RAA Incident Stats

[pmccarthy]

With a bit of time to spare, I analysed the incident data on the RAAust website from 21 Oct 2015 to today. I added up the number of types reported, for all those greater than one. This should give some idea of how active and numerous each type is. Of course, there may be reasons for under- or over-reporting depending on the type. As you might expect, the first three are widely used for training.

 

The source data isn't "clean", and each type may be reported under several headings and spellings.

 

I hope this is interesting.

 

Jabiru 152

 

Aeroprakt 53

 

Tecnam 46

 

BRM Aero 25

 

Evektor 21

 

Airborne 17

 

The airplane factory 16

 

Sling 10

 

Brumby 9

 

Skyfox 8

 

Zenith 7

 

Pipistrel 6

 

Morgan 6

 

Thruster 6

 

Flight design 6

 

ICP Savannah 5

 

Fly Synthesis 4

 

Eurofox 4

 

Alpi 4

 

Cubcrafters 3

 

Aerochute 2

 

Just aircraft 2

 

Avid 2

 

Skyranger 2

 

Sonex 2

 

 

[HeliPilot70]

Very interesting, but the accident rate for each brand would be more informative. Is it possible to get the total number of each brand registered and divide these values into it to give the rate?

 

 

[pmccarthy]

That could be misleading unless we knew the hours flown for each type. The incidents include more "pilot error" than type problems, so it would also not really tell you anything about the safety of a particular type. I just view this data as a reflection of how much flying is being done by each type, assuming the chance of an incident is about the same for all of them.

 

 

[Jaba-who]

That could be misleading unless we knew the hours flown for each type. The incidents include more "pilot error" than type problems, so it would also not really tell you anything about the safety of a particular type. I just view this data as a reflection of how much flying is being done by each type, assuming the chance of an incident is about the same for all of them.

I don’t think you can assume that or anything.Unfortunately it’s a difficult area to get meaningful data. But in reality you need the whole lot. Numbers of aircraft, numbers of hours flying, division of incidents into pilot error vs airframe vs engine mechanical etc.

 

until you have the full range of data any use of the figures is open to question.

 

 

[turboplanner]

Well at least someone made an effort. Those who’ve set up databases would know the numbers are small enough and easy enough to be automatically sorted and reported every month for small change.

 

 

[Yenn]

As previously stated the stats need further analysis, but some things do stand out.

 

Jab has a very big number, but that includes a large number of aircraft and also a high proportion of training time.

 

BRM, I have never seen one flying and I must admit they look good on the ground at the RAAus Narromine fly in. 25 seems a high number. I shall have to look at what was reported.

 

Savannah looks pretty good considering the number of them I see flying.

 

 

[HeliPilot70]

By totaling up the number of aircraft from each manufacturer on this list (unfortunately it is from 2012) RA-Aus aircraft register, in number sequence we get the following:

 

As Yenn mentioned the top listed aircraft brands are very common for ab-initio training, so applying an average number of hours per aircraft per year doesn't work, but the figures are very interesting none the less.

 

2012

 

Jabiru 152 763

 

Aeroprakt 53 69

 

Tecnam 46 132

 

BRM Aero 25

 

Evektor 21 38

 

Airborne 17 236

 

The airplane factory 16

 

Sling 10

 

Brumby 9 11

 

Skyfox 8 106

 

Zenith 7 72

 

Pipistrel 6 19

 

Morgan 6 24

 

Thruster 6 115

 

Flight design 6 27

 

ICP Savannah 5 45

 

Fly Synthesis 4 10

 

Eurofox 4 14

 

Alpi 4 17

 

Cubcrafters 3 10

 

Aerochute 2 223

 

Just aircraft 2 1

 

Avid 2 10

 

Skyranger 2 17

 

Sonex 2 21

 

 

[pmccarthy]

Interesting data. Aeroprakt have increased numbers a lot since 2012, particularly in schools. A lot of Thrusters are probably not Flying, or not Flying where they would make incident reports. Ditto Aerochute.

 

 

[kaz3g]

I told a friend the other day there was no way I'd get into one of those Aerochute things...bloody dangerous.

 

Well, the figures certainly give a lie to my bias, that's for sure.

 

Kaz

 

 

[Roundsounds]

The organisation should periodically review the data to determine accident rates by type/location/phase of flight/nature of Flight/quals of pilot etc. Based on the findings targeted training programs would be developed to correct any issues. So far, I don’t see any evidence of this?

 

 

[turboplanner]

I told a friend the other day there was no way I'd get into one of those Aerochute things...bloody dangerous.Well, the figures certainly give a lie to my bias, that's for sure.

Kaz

[turboplanner]

These are just a few informal figures; they don’t represent all figures or all aircraft, and they don’t have a breakdown of injuries. Something might look good in the raw figures but have a history of serious injuries.

 

 

[Roundsounds]

These are just a few informal figures; they don’t represent all figures or all aircraft, and they don’t have a breakdown of injuries. Something might look good in the raw figures but have a history of serious injuries.

Yes, agree. Why collect this data if it’s not being put to good use. I think it’s seen as a regulatory requirement, but they don’t see the value in analysing it.

[pmccarthy]

These reports are not in the form of a database. It is clear that data such as aircraft and incident type are not entered by pull-down menus but just free text. So to make it possible to analyse I guess would cost $10-20k in setup costs and judging from other comments here there would be some member outrage about spending that money.

 

 

[Roundsounds]

These reports are not in the form of a database. It is clear that data such as aircraft and incident type are not entered by pull-down menus but just free text. So to make it possible to analyse I guess would cost $10-20k in setup costs and judging from other comments here there would be some member outrage about spending that money.

I think you’ll find they are in a data base designed using the same fields as ATSB. This was done several years ago to make the madatory reporting easier, The stuff presented to members being a report of some sort created from the database. The money has already been spent.

[turboplanner]

These reports are not in the form of a database. It is clear that data such as aircraft and incident type are not entered by pull-down menus but just free text. So to make it possible to analyse I guess would cost $10-20k in setup costs and judging from other comments here there would be some member outrage about spending that money.

I’m working with bigger databases in Excel, which automatically analyses the data.

[WJSYBSU]

To quote the bard..

 

There are Lies,Damn Lies and then there are statistics.

 

 

[turboplanner]

That’s OK for a play, but there is also the truth, and that’s what we are looking for, because that ensures good decisions.

 

 

[pmccarthy]

Twas not the Bard, Mark Twain said it was Disraeli.

 

 

[Yenn]

What is published by RAAus is not concise enpigh for us to make any decisions about safety of a type of aircraft or procedure. Some of the reports do little except state that "something" happened "somewhere"

 

I also know for a fact that some get published, without the publisher even understanding what has been said.

 

 

[DingerPPC]

I told a friend the other day there was no way I'd get into one of those Aerochute things...bloody dangerous.Well, the figures certainly give a lie to my bias, that's for sure.

Kaz

Safest form of flying.