AS332L2 Ditching off Shetland: 23rd August 2013
). Added to the analogue nature of the inputs and/or outputs (TBH, I can't remember which it is) and you have an autopilot that is less than ideal for some of the tasks that it is used for.
Join Date: Apr 2009
Location: Vancouver
Posts: 3
Likes: 0
Received 0 Likes
on
0 Posts
Prioritize airspeed
The many crashes I see in the 76 sim are nearly always preceded by a loss of airspeed, whether or not an autopilot is used or installed. Certainly in the 76 if you look after the airspeed everything else will follow. If you don't look after the airspeed everything will fall apart. This applies in manual flight or in autoflight. Hence, if you are climbing, descending or on approach I firmly believe that you should have the autopilot look after airspeed first. Full coupling on approach is better, but if you can only engage one vertical mode, make it airspeed.
I can produce a crash 100% of the time in the sim when 2 cue by having the students set the power too low in a climb with VS engaged, or too low when approaching an ALTPRE level off.
I wonder whether autopilots tend to reduce our attentiveness and monitoring of flight instruments.
I can produce a crash 100% of the time in the sim when 2 cue by having the students set the power too low in a climb with VS engaged, or too low when approaching an ALTPRE level off.
I wonder whether autopilots tend to reduce our attentiveness and monitoring of flight instruments.
The reason for the L2 autopilot 'leading with the cyclic/attitude' was explained to me because it isn't a true 4-axis autopilot (although it is sold as such). It is in effect a '3 plus one' autopilot, ie a 3-axis autopilot with the 4th axis tagged on. This is why the L2 will always use attitude first to control or achieve speed or height, followed by power. (eg when the GA button is used ).
Added to the analogue nature of the inputs and/or outputs (TBH, I can't remember which it is) and you have an autopilot that is less than ideal for some of the tasks that it is used for.
). Added to the analogue nature of the inputs and/or outputs (TBH, I can't remember which it is) and you have an autopilot that is less than ideal for some of the tasks that it is used for.
PA,
I would have liked for the AAIB to include the CVR data....a couple more days would not make a difference one way or the other as to the future of the SP on the North Sea. It would have made a huge difference in the ability of the Operators and others to provide a better explanation of what happened and why. It could still be well short of a full report but the inclusion of the CVR data would remove a lot of speculation yet to come.
One thing I noticed in the Sim was when things started going wrong....one of the first reactions of a lot of Pilots was to de-select the Autopilot and revert to hand flying.
Part of that stemmed from the newness of both Sim Training and Autopilots. I would think as Automation and one's experience/familiarity with modern Autopilots this tendency would decrease.
The single most glaring mistake I saw was the failure to increase power upon executing a Missed Approach.....and when that happened....we saw a slow motion crash. The Crew would decide to go Missed....hit the G/A button....make their Radio Call....start looking for charts, maps, changing radio frequencies or whatever.....and the Autopilot would do its best to comply but at some point the climb rate could not be produced by the decrease in Airspeed and then the loss of control happened.
Generally, Unusual Attitude Recovery training is done with lots of airspeed and at higher altitudes. Low Airspeed events close to the ground are where helicopters are the most at risk.
What is the system in the UK for such training and testing on Base Checks?
Do you practice low or zero airspeed upsets?
I would have liked for the AAIB to include the CVR data....a couple more days would not make a difference one way or the other as to the future of the SP on the North Sea. It would have made a huge difference in the ability of the Operators and others to provide a better explanation of what happened and why. It could still be well short of a full report but the inclusion of the CVR data would remove a lot of speculation yet to come.
One thing I noticed in the Sim was when things started going wrong....one of the first reactions of a lot of Pilots was to de-select the Autopilot and revert to hand flying.
Part of that stemmed from the newness of both Sim Training and Autopilots. I would think as Automation and one's experience/familiarity with modern Autopilots this tendency would decrease.
The single most glaring mistake I saw was the failure to increase power upon executing a Missed Approach.....and when that happened....we saw a slow motion crash. The Crew would decide to go Missed....hit the G/A button....make their Radio Call....start looking for charts, maps, changing radio frequencies or whatever.....and the Autopilot would do its best to comply but at some point the climb rate could not be produced by the decrease in Airspeed and then the loss of control happened.
Generally, Unusual Attitude Recovery training is done with lots of airspeed and at higher altitudes. Low Airspeed events close to the ground are where helicopters are the most at risk.
What is the system in the UK for such training and testing on Base Checks?
Do you practice low or zero airspeed upsets?
Last edited by SASless; 5th Sep 2013 at 20:12.
Yes unfortunately I think this is a worry. The better the autopilot, the more tempting it is to stop monitoring. But what is the answer? Get rid of autopilots and make all flying manual? (back to the 1970s) or have a bit of a new paradigm in training where there is much more attention placed on monitoring skills? I'm not sure I have ever received any training in monitoring (or, to be honest, given it) but I have certainly "given feedback" when monitoring has been found to be lacking, a relatively common occurrence. Its a skill that we expect everyone to have innately, but maybe that is expecting too much?
Join Date: Apr 2009
Location: Vancouver
Posts: 3
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by SASless
One thing I noticed in the Sim was when things started going wrong....one of the first reactions of a lot of Pilots was to de-select the Autopilot and revert to hand flying.
Join Date: Jul 2010
Location: Aberdeen
Age: 43
Posts: 46
Likes: 0
Received 0 Likes
on
0 Posts
Could the guys up front maybe have been ex 225 drivers and gone back to the L2 since the 225 was grounded perhaps thinking that the protection systems available on the 225 would kick in?
Join Date: Apr 2009
Location: Vancouver
Posts: 3
Likes: 0
Received 0 Likes
on
0 Posts
I agree with HC and SASLESS about sim training. Maybe a bit less plodding through procedural IFR and simple engine failures, and more recreation of circumstances that lead to loss of control or CFIT. I have been in simulators with some instructors who seem to want to avoid letting the crew crash at all costs. There's nothing like a spectacular crash in the sim to drive a lesson home.
Even in 4-axis, the L2 autopilot isn't very good. When using ALT-A in the climb or descent (such as during a NPA), then it often flies through the required altitude before reacquiring. And it's control of the 4th axis always has the potential to overtorque the engines.
So hours on type may not be the crucial crew data but hours on other type. That would make some sense in relation to the auto systems tutorials of the last few pages.
HC
I have made the point before that more and more pilots seem to be becoming system operators rather that pilots. If you were confident in your ability to fly the a/c you wouldn't need "monitoring skills" the auto pilot would just be a tool you use to reduce the workload when it was required.
Although the 365N2 autopilot was simple it would do a coupled ILS. As I spent most of my time offshore the only time I could practice my IFR skills was bringing the a/c ashore or doing rig radars while holding for a beach flight to pass through. I therefore flew manual ILSs when I could. If the weather was on limits then I would use the coupled ILS to reduce my work load.
The emphasis nowadays seems to be use the autopilot at every opportunity which seems to be reducing the flying skills of the crews - hence we seem to have pilots - both fixed wing and rotary who allow the autopilot to take control in unusual situations with tragic results.
I still can't understand a training regime which insists that the autopilot is engaged at Vy and, I believe, has a monitoring system which asks pilots why they didn't should they try and hand fly. No wonder pilots become disconnected between actually flying and operating a computer game!!
In my day you didn't have to have a "monitoring" skill you flew the a/c and if the coupler was engaged you made sure the instruments were indicating what you would expect if you were hand flying. It helped that the coupler wasn't perfect and sometimes needed a nudge with the flying controls to do what you wanted.
HF
I'm not sure I have ever received any training in monitoring (or, to be honest, given it) but I have certainly "given feedback" when monitoring has been found to be lacking, a relatively common occurrence. Its a skill that we expect everyone to have innately, but maybe that is expecting too much?
Although the 365N2 autopilot was simple it would do a coupled ILS. As I spent most of my time offshore the only time I could practice my IFR skills was bringing the a/c ashore or doing rig radars while holding for a beach flight to pass through. I therefore flew manual ILSs when I could. If the weather was on limits then I would use the coupled ILS to reduce my work load.
The emphasis nowadays seems to be use the autopilot at every opportunity which seems to be reducing the flying skills of the crews - hence we seem to have pilots - both fixed wing and rotary who allow the autopilot to take control in unusual situations with tragic results.
I still can't understand a training regime which insists that the autopilot is engaged at Vy and, I believe, has a monitoring system which asks pilots why they didn't should they try and hand fly. No wonder pilots become disconnected between actually flying and operating a computer game!!
In my day you didn't have to have a "monitoring" skill you flew the a/c and if the coupler was engaged you made sure the instruments were indicating what you would expect if you were hand flying. It helped that the coupler wasn't perfect and sometimes needed a nudge with the flying controls to do what you wanted.
HF
Bravo, what you say is true but descending then levelling at say 110 kts isn't going to cause an overtorque. Yes it does dip below the ALT.A height, but the solution is just to set 50' above MDA at the "top of the drop" (or more, if high vertical speed) and then beep the alt down a bit as its levelling. So whilst you are correct in that the L2 autopilot is far from perfect, I maintain its quite capable of flying a successful NPA in 4axis. I should know, as a training capt doing endless OPCs in it (especially before we started to use Helisim) I've done (or watched from the other seat) enough!
HF I think your last para is very true. As I said before, the better the autopilot the less perceived need there is to monitor. But are we saying therefore that manufacturers should intentionally build in software bugs just to keep us on our toes? Don't think so!
Personally I was quite happy to use the 225 autopilot to its full capability because I am intrinsically lazy. My "old fart" colleagues don't seem to have lost the ability to fly, although some still struggle to understand the technology. The newbies of course lap up the technology, but I see a determination in them to also master manual flying and most ask to fly manually quite regularly. So, whilst a balance of both these disparate skills is required, I think we are pretty much doing that already.
Personally I was quite happy to use the 225 autopilot to its full capability because I am intrinsically lazy. My "old fart" colleagues don't seem to have lost the ability to fly, although some still struggle to understand the technology. The newbies of course lap up the technology, but I see a determination in them to also master manual flying and most ask to fly manually quite regularly. So, whilst a balance of both these disparate skills is required, I think we are pretty much doing that already.
Last edited by HeliComparator; 5th Sep 2013 at 21:12.
Join Date: Aug 2013
Location: Northern Lights
Posts: 48
Likes: 0
Received 0 Likes
on
0 Posts
HF
I don't know what regime you are describing in your post.
On the only 4 axis machine I have flown, the company -- and the other pilot -- would insist on 4 axis at night or in Poor IMC but I've regularly done manual climb outs to altitude, manual ILS in a variety of weather -- allbeit usually pretty good
-- and, offshore and on, will manoeuvre visually to the deck/runway from a few miles out. Plenty of opportunity for maintaining manual skills.
On the only 4 axis machine I have flown, the company -- and the other pilot -- would insist on 4 axis at night or in Poor IMC but I've regularly done manual climb outs to altitude, manual ILS in a variety of weather -- allbeit usually pretty good
-- and, offshore and on, will manoeuvre visually to the deck/runway from a few miles out. Plenty of opportunity for maintaining manual skills.
Join Date: Mar 2006
Location: SW
Posts: 208
Likes: 0
Received 0 Likes
on
0 Posts
Having read most of this and other N Sea threads with interest I've got a few questions (some more tongue in cheek than others):
1) If the autopilot can fly the heli down the approach with 1 pilot monitoring, what is the other pilot doing during the sortie?
2) How much do these semi-sub rigs move? (either in terms of degress of pitch or metres etc).
3) What visual references do you get when in close proximity to the FLt deck or over it in order to hover?
Thanks (in particular to those who have kept the standard on this thread thought-provoking and forward-looking.
1) If the autopilot can fly the heli down the approach with 1 pilot monitoring, what is the other pilot doing during the sortie?
2) How much do these semi-sub rigs move? (either in terms of degress of pitch or metres etc).
3) What visual references do you get when in close proximity to the FLt deck or over it in order to hover?
Thanks (in particular to those who have kept the standard on this thread thought-provoking and forward-looking.
Hold the Phone Jack!
What is this "Monitoring" crap?
One guy is driving the bus....all the time....George may be wiggling the sticks and doing all the monkey skills stuff....but one of the two highly paid professionals in the front seats is supposed to be "flying" the machine.
Call it "monitoring" but that in itself connotes a detachment from an assigned Duty I am not comfortable with. I see one's attention to what is going on as being "involved"....intimately involved and not merely sat back and waiting for an Alarm Bell to sound to get you to sit up in your fancy fleeced lined easy chair.
Even when George is at the Helm.....he is being given Helm Orders by a Pilot.
Are you guys getting so comfortable with not having to wiggle the sticks yourself that you are not paying attention to what is going on to the degree you should?
Do you just reach up and reset a Bug perhaps and not tell the other Pilot why you are doing it and what data point you are looking for or do you leave him/her out of the Loop?
The "Monitoring" Pilot is the Non-Flying Pilot.....and the guy telling George what to do is "Flying" Pilot with every thing that title requires. If the other Pilot doesn't know what your intentions are.....how can he "Monitor" your actions to ensure you don't make a mistake or fail to catch the Autopilot doing what it should?
Pilots Monitor one another.....one of them Flies the helicopter and controls the Autopilot.
What is this "Monitoring" crap?
One guy is driving the bus....all the time....George may be wiggling the sticks and doing all the monkey skills stuff....but one of the two highly paid professionals in the front seats is supposed to be "flying" the machine.
Call it "monitoring" but that in itself connotes a detachment from an assigned Duty I am not comfortable with. I see one's attention to what is going on as being "involved"....intimately involved and not merely sat back and waiting for an Alarm Bell to sound to get you to sit up in your fancy fleeced lined easy chair.
Even when George is at the Helm.....he is being given Helm Orders by a Pilot.
Are you guys getting so comfortable with not having to wiggle the sticks yourself that you are not paying attention to what is going on to the degree you should?
Do you just reach up and reset a Bug perhaps and not tell the other Pilot why you are doing it and what data point you are looking for or do you leave him/her out of the Loop?
The "Monitoring" Pilot is the Non-Flying Pilot.....and the guy telling George what to do is "Flying" Pilot with every thing that title requires. If the other Pilot doesn't know what your intentions are.....how can he "Monitor" your actions to ensure you don't make a mistake or fail to catch the Autopilot doing what it should?
Pilots Monitor one another.....one of them Flies the helicopter and controls the Autopilot.
Last edited by SASless; 5th Sep 2013 at 21:42.
1) If the autopilot can fly the heli down the approach with 1 pilot monitoring, what is the other pilot doing during the sortie?
2) How much do these semi-sub rigs move? (either in terms of degress of pitch or metres etc).
3) What visual references do you get when in close proximity to the FLt deck or over it in order to hover?
2) How much do these semi-sub rigs move? (either in terms of degress of pitch or metres etc).
3) What visual references do you get when in close proximity to the FLt deck or over it in order to hover?
2) It depends on the sea state. Anything up to several degrees in either axis and several meters up or down. However, the aircraft have very strict limitations for landing. For an aircraft the size of L2/225, one set of these limits is +/- 3 degs pitch or roll, 3.5 degs inclination, 5 metres of heave, with a heave rate of 1.3m/s.
3) Depends on the size of the deck and the available light. But, generally speaking, more than enough. Things get tricky when these visual references are reduced. Particularly with a bow deck at night (because all of the lights will be beside or behind you and there might not be a visible horizon).
S-O-L -
1) remember, there are 3 pilots on an automated aircraft, there is the autopilot, the pilot monitoring the autopilot, and the pilot monitoring the pilot monitoring the autopilot. Simples!
2) depends on the sea state, the installation, and whether a particular installation is ballasted and anchored down, or ballasted up for relocation. If you mean what are the limits of movement, it depends but in general 3 degrees in pitch and roll. There are also heave and heave rate limits. 3 degrees probably doesn't sound a lot to you, but since the point of rotation is well below deck height, associated with the pitch or roll is a lot of sway (ie lateral movement) so you can end up chasing the helideck around!
3) approaching, you have a BIG chunk of illuminated metal. Over the deck, it depends a lot on the wind direction. If you have the structure above the helideck in the field of view (derrick, cranes, flare stacks etc) you have fairly good visual references. However if you are pointing away from the structure looking out over the sea, very little, especially at night. You learn to hover looking down at the helideck just in front of you, something your ab initio instructor told you never to do!
1) remember, there are 3 pilots on an automated aircraft, there is the autopilot, the pilot monitoring the autopilot, and the pilot monitoring the pilot monitoring the autopilot. Simples!
2) depends on the sea state, the installation, and whether a particular installation is ballasted and anchored down, or ballasted up for relocation. If you mean what are the limits of movement, it depends but in general 3 degrees in pitch and roll. There are also heave and heave rate limits. 3 degrees probably doesn't sound a lot to you, but since the point of rotation is well below deck height, associated with the pitch or roll is a lot of sway (ie lateral movement) so you can end up chasing the helideck around!
3) approaching, you have a BIG chunk of illuminated metal. Over the deck, it depends a lot on the wind direction. If you have the structure above the helideck in the field of view (derrick, cranes, flare stacks etc) you have fairly good visual references. However if you are pointing away from the structure looking out over the sea, very little, especially at night. You learn to hover looking down at the helideck just in front of you, something your ab initio instructor told you never to do!
Join Date: Mar 2006
Location: SW
Posts: 208
Likes: 0
Received 0 Likes
on
0 Posts
Guys,
Thanks for that. I'm surprised how much they can move despite being such huge submerged chunks of metal. I suppose that if you're flying 4 or more routine sectors in a day, all straight line over the water then having 3 pilots doing the same job has it's benefits in terms of redundancy and safety. It's a fair way removed from how we operate though.
Thanks for that. I'm surprised how much they can move despite being such huge submerged chunks of metal. I suppose that if you're flying 4 or more routine sectors in a day, all straight line over the water then having 3 pilots doing the same job has it's benefits in terms of redundancy and safety. It's a fair way removed from how we operate though.
Hold the Phone Jack!
What is this "Monitoring" crap?
One guy is driving the bus....all the time....George may be wiggling the sticks and doing all the monkey skills stuff....but one of the two highly paid professionals in the front seats is supposed to be "flying" the machine.
What is this "Monitoring" crap?
One guy is driving the bus....all the time....George may be wiggling the sticks and doing all the monkey skills stuff....but one of the two highly paid professionals in the front seats is supposed to be "flying" the machine.
The autopilot does all the work. Its programmed (via the FMS) to fly the whole route to the rig (and back, if the pilot forgets to fiddle with the knobs). One pilot is rambling on and on about how hard done by he is, how little he gets paid, how the whole operation would fall apart without him, how the wife is giving him a hard time etc. the other pilot is pretending to listen to the same stories again and really can't believe the other pilot has forgotten he said these exact same things on their earlier flight together, and is in fact playing Angry Birds on his iPad hoping his colleague won't notice.
Once the manufacturers install an audio gong that sounds as the aircraft reaches the MAPt, when the pilots have to start paying attention, it will be even easier -but don't tell management otherwise there will be paycuts!
SAS You are so 20th century!