Helicopter accidents

[Nigel Osborn]

There still seems to be far too many fatal accidents around the world, one can understand if it's war time activity but not general flying.

At Flight Safety in Florida some 20 years ago, one of their excellent instructors had done a very comprehensive check on all accidents around the world. His startling finding was that 94 % ( I hope my memory is correct!) were caused by pilot error & only 6 % by mechanical failure. This seemed unbelievable but he showed me the figures & it looked true.

He included in pilot error cases where mechanical failure was incorrectly handled leading to a fatal accident. For example a S 76 had a number 1 engine fire but closed down number 2 & then number 1 failed. Above cloud, over hills, result 14 dead. A Wessex 5 had a number 1 run away up, closed down number 2, number 1 then over speed tripped, result 4 dead.

The accidents in the last few months all look like they could have a pilot problem causing the fatalities. CFIT is common, everyone knows it's dangerous but it still happens too often.

What can be done to avoid these unnecessary deaths? During 45 years of various types of flying in dangerous countries like Borneo, PNG, Tasmania & other places, it struck me it was important to learn to fly in such a manner that you didn't cause an accident, a bit like learning defensive driving when you learn to drive a car.

What do other guys think or am I wasting my time?:confused:

[Pete O'Tewbe]

a bit like learning defensive driving when you learn to drive a car.

Sounds like you are describing Threat and Error Management (TEM).

[heli1]

40 years of recording helicopter accidents in "Accident Spot" in Heli Int tells me that the survey is not far off,Time and time I see the same simple mistakes repeated.
The wise words I took in many years ago were" Don't fly if you've just had a row with the wife" and discipline yourself to always carry out a full preflight check,irrespective of of any previous check done. Bottom line is ,most accidents are caused by pilot inattention,not error?

[Thomas coupling]

CRM, MCC, Human Factors.....where have you been Nigel?
Human error will always be the number one problem because it is completely out of the control of the "system". Everything else is black or white or fixable.
The best pilot in the world can leave for work this morning after a stonking argument with his family and go onto crash the aeroplane.

Human nature. And until pilots are removed from cockpits (which won't be long now) this will never be resolved.

[handysnaks]

Will pilots be removed from the cockpit, or will the cockpit be removed from the aircraft?

[Fareastdriver]

Then who do you blame if something goes wrong?

All other automated transport systems can stop, flying machines cannot.

[chopjock]

Ask the authorities to allow more use of simple helicopters, less to go wrong and less for the pilot to have to work out under stress.

eg, single engine and single fuel tank and allow auto pilot use on them.

That way pilot does not shut down the wrong engine, select the wrong fuel tank switch or have to see in the cloud because the auto pilot could take over.

Unfortunately the authorities over regulate and think they know better mandating over complicated machines in the name of safety.

[topendtorque]

Not wasting your time at all Nigel.

From your residence area in OZ you will have seen a repulsive wave of accidents these last few years in the Mustering Industry.

Not trying to predict anything but there is a determined push from some in Industry and CASA to address these very problems after the previous mentoring programs from the bigger companies all ran down.

I just pray that when I say, watch this space there will be some gaps filled in over the space of the next twelve months or so..

The OZ ag industry has a way to go as well. Jungle talk suggests a couple of years back that we over here had the same number of accidents as in the states but, (drum roll) only one tenth the number of aircraft as is used in the US

cheers tet

[Pittsextra]

I think there are a few things that could be improved on the human factors side and actually it wouldn’t take very much effort because the structures already exist.

The first thing that is required is greater resourcing at the AAIB so that they might be able to engage and report in both a timely manner and without having to rely on manufacturer resources. It surely can not be seen as desirable to have the situation, as we saw with the EC225, where we saw effectively an educated punt on the problem until aircraft No.2 ended in the sea. Nor can it be desirable to have the timescales we currently have to produce final reports. The EC135 accident in Glasgow reporting is very odd in so far that the process requires a seemingly additional 6 month delay in making a report public because of the input needed from interested parties.

We await to discover how or if human factors played a part in the Glasgow accident but actually if you do nothing but take PPrune as a window into the pilots of the EC135 years before the accident its fuel pump system was unclear to some. It would be amazing if that situation was oblivious to those engaged in the training and ratings for the type.

The AW139 accident that crashed in Norfolk is another one where red flags were waved consistently over the years, again on this forum. This time over the attitude of the owner/operator.

The findings of the accident in central London would suggest 3rd parties would not have been oblivious to the conditions the accident machine was being operated and failed to recognise or exercise control. Same as the S76 just off the top of trees in Kent, this time the co-pilot, yet despite being so terrified by his ordeal he recognised he become unable to speak “No entry was made in the aircraft’s technical log relating to the go-around and no air safety report or MOR was raised”….. The AAIB report for this event coming out over 2 years later!

Neither is there a co-ordinated, communicated or easily searchable set of military accident reports – which if we are looking to resolve human factors would be of great value. Such as the recent ETPS Gazelle tail strike event which looked at EOL’s or lessons from the Squirrel accident with a student pilot from RAF Shawbury. Even if these lessons are simply to identify that even the best of the military can make mistakes – and how quickly and easily things can go wrong - which allows the humble PPL to renew his focus.

All these things should be easy wins because its not as if pilots don’t like to talk or engage. The oil and gas industry have its own safety body, we have our own CAA, we have bodies/organisations dedicated to aviation from The Honourable Company of Air Pilots to the Royal Aero Society, LAA struts, hundreds of flying clubs topped off with LPC’s, type ratings, company training organisations and forums like this. It shouldn’t be hard, it just needs openness and a lack of ego.

[500e]

Pits
Totally agree but I think there are to many bodies & some with vested interests.
As you say
"The EC135 accident in Glasgow reporting is very odd in so far that the process requires a seemingly additional 6 month delay in making a report public because of the input needed from interested parties".

[Thomas coupling]

Chop jock - you haven't changed have you? :ugh:

Nigel has simply raised an age old issue which by definition cannot be engineered out of the system. All the authorities can do is legislate to within an inch of the industry's life and prosecute those who break the rules.

BUT,

automation is coming: AG flying will probably be the first. Already one can purchase driverless tractors and farm machinery. We have pilotless helicopters and of course pilotless aeroplanes have been around for decades.

Give it another 20yrs and you will see the first bulk purchase of pilotless cargo planes and many 'mundane' aerial tasks.

Perhaps human factors can be engineered out after all? :hmm::hmm:

[Pittsextra]

I guess this will be interesting on the same topic come November..

http://aerosociety.com/Events/Event-List/1954/Predicting-the-Fatal-Flaw-Can-We-Do-Things-Differently-in-Aviation-Safety

I know Jim Lyons is on this forum from the rotorcraft group so perhaps in between now and then there might be some input and therefore reflection upon recent helicopter accidents in this conference?

[ShyTorque]

Perhaps human factors can be engineered out after all? :hmm::hmm:Perhaps so, then of course nothing can go wrong, go wrong, go wrong....

Engineers and designers aren't human, after all. ;)

[Nigel Osborn]

Some interesting comments to my post. Today there are more medium to large helicopters around than 40 years ago & are generally crewed by two pilots & have fancy equipment fitted. So why do these helicopters so often crash due to pilot error? CRM has been around for a long time too, no longer can the captain tell the co-pilot to sit on his hands on approach. If either make a mistake, & we all do, the non flying pilot must speak up & point out the error.

In single engined helicopters the pilot may be should have done a CRM course too as then he can monitor his performance & quietly talk to himself about what is going on & the options available to get himself out of trouble if need be.

High time pilots can & do make mistakes but often due to their experience can realise they have made a mistake & have time to do something about it that the inexperienced may be too slow to notice.

Something must be wrong with the basic training & maybe more importantly their continuation training in later years. I once flew with a Bell 47 pilot who had 8000 hours of bush flying on type & one night he asked me what VRS was about because there had been a Bell 47 accident caused by VRS & he didn't understand the problem. In other words after getting his license years earlier, he had no more training.

So surely there must be more training in general with more emphasis on how to avoid accidents, like an advanced car driving course.

[John Eacott]

Nigel,

There are a couple of proactive approaches in the past ten years or so, here in Oz and globally. TEM (Threat and Error Management) is part and parcel of the training system now, and CRM is no longer Cockpit but Crew Resource Management, to account for those SP operations.

[Geoffersincornwall]

There may be some mileage in reading the thread I started about the selection and training of all kinds of flight instructors. It's running right now if you are interested.

I point this out for I feel strongly that the future of our industry may well depend on creating an instructor population that knows how to teach, has a lot of experience and is technically sound.

With the aid of modern simulators we can make a difference but only if we make every minute in the box count.

G.

[Nigel Osborn]

Hi John

I am too old to have done the TEM but I did CRM several times & found it worth it.

Although all these flash programs exist, they don't seem to have lessened the incidents of CFIT & other bad decisions. A good friend of mine ex Hobart, about 8000+ hours, very much a vmc pilot, couldn't handle night nor IMC flying, did his first twin conversion on a Bolkow 105 when about 70 years old in PNG. He flew 2 men to crew change in the mountains, the weather was bad so he landed in a paddock to wait for the weather to improve. It didn't so he decided to return to base. He climbed up several 1000 feet, entered cloud & at cruise speed he flew into a mountain that he knew was there. All 3 were killed. He had 3 options, stay the night in the paddock which was safe, not take the direct route over the mountain but fly around it or take the action he did & crash.

He had stacks of mountain experience in Tassie, Indonesia & PNG but never got of top of instrument flying. After leaving Tassie in 1989, he basically was not exposed to much check & training, just the barest minimum. That is my point, even experienced pilots need to be checked on a regular basis & really taught how to avoid accidents. Pretty flying is all very well, keep with 10 feet of altitude, 2 degrees of heading, 5 knots of air speed but don't worry about the deteriorating weather or last light or many other things.

[cpt]

.....coming back to unmanned transport flying machines, I believe that the human issue shall not solved at all .... there will be "pilots" or whatsoever on ground, as well as operations planners, maintenance engineers, systems controllers/software designers ....
Of course, we, "flying" pilots shall not be the last link in the chain anymore but I doubt greatly that the accident ratio would benefit a 94% gain.

It should be interesting to get a feed-back in regard with human factors from drones pilots within the armed forces.

[John Eacott]

Nigel, like all changes in procedures (or mindsets) there is usually a delay before effects are seen.

You and I come from a generally Australian flying background where checks have been (to put it bluntly) few and far between, against the UK for instance where a 6 month and 13 month type check is required to be signed off in your licence. CASA are changing the checking requirements, eg the HFR, but it is still generally driven by a company rather than the statutory authority in this part of the world.

Part 61 is an overkill and badly contrived, but the principal of a more regular and comprehensive check isn't a bad thing overall. The HFR (helicopter flight review) being every two years is insufficient, especially as I have to do one every 6 months because of my advancing years ;) At least annually is quite justifiable yet this was overlooked by the Part 61 changes even though low flying checks are now called for annually, despite having never been called for previously!

Anyone able to give an input on the FAA requirements?

[paco]

John - although the checking is frequent, the quality of TREs in Europe is fairly patchy. You will not get a standard ride from one place to another. Contrast that with Transport Canada where, no matter where you are, you will get the exact same standard ride. TC have their problems, but their standardisation is good. You only have to look at the threads on pprune to see how patchy basic training is - somebody puts up a question, and there are 48 different answers where there should really be only one.

From where I sit, we get a wide variety of students from all backgrounds, and I think that the standard of PPL training is critical. The FAA mandate 60 hours of ground school, which must be signed for in the log book and the FAA randomly inspect the schools. None of that happens in the UK - instead, the schools administer the exams! I think at the very least, if you study at one school, you should be examined at another, or by whichever Authority.

The other part of the pattern is attitude and self-esteem. For some reason, if a fork lift driver says that he cannot lift more than 50 packets, everyone believes him. Why don't they believe us? Because we are too eager to please! We, as an industry take on too many of other peoples' problems.

Phil

[n5296s]

Anyone able to give an input on the FAA requirements?
For Part 91 ops, close to non-existent, compared to the UK for example. For an aircraft under 12500 lbs, a BFR every two years, in something that flies. I can do my BFR in a hot-air balloon (well, *I* can't because I don't have a PPL-Balloon, but if I did I could) which would make me good to go in anything under 12500 lbs and non-jet I have a license for. (The SFAR 73 check for the R44 is considered a BFR, and makes me good to go in my 182 for example).

Insurers may set more stringent requirements - and do for turboprops for example, probably turbine helis also. And specifically for the Robbie an SFAR 73 ride is needed once per year.

I don't know what the rules are for Part 135 (fly for hire), not being a Part 135 pilot. I think it's established in the operator's Part 135 certificate.

[John Eacott]

Phil,

I think the issue here (as per the OP) is continuation training and checking rather than basic training of students/PPL(H) initial stuff.

Your point about TC standardisation is moot, since only the big companies follow such a practice. Or at least, used to follow unless the beancounters get involved?

[Gordy]

paco

Where does the FAA mandate 60 hours of ground school? Certainly not under part 61.

As for continued checking, under part 135 you have to undergo recurrent training and checking every 12 months.

[rotorspeed]

I think accident rates could be reduced if a greater proportion of initial and recurrent training involved dealing with issues that are actually causing most of the accidents. Someone with more spare time than me might do some research and summarise what these are, but training in poor visibility conditions, with aircraft near MTOW, high winds, difficult approaches, distracting cockpit environments, spurious distracting warnings, etc etc, would all be very useful in highlighting weaknesses for pilots and generating both more competence and more respect for the hazards of such things. I've long thought that too much training focus is often on dealing with mechanical emergencies - which very rarely occur. The vast majority of accidents involve perfectly serviceable aircraft.

Of course it's important (and not easy) to do this without adding significant risk getting involved in risky procedures, as with S/E EOLs. Just how competent the trainers are is also a significant issue - sometimes instructors have inadequate experience themselves of real world operational risks.

Pittsextra - you quite reasonably attach great importance to AAIB accident reports, but I am not sure how much help it would really be if they were produced in even half the time. Accuracy is critical. I think most of the time the basic causes of accidents are fairly obvious and indeed with regard to what can be learnt to avoid future acidents, often covered in a small range of hypotheses here on Pprune! If we consider a few examples, the Haughey 139 was almost certainly pilot disorentation manually flying in IMC, Kent S76similar, Sumburgh Super Puma probably dozy monitoring of an IMC approach, Glasgow 135 probably poor fuel managment, though less certain admittedly -I'm sure 135 pilots are more careful on transfer pump use now though.

Of course there are occasions where particularly say mechanical failure - and I think you're pretty hot on this, eg ? AS332 MGBX - are identifed and rapid industry awareness vital. There again, I suspect that even if the AAIB report isn't out, between the CAA and manufacturers preventative management is usually well underway.

Most importantly though as pilots we've got to take responsibility for our own actions - usually, but certainly not always, accidents are fundamentally our fault, not that of the "system". And let's not forget that the "system" ultimately relies on judgements of individuals - like pilots - and buck passing should be avoided. As helicopter pilots we enjoy greater freedom of making our own judgements, but with that comes more responsibility.

[paco]

We had a CFII here not long ago (Miguel) and he gave us that information.

I take your point, John, but they are linked. Continuation and checking is fine if you get the standards right, but they are not necesasrily here in UK. I have mostly worked for small companies in Canada, over a space of over 6 years, and from TC inspector to DFTE I've had more consistent check rides than anywhere. But then, that was some years ago :).

Before JAA, base checks (6 monthlies) used to be done at 90% MAUW, and I think it's a shame that it stopped.

Phil

[Pittsextra]

Here's some real data since 2010. Total of 44 reported accidents under "GA" with 23 PPL's, 5 student pilots, 14 CPL's and 2 ATPL's.

Of that lot there were 4 fatalities, which were 2 x CFIT (one a Gazelle driven by a PPL, the other a 109 driven by an ATPL). There was an R44 PPL inadvertent IMC and finally a new PPL mast bumping an R22.

Of the GA accidents there were 11 mechanical failures, 13 accidents through loss of yaw control/catch a skid/low speed mishadling, mostly in Robinson products and mainly very low currency.

So between the mechanicals, fatalities and low currency yaw control (which were hover taxi/landing/take off/parking events - I have been a little lazy lumping many of these together) that accounts for 28 of the 44 accidents. The balance were some flight instructor / student accidents, couple sling events and a couple of hitting bushes/tress whilst taxing.

In the same period there were 27 accidents reported under "Commercial Air Transport".

That also included 4 accidents with fatalities of which 3 (AW139 Norfolk, EC135 Glasgow and Sumburgh Puma) are yet to formally report and the other was the CFIT in central London.

There were 7 mechanical failures, 4 events of hitting objects on the ground a vortex ring heavy landing and a couple of loss of yaw controls.

You draw your own conclusions from that but personally I think it shows that there are actually more mechanical failures than anything else and the training to deal and recognise these failures seems to be effective as there were no fatalities as a result of them all across all licence holders and experiences.

It also shows that those most at risk of a CFIT are higher time pilots in more complex types and that if you take care of not hitting things on the ground and stay current such that you don't loose yaw authority you chances of rapping up a helicopter seem small!!

I guess the timings of AAIB reports seem important since any learning can be made sooner at the most simplistic level. For example if we believe the AW139 was a CFIT accident then it happened whilst waiting for the other report on the London accident. The formal report may have done nothing but it may also have been a timely reminder to owner and pilot..

If that is a stretch of the imagination too far then what about the odd example in 2012 where one company had two accidents inside a few weeks with slings - clearly the information flow wasn't happening internally so it was unlikely to be fed to other pilots doing similar tasks.

On the other end of things is the fact that whilst the events that end in fatalities certainly catch peoples attention regardless of the time scale between accident and report some of the other events may be lost in the mists of time.

[CRAZYBROADSWORD]

I have to admit I do watch a lot of air crash investigation and read all the reports and you do get this common theme of not one but a number of unlikely events all lining up to cause the accident , and not one thing on its own and as each event unfolds you are screaming at those involved to figure out what comes next but they never do .

As these accidents have happened and will happen again I am a firm believer in new technology being designed and put into practice to try and eliminate some of the issues that pilots face, the new trials of light weight mode s transponders is one such idea and we should always look positively at new tech! How many people poo pood iPads in the cockpit and now they are everywhere

TCAD , synthetic vision , IR cameras , night vision goggles etc should all be looked at and if they are useful made available along with the training to use them . That said if it moves it can crash skate boreds , pushbikes , cars, planes , helicopters the only way to stop them all having accidents is not to use them but remember horses bite !

[Pozidrive]

Pittsextra, your second-to-last paragraph refers to two similar incidents at the same company within a few weeks.


The AAIB investigate, which takes time, but isn't responsible for enforcement. In your example, would the CAA be checking the company's procedures as soon as possible, or would that happen after the AAIB reports?


I'm comparing this with the situation in other industries where the HSE can immediately issue improvement notices or prohibitions.

[timprice]

With all this regulation and Helicopters that are far more realiable, why is the accident rate so bad?

[Pittsextra]

Pozi - to answer your question honestly I don't know the full process of who talks to whom and when, but I'm sure someone on here would know.

However I used the accidents for those two sling load reports just as a convenient/ lazy way to demonstrate the need for timely AAIB reporting.

It was clear that the accident from G-ORKY just 8 days before made no impact on the pilot of the other machine (G-BXGA) in the same company, let alone the wider industry. One could i suppose fairly argue that he may not have felt anything needed changing with the huge experience the pilots had and so the specifics of these two accidents may have no lessons what-so-ever (and of course no one would expect a report inside two weeks). The point I was trying to make was that it is important that AAIB reports are timely because very often these are the only ways others can see the detail and its a cheap way to learn.

Actually with more time on my hands I looked at the AAIB reports for helicopters in GA all the way back to year 2000. 248 reports from 2000-date.

In the context of that the biggest accident group are students (66 total) and the bigger sub set of that are students with an instructor on board (39!). And with an eye to the instructor thread thats running of the 39 that shunted with an instructor aboard 26 had more than 1000hrs, 20 more than 2000hrs...

Then by far the next bigger group are those mis-handling events (61 reports) at low speed on the airfield (uncontrolled yaw, tail strikes, dynamic roll over, start-ups with the lever raised, hover taxi wobbles i.e. its a lazy catch all for all ball dropping at the airfield).

If you break that lot out then the biggest factor is flying a Robinson (35) which is ahead of currency, where 30 pilots had less than 200hrs TT, 20 pilots had less than 3 hours in the last 30 days.

Next biggest category is mechanical failure (43).

Without wishing to bash up robbos, they do feature alot and ironically of its a factor that could help to explain the improved GA accident rate since 2010, i.e. less robbo's. 126 reports of the 248 featured a Robinson.

There were 23 fatal GA helicopter events since 2000, 13 inadvertant IMC or weather related.

If you strip out the student/instructor and mechanical reports then of those doing it for themselves 48 had < 200hrs, 38 had >1000hrs of which 9 >10k hours. Interestingly the fatalities are clustered in the three bands, 6 circa 100hrs and less. Another 6 from 200-500hrs and 5 with experienced guys all IMC/CFIT.

[mickjoebill]

Here is a link to a list of helicopter accidents involving aerial filming that I collated. Year 2000 to 2013.


https://www.dropbox.com/s/remxlz52pjwokdo/Aerial%20filming%20accident%20database%20V7.xls?dl=0

The work to research these incidents was prompted by the realisation that aerial filming claims the lives and injures more Tv/film technecians than all other causes combined (excluding war zones)

A shocking record of failed industrial safety.
Since distributing the list the rate of fatalities declined, up until the recent mid air that claimed 5 media crew and 5 other souls :(

The rate of incidences has declined though, perhaps due to
1/ making crews more sware that their behaviour before and during the flight affects flight saftey
2/ increase use of drones in low budget shoots
3/ greater availability of nose mounted stabilised cameras
4/greater awareness by pilots of the hazards of "luvvie fever"


Mickjoebill

[Pittsextra]

On that Mickjoebill - to be fair this one looks like it was properly planned:-

http://www.aaib.gov.uk/cms_resources.cfm?file=/dft_avsafety_pdf_502261.pdf

Yet it went wrong. The AAIB report doesn't indicate why it went wrong or the process that led to the pilot not seeing the wires, yet he stated he specifically chose that part of the river inc. exemption from the CAA.

Is the recce for these things physical or via a google earth or similar??

[mickjoebill]

Is the recce for these things physical or via a google earth or similar??
A physical ground recce is rare due to cost. Combination of google pilots knowledge is the norm, followed by "lets have a look". So it can get pretty intense if a director has high expectations and a low budget.

One aspect of aerial filming that requires particular diligence to counter is a change of plan by tv/film crew in the following ways:
1/ subtle alteration of a flight path that contravenes previous agreements.
2/ request to change the shot during the shot
3/ ground activity altered without informing pilot
4/ heated, lenghty, irrelevant or misleading requests which pilot has to decipher whilst airborn.
5/ after a successful "take" the pilot succumbs to the request to go lower, closer or slower.
6/ tracking uncoordinated ground action whilst maintaining ability of cameraman to get the shot.

In respect to number 6 this is far more difficult if the camera is side mounted than nose mounted. Not surprising that the incident of accidents is higher for side door flights than nose mounted flights.
Sadley the instances of crew being killed when shooting out a side door are significantly higher than when using nose mounts, for obvious reasons.

A pilot recieves instructions by "non pilots" to position the helicopter to very specific positions often without due regard to the capabilities of the aircraft or the pilot, the weather or rules of the air!

Pilots in all experience brackets succomb.
Accident investigations rarely give the pressures and distractions by the film crew enough weight.

Since the media crew are casual or freelancers one relys on social media, conferences, and the like to educate TV folk.


Mickjoebill

[FLY 7]

Pitts, very interesting statistics.

[Nigel Osborn]

I wonder if this S76 C++ accident this week fits into the 94% of accidents which includes pilots not handling mechanical problems correctly.

Happily no fatalities this time but I hope the full report comes out quickly, so all can learn.

[Thomas coupling]

For as long as I have been in aviation and undoubtedly before that, the overwhelming cause of ALL air accidents - globally (betwen 65-75% dependent on the source)are down to Human Factor or 'pilot' as we are colloquially known :ugh:
These stats are stable at this level and have been for many years. Automation is the only way forward from this day on and as soon as we have pilotless aircraft, you will notice a sizeable change in the stats.
To ERR is human......................................:sad:

[Pozidrive]

Pittsextra, thanks for Post #30. Quite agree the benefits of timely reporting.


I have a particular interest in a couple of recent cases which have taken three or four years to come to court. So that's three or four years before the facts are in the "public domain" - facts as opposed to speculation and rumour. I suspect this is down to the lawyers, digging ever deeper, which all requires a response from the other side.

[Pozidrive]

Thomas coupling,


Automation is installed, programmed and maintained by humans - so the Human Factors and potential for error is still there.

[Thomas coupling]

Pozidrive - stop being obtuse. You know what I mean. Computers will of course have been programmed by humans (initially anyway) but these computers will not have to err the way humans do - en route.

[Pittsextra]

If we take AAIB reports since 2000 for both GA and commercial there were a total of 329 incidents.

However the 66 GA student training flights (there were 3 commercial "student" accidents) one assumes that they are flying for fun as a primary motivation so that's 20% of accidents automation is unlikely to capture.

There were 3 commercial "filming wire strikes" again unlikely to be captured by automation, likewise the 8 events where the machine ran out of fuel (excuse the crude term - inc. fuel mismanagement like (say) G-DPPH). There are 2 bird strikes, 3 where slings/sling loads hit the machine, 4 carb ice, 4 lightning strikes and 74 mechanical failures. So thats almost half of accidents since 2000 unlikely to be captured by automation.

[mickjoebill]

So thats almost half of accidents since 2000 unlikely to be captured by automation.

A note on Official reports from some who has read all of them which have involvied aerial filming.
Comparing accident reports to 1st hand accounts you would think that reading the official report that the media crew had no influence on the pilot.
We know they are capable of corrupting, distracting and confusing.

In developed countries there were only a few incidents that could/should have been reported that were not. (Incident defined for my list as damage to aircaft or camera or injuries to crew.

In a large number of incidents it took additional sources of information and research to join the dots as to the detailed purpose of the flight.
For instance a crash of a two seater with photograher on board is one thing but if it transpires the photographer was also shooting video and they were orbiting a subject and spun in then it is of relevance if one is to educate pilots of the issues in trying to achieve a smooth orbit of a subject on a breezy day.

A comparitivly large number of "close calls" go unreported and it is not PC to discuss in public thus depriving the industry of, in my view, at least a three times greater incident base from which to learn.


Mickjoebill

[500guy]

Here are some numbers for the US from a NTSB Database Search I did.


Leading Causes of Fatal Turbine Rotorcraft Accidents (US) 2004-2014




Fatal Accidents (totals past 10 years) (108)


Collision with Wires/Obstacles 26



Inadvertent IMC 15



Main Rotor Blades/System Failure 9



Spatial Disorientation 9



Mid Air Collision 7



Fuel Exhaustion 5



Engine Failure Mechanical (Single Engine) 5



Undetermined 5



Pilot Medical Incapacitation 3



Over Gross Weight CFIT 3



Tailrotor/Drive Failure 3



Dynamic Rollover 2



Undetermined (Not Engine Related) 2



Passenger Bumped Controls 2



Ground Personnel Contact with Rotors 2



Failed Attempt at Aerobatic Maneuvers 2



Ext. Load Contacted Tail rotor 1



Hydraulic Failure 1



Bird Strike 1



Engine Failure (Twin engine Helicopter) 1



Twin engine helicopter took off with one engine at idle (CFIT) 1



Controlled Flight into Water While Filming 1



Took off with Pedals Locked 1



Loss of Tail rotor Effectiveness 1

[CRAZYBROADSWORD]

Would like to see stats from similar industries as a comparison helicopter flying will never be without risk but I still feel more nervous in a car