ATSB failure in Benalla investigation
 

Many will remember that after nearly 18 months of investigation, the ATSB came out with its report on the Benalla fatal accident with a press release which said, “Unfortunately, despite these prolonged efforts, the reason for the tracking error could not be determined.”

What a joke. Evidence has been provided to the Coronial Inquest in Victoria that if the aerial is simply disconnected or has a fault, the Trimble GPS will fly the exact route of the accident aircraft. How could the ATSB get it so wrong?

More to the point, how could they write a final report after such and extensive time and not mention the “elephant in the room.” That is, the lack of the use of radar in uncontrolled airspace in Australia to prevent Controlled Flight Into Terrain accidents.

It is interesting that the Counsel assisting the Coroner, Mr John Langmead (a pilot who had told me that he had done previous work representing the ATSB), did everything he could to protect the ATSB from any investigation. Here is a quote from his opening address.

Obviously not the “experts” who wrote the ATSB report which completely missed the most fundamental issue of all. That is, as stated above, the Trimble unit is so flawed in its design it can go into Dead Reckoning and completely replicate an instrument approach many miles away from where it actually is.

There are others at the Inquest who are going to make sure that any matters in relation to the air traffic service – i.e. the airspace and the use of radar – are not really looked into. Note the following quote from Ian Harvey, the learned counsel representing Airservices Australia at the inquest.

Of course, Mr Harvey’s proposal completely ignores the fact that the ATSB report is totally wrong, and therefore would certainly require some “adumbration” (whatever that means) or “elaboration”. The Airservices internal report would no doubt have not addressed in any way the lack of Minimum Safe Altitude Warning (MSAW) system in radar covered uncontrolled airspace in Australia.

The counsel for the family, Mr Jay Anderson, proposed that I be called to give some of the evidence that others so much wanted suppressed. Interestingly, the counsel assisting the Coroner, John Langmead, was violently opposed to me appearing. In a 10 minute “rant” to the Coroner, he came up with every possible legal technicality as to the reason I should not be allowed to appear.

The Coroner took Mr Langmead’s advice, and a copy of the Coroner’s decision is on my website here, together with my statement that was prepared and paid for by the Henderson family.

I think the whole issue is very sad for the families concerned, and it is telling that the ATSB has not issued an immediate and urgent warning to all pilots on how their report was completely wrong, and that pilots should be aware that the Trimble unit may go into Dead Reckoning mode – and without manual inputs can fly a complete approach in a different location. I hereby warn pilots of this issue. My suggestion is to throw out your Trimble units and get a decent Garmin today.

A NDB at the airfield would have been a good defence to the tracking anomaly.

Yes, but that would not have helped if the aircraft was on track but descended too early - the most common form of CFIT accident.

Why not use radar when it is available?

Perhaps, if the Trimble has such a fundamental flaw, it is CASA or the manufacturer that should be queried. Didn't they approve it for use? Surely the manufacturer knows about this defect?

Jeppesen has recently announced they will not be providing database updates for some Trimble models after Mar 2009.

AOPA Online: Jeppesen to end updates for older IFR GPS receivers

AsA cannot man the current airspace, let alone class E to 700ft everywhere. :rolleyes:

Dick, my understanding of the ATSB role is that they determine what happened from the facts they are presented with. They could not determine that the aerial had become disconnected or was faulty. That would be supposition.

Nor is it their role to decide what should have happened. Yes, they can make recommendations to prevent future crashes but only based on certain facts which in this case, were not certain.

I can understand your frustration with the present system but to criticise the ATSB for not presenting a report that would be based on probability is not going to help.

I have just heard from a retired US air safety expert and air traffic controller. He said in a situation such as Benalla, and

I wonder why the ATSB didn’t recommend that Airservices look at such a proposal here? Yes, I’ve heard all the claims about cost etc, but surely to do it at one airport (say, Proserpine, with large jet movements) the cost would not be prohibitive and the safety improvement would be important.

The nice thing about coronial inquests (I've been an expert witness at several) is that they usually look at aviation accidents with a different mind and different opinions to those of us who are involved in aviation in our daily lives.

There seems to be a view on these pages that the ATSB is beyond reproach. I agree that our ATSB investigators are some of the best in the world, but they can, and do get it wrong from time to time. It is important therefore, that the ATSB, and in this case Airservices and CASA, are subjected to external scrutiny because this is how we can learn and develop even safer systems.

I am surprised at the extent to which some parties are going to try to limit the evidence presented to the coroner. Wouldn't it be better if all the potential contributors and mitigators where considered?

After the Toowoomba C90 and Lockhart Metro Coronial findings, for justice to be done - and seen to be done - one would have thought it appropriate for the Coroner to avail himself of all possible evidence and submissions, from all parties.

After all, the Coroner has the discretionary power to accept, ignore or reject any evidence he believes irrelevent or incorrect.

When certain names consistently appear in every Coroners Inquest in respect to aviation matters, one can't help but feel the possibility of manipulation could occur.

I'm not a man of huge means, however I would consider contributing funds to some kind of legal challenge if that is at all possible.

The Coroner has however made it clear that your testimony does not fit with the initial purposes of a Coronial inquest, however it should be included in testimony in order for the coroner to make recommendations for the future so this doesn't happen again - without your testimony, this can't happen.

As a minimum Dick if you haven't already, you should be taking this directly to the media with a statement near identical to what you have posted here. If the public knew that this was a known Trimble 'condition,' then the fact that this was known should not be ignored. If you believe that you can directly blame the deaths of 6 people on equipment that is fitted to other aircraft that could do the same thing, then people will surely listen and the media should definitely work with it. Go for it! It might take the heat off Qantas and put some on CASA and the ATSB for change!

For DR? needs a GS and Start Position to be entered manually. Who in their right mind would trust a device AFTER entering that information? DR is a DEMO mode in the Trimble.

Better yet, how much information is available to a pilot off an expired datacard? What functions are locked in a Trimble with an expired Datacard? Can a pilot activate APR with an expired Datacard?

In a shocking case of irony, the aircraft was one week away from a major avionics upgrade.

OZ, You are totally wrong! The Trimble manual may say that it needs to have manual entry but in practice the unit will keep "flying" the last track and groundspeed.

Wait another couple of years and CASA is sure to advise pilots of this!

Wrong Dick! I think you best ask someone who actual runs one. POH doesn't support your theory.

.... and the Lockhart R aircraft was scheduled to have terrain warning installed!

A $1600 Garmin 296 GPSMap with the terrain warning activated may well have prevented both the Lockhart R and the Benalla crashes!

Kinda makes you think doesn't it!

Dr :8

If the Trimble really has such a major flaw/defect, would the FAA be sitting on their hands?

Would it really be left to the A - Australian - TSB or CASA to tell the world?

OZEBUSDRIVER, You are amazing, do you work for the ATSB?

I can assure you the documents are wrong and thats why the ATSB got it wrong - they believed the documents without the simplest checking.

CaptainMidnight-Yes!

Dick

Maybe you are corret, however assuming that the GPS was fine, can APPROACH MODE actually be selected with an out of date data card??

This is a question not answered.

If not, perhaps a pilot entered approach was being used and it was wrong. Finger trouble is not new.

J

Garmin 430 - yes!
Trimble - don't know!

Dr :8

Fair enough..... but OZBUS seems to have info that the Trimble will not.

An IFR GPS that does not play by its manual is a bit strange, so what does it actually do?

Seems nothing much said in the ATSB report this yet Dick is convinced it does not work as advertised. Something fishy here!

J

I had the distinct pleasure of knowing 'K' the PIC on that faitful day thru a few years of flying in & out of BLA and simply chatting to him over a cuppa in town whilst he waited as we all did on charters back then. I have flown into BLA numerous times visually & via the NPA in PA42's & B200's & I'm 100% sure that there had to be a unique error in the Trimble GPS unit (aerial or no aerial problem) that wasn't picked up or noticed by 'K' during that particluar approach. The only thing that may have distracted 'K' on the day was his pax on board, a couple of them would have been very keen to observe the goings on especially during the approach. I'd like to think that it was an insidious error that probably none of us pilots would have picked up that ended "K's life & the unfortunate pax on that day. Trimble or no Trimble I rely on any GPS with the absolute minimum of trust. If it's a stand alone approach as in no other ground based aid at the airfield (as is the case at BLA) I add a bit for 'mum & the kids' whilst outside the circling area, sure that's strictly not correct but that's just me. Most GPS's of these types (Trimbles & KLN's etc) have poor situational awareness as in small screens in monochrome colour, this just adds to the workload single pilot.
We can argue this sad story 'till the cows come home, we lost a true airman that day, none finer!


CW

Dick,

Its been some time since using a trimble, but the TSO trimble still has a RAIM function, still sends out warning messages.

A very good number of GPS units operate in DR mode when nothing better is available, but appropriate warnings are given to the user.

Dick, I just want to hear from Drivers with experience with this device. There is nothing sinister in that. I haven't flown with one and I am only going by a POH. At least it is better to get information here than sensationalise it in the press.

If such an alert was to be issued by the controller, would the pilot have to initiate a missed approach on the basis of that "descending below MDA" alert, or acknowledge the alert, checking that he was tracking the approach as per GPS indications, and continue happily into the rising terrain (according to Dick's "Trimble in DR mode - correct approach but wrong location" hypothesis)?

I'm with Dick on this one.

Surely the most important aspect of this accident for the ATSB to be looking at is what caused the GPS to generate consistent (but wrong) positions.

The GPS entering some sort of "special mode" would have to have been a focus of theirs. That "special mode", as Dick points out, could well be the "dead reckoning" mode.

In an enroute, high altitude scenario, such a mode could be a somewhat useful thing (for short periods) should the unit be unable to track sufficient satellites for proper position fixing, so one can see why the unit would possess such a feature.

And one way to deprive it of satellites is to disconnect the antenna - or allow the antenna to become disconnected via a fault.

Highly unlikely that the unit would not have given some warning that it has entered DR mode. Was the warning sufficient to alert the pilot? Did the pilot see the warning but not understand it?

Irrespective of what the manual says, the ATSB should have tested whether inadvertent "mode reversion" was a factor in this accident. One cannot expect the manual to explain all possible behaviours in all possible scenarios.

Have to agree with Dick about the Radar warnings too. Note that Dick is not saying that Radar must be installed everywhere - he is simply saying that if it can be used to save lives, why isn't it?

Will a TSO C129 GPS Navigator go into Approach Mode with an out of date datacard? Yes it will, it is a feature of the TSO based around a philosophy that it would be better in an emergency to be in the situation of having to fly an approach with an out of data card than to have an otherwise good navigation system that was unusable.

Does the Trimble 2101 go into DR mode when it loses signal? From memory yes, but it's been a while since I used one.

The T2101 was one of the first TSO C129 navigators available. It's pertinent to note there were a number of major software changes during the early deployment. Notable amongst several is that on some versions the RAIM annunciator would be illuminated when RAIM was available and on other versions the RAIM annunciator illuminated when RAIM was not available.

"Have to agree with Dick about the Radar warnings too. Note that Dick is not saying that Radar must be installed everywhere - he is simply saying that if it can be used to save lives, why isn't it?"

Here are a few reasons that make that proposition questionable:

• Presumably aircraft is on CTAF frequency ... how does ATC talk to him?
• Aircraft is in uncontrolled airspace and could be manouvering for other (unknown to ATC) safety reasons ... eg. traffic, turbulence

Basically, unless the aircraft is in an airspace, over which the ATC has a control and communication responsibility ... then the proposition is fraught with danger, fuelled by ambiguous responsibilities.

Solutions:

• If there's radar coverage, make it a minimum of Class E Airspace ... ATC staff numbers ?????
• OR, put in another class of staff to monitor and communicate with aircraft in Class G airspace in radar coverage... it ain't gunna happen.

Some interesting reading on the history of Class G, TAAATS, ATC, DTI, and key players.

http://www.aph.gov.au/senate/committ...ort/report.pdf

Australian Parliamentary Library - Background Paper 10 1997-98

Both the ATSB and CASA have read the very thorough report by professional pilot John Chew, which was presented to the Coroner last week. This report shows beyond reasonable doubt that the Trimble unit can go into dead reckoning and fly the complete approach in the dead reckoning mode.

Of course, there are normally warnings – but just as a fault can put the unit into dead reckoning, a fault could remove some or all of the warning messages.

Other GPS units, such as the Garmin, simply freeze when there is a loss of signal in the air – i.e. they do not go into dead reckoning.

Why are both the ATSB and CASA hiding this important finding from the pilot community? Is it once again someone trying to protect the obvious terrible failure of the ATSB final report? After that report came out, the Deputy Director of the ATSB, Alan Stray, said words to the effect, “We will never know what caused the aircraft to get off course.”

The ATSB also failed to do a proper investigation - i.e. simply fly the route and disconnect the antenna to see if, in the prevailing winds, the unit will fly a totally false approach -which it clearly does.

I will today write to the Minister and ask him to see what he can do about getting CASA and the ATSB to announce to the industry the important finding by John Chew. I have a copy of John Chew’s statement, however I don’t believe I can post it on PPRuNe. It is up to the ATSB and CASA to release this statement – they have it as well. We may then prevent another crash in the next few days. Who knows?

Jabawocky, yes, in a Trimble unit the approach mode can be selected with an out of date data card. The only message (after the unit is turned on) is “data card out of date” or something similar.

Peuce, you state:

This is a factual statement but your mind seems to be completely closed as to better ways of using our radar. For example, in the UK, air traffic control actually “control” IFR aircraft in Class G. The Plymouth tower manned by air traffic controllers sits in Class G airspace. Those who want an IFR service are actually “controlled” by air traffic control – either in the Centre or in the tower.

Now I know this must be almost beyond comprehension to you, because ingrained in the view of Australian controllers and pilots is that in uncontrolled airspace, a controller must never issue an instruction – unless possibly it is an emergency.

It just so happens that the British regulations (although not ICAO compliant) actually lead to greater safety. We are a sovereign country and we can do the same thing ourselves – even though I would not recommend it. I would put in Class E airspace.

How can Class E at, say, Benalla, increase the number of ATC staff required? You would be lucky to get one IFR aircraft at Benalla in a day, so there would be no separation required. You may be amazed, but in Class E airspace the pilot remains on the Class E frequency until visual. This is because no one else can be in the cloud at the same time without a clearance. This maximises the use of radar.

Fortunately there are many more young Aussie controllers coming along with a broader view, and they are asking sensible questions, such as, “If other countries can provide a control service for IFR aircraft where they have radar, why can’t we in Australia?” The answer is that indeed we can, we just need to have open minds and be prepared to copy the success of others.

Regarding the number of controllers, that piece of airspace is divided into two when necessary, ie, when traffic warrants. Where Benalla is located is in the HUM/OVN/DOK sectors. Those sectors provided arrivals/departures/sequencing for Melbourne, Canberra and Albury. They also look after the often busy Wagga Wagga. DOK sector has Mangalore in it, which is now becoming a very busy training establishment with many IFR movements per day. The following aerodromes are located in that airspace and have greater than or equal to the IFR traffic on Benalla, with published instrument approaches (off the top of my head):

Mangalore, Mansfield, Corryong, Albury, Shepparton, Benalla, Wangaratta, Corowa, Wagga Wagga, Tumut, Yarrawonga, Mount Hotham, Mount Beauty and Canberra (at night).

The range set on the screen to cover these sectors is around 400NM, which has to take in Canberra and Melbourne.

Given the staffing at the moment, they are often combined onto the one position which also takes the SNO/YAS/JVS sectors, with the following IFR capable aerodromes:

Latrobe Valley, West Sale, East Sale, Great Lakes, Bairnsdale, Orbost, Malacoota, Merimbula, Moruya, Nowra, Cooma, Polo Flat, Goulburn and Cootamundra.

The range they would be operating takes in Sydney to Melbourne. In addition to providing class A, C, E and G services, they separate crossing trans-Tasman traffic with north and southbound traffic to/from Tasmania, liaise with Esat Sale and Nowra for clearances in restricted areas, and are responsible for IFR movements below A090 into and out of Sydney to the south and east.

In all, there are about 100 separate instrument approaches in that airspace. That's a lot to keep tabs on, let alone monitor to (almost) the ground. Most of those aerodromes don't have radar coverage to the deck.

I think we would need more controllers.

Cheers,

NFR.

Geee.....A good(thorough) low level ADSB roll out and a few more controllers and Dick could have his Class E all they way:ok:

J:E

I would like to get hold of a copy of John's statement. I personally know him. I know he would accurately report on all possibilities, However, I would want to read his words!

I am pretty sure John would state that for any failure, it would also produce a warning message. If it were possible to revert to DR there are warnings for this( prompt for GS and position). If the aerial failed, there is a specific warning for this, If RAIM was lost there is a specific warning for each mode.
If it were possible to revert to the DR DEMO mode, according to the POH all features are replicated except for APR mode to be armed. Going by this scenario you would have to have the aerial failure and loss of RAIM without getting a warning message at any time? On top of the non standard flight and non standard route request, That is a lot of holes to line up!

For the message light to fail? How many more fail modes are we going to add to get the scenario we want? The more failures, whilst possible, increase in improbability. The same scenario also works with the wrong approach selected with no warnings and no failures. It is also quite possible that the pilot may not have armed the GPS with this approach. It could have just as easily been input by the person sitting next to the pilot whilst under supervision. What if, what if, what if, Even for John, he would only put down what MIGHT have happened, what was possible. Nobody will ever know what did happen.

The one question that must be answered no matter what scenario is envisaged- Why did the pilot descend below the FAF height of 2295ft without the GPS arming APR 2nm before the FAF? No mode supports arming APR unless 2nm from FAF and RAIM availability. No Aerial No RAIM NO APR mode. If no warning device why did pilot descend below the FAF.

Dick, your accusations, whilst trying to absolve the pilot, creates more and more problems that still must be answered.

DICK,

I don't care what the airspace is designated ... G, K, T or Z .... if you want Controllers to control aircraft in it, then you need:

+ Controllers
+ Clear responsibilities
+ Appropriate Facilities

The facts are the facts ....

As an aside, if "...we are lucky to get one IFR into Benalla a day ... ", then by making it controlled airspace, are we really allocating the resources where they will do the most good ?:confused:

Ozbusdriver, in no way am I trying to absolve the pilot of anything. I’m simply stating that the ATSB got it wrong. That is, they stated clearly in their final report that the only way the GPS would fly on dead reckoning in that situation would be to enter the information manually. This is wrong.

In practice, the GPS will fly the route if it goes into dead reckoning for any reason – and in fact fly that exact route if the wind is in the right direction. No proper certified GPS should be able to do such a thing. The Trimble GPS should follow the flight manual details – i.e. require manual inputs.

Surely if a trained 747 pilot can take off at Tenerife without a clearance, any human being can do anything. We are not infallible.We can miss warning messages or they can fail.

Have a look on my website here. I have posted John Chew’s statement.

Peuce, I agree regarding controllers’ clear responsibilities and appropriate facilities – I always have. However you seem to have an incredible misunderstanding. You state:

Most certainly, because we may be able to stop a controlled flight into terrain accident. The last one killed 6 people there. You seem to think that IFR services are only related to collision protection between aircraft. In fact, there would be at least 5 times as many collisions with the ground compared to collisions between aircraft. If we can use air traffic control and radar to help prevent this, we should.

My understanding, open to correction, is that radar does NOT go to ground level at BLA.

A TSO 146 NAV seems a more appropriate solution for aircraft in Class G?

Dick,
If the UK controllers are 'controlling' aircraft in class g, then it is not class G, it is a heath-robinson construction.

The only way control is exerted is by issue of clearances. Acharacteristic of class G is that clearances are not issued.

Perhaps you meant to say something else, perhaps that controllers issued advisories or somesuch? In any case it is not standard ICAO, or, dare I say it, standard FAA.

For over 15 years stubborn people have been claiming that because radar does not go to ground level everywhere, we cannot then use it anywhere in an enroute environment to reduce controlled flight into terrain accidents.

Nothing could be further from the truth. In the case of the Benalla accident, the aircraft was below the lowest safe altitude when 25 miles away from the aerodrome. If the MSAW alarm had been activated and the pilot was told, there is a good chance everyone would be alive today.

Others have constantly claimed that better equipment in the aircraft is “a more appropriate solution for aircraft in Class G” rather than understanding the obvious. That is, a good TSO 146 nav, plus the extra added protection of a radar MSAW will be even safer.

Around the world, pilots have had the best navigation equipment possible, but with one error of descending slightly early, everyone dies. Look at the Flight Safety International documentation of CFIT accidents. They make it quite clear that radar and air traffic control is one of the best mitigators.

40years, no I didn’t mean to say something else. As I said, in the UK there are large amounts of uncontrolled Class G airspace, where the controllers will actually “control” IFR aircraft.

My suggestion is that you ring the head of the Office of Airspace Regulation at CASA, Mr Peter Cromarty (on 02 6217 1408) and ask him. He will verify that even though it is not ICAO compliant, it is the British way of handling IFR aircraft in such airspace.

Dick

I have followed the Benalla crash from day one with some interest, as it happened the day after I escaped from Tumut after being stuck there for 3 days in atrocious weather.

I cannot explain to myself how, or why it happened! Likewise I was surprised that ATC didn't say something because in my part of the world (NQ) they have on a number of occassions when I have done something they weren't expecting - like have an AP get a dose of the wanders.

None-the-less, are the circumstances of this accident now a little historical?

Those of us who fly in areas well away from radar coverage need to put systems in place to maximise our safety. My answer to that in my Bonanza is to have T129 G430 with terrain warning - which I have recently upgraded to TSO145/146a, which has the terrain warning incorporated anyway. In addition, I carry a Garmin 496 which also has terrain warning.

The cost of equiping an aircraft with these systems is relatively small when taken in the context of the cost of acquiring a suitable aircraft and operating in the IFR environment.

When the Benalla accident occurred, my Bonanza didn't have a TSO'd GPS - just a Garmin III Pilot on the dash, but it is now equipped as above.

As I have indicated in here previously I flew the LHR approach in IMC the day after the Metro accident, and if fact had a conversation with the rescue chopper that was hovering over the accident site at the time. At that time my 30 yr old Bonanza was fitted with the terrain warning system mod to its G430. Unfortunately the aircraft flying RPT was not so well equipped.

I think a bigger danger to my health these days is a mid-air collision - vividly illustrated to me recently.

I think the biggest enhancement of aviation safety in this country would be the availability of vertical navigation at remote locations - ie WAAS, and affordable enhanced traffic avoidance. I fitted a PCAS system to the Bo after my recent close call, but the jury is still out on how effective that is.

Dr :8

By the way, in the USA ATC's will separate IFR aircraft when in radar covered class G. It's only in Australia that ATC's are not allowed to do this and are specifically trained just to give a traffic information service.