing develops - another but different judgement - reactive, requiring expertise. Note several accidents involving decreasing friction on wet rubber deposits towards the end of a runway.
So should our safety statistics include all occasions where the safety margin was compromised - not listing a human error of judgement, but to identify the circumstances where a judgement could be compromised.
One possibility is to use FDR and assess the deceleration achieved vs expected capability; distance is more difficult to capture.
But even if this collects relevant data are they shared for general safety activities within the industry.
No? Which might be one of the many problems with safety data collection; and perhaps understanding this is what I am looking for.
The shortage of investigative data would be an asset if the number of incidents were few, yet for overruns it is the opposite which is of concern; many hull loss events, fortunately few fatalities, but this does not equate to being safe, nor managing the risk of a fatality from any overrun.
There is a chronic shortage of data suitable for meaningful analysis to justifying safety activity.
An increasingly popular view is that the industry needs to look at what goes right opposed to measuring what goes wrong. So far this thread would support the need for a wider view, ‘what happens every day’, because most of this is ‘right’, but will this data (even if collectible) enable adequate meaning.
The point about ‘near miss’ triggers thoughts about what is a near miss; the gear collapse above is a ‘negative’ event, it's just not put into the box labelled overrun. Thus the choice and definition of safety boxes biases data and thoughts.
Thus what should a definition of ’near miss’ include.
All landings should be based on having a safety margin in the landing distance, but how often are these margins compromised, is this a near miss. One view is that only when the safety of a landing is assured then the remaining runway can be used for roll out; but this is a judgement - proactive.
Daily operations might suggest otherwise, that pilots only use what is perceived to be required to stop in the full length; then change the action as the landing develops - another but different judgement - reactive, requiring expertise. Note several accidents involving decreasing friction on wet rubber deposits towards the end of a runway.
So should our safety statistics include all occasions where the safety margin was compromised - not listing a human error of judgement, but to identify the circumstances where a judgement could be compromised.
One possibility is to use FDR and assess the deceleration achieved vs expected capability; distance is more difficult to capture.
But even if this collects relevant data are they shared for general safety activities within the industry.
No? Which might be one of the many problems with safety data collection; and perhaps understanding this is what I am looking for.
Doesn't help move you forward but even when I have through multi stage searches managed to find a report, the quality of analysis is often pretty abysmal. But I'll alert you if I find anything that will help.
The report shows just under 250 2007-2016 commercial air transport accidents, of which around 70 were runway excursions, and identifies safety issues relating to each category. Clearly there isn't enough space to list the causal factors identified for individual occurrences, but that doesn't mean the data doesn't exist.
Thread Starter
Dave, perhaps not a conclusion that data does not exist, but an observation that without a clear reference to EASAs data we don’t know how accurate it might be, nor if everything has been accessed. Also, without supporting analysis for the categorisation it is impossible to understand the relevance of each contributor.
It's easy to tick a box based on a probable cause / contribution; e.g. human factors can affect everything, tick, similar for organisation, tick, but how does this contribute to a safety programme?
It's easy to tick a box based on a probable cause / contribution; e.g. human factors can affect everything, tick, similar for organisation, tick, but how does this contribute to a safety programme?
