Originally Posted by
Ascend Charlie
As has been mentioned before, Negative Habit Transfer is something to be avoided. Learning to do things in a certain manner because that is the way the sim works, will cause confusion in the real machine when you put in your control movement and the aircraft does less or more or nothing at all.
A sim will behave the way it is programmed to do, and it is NOT exactly the way a real machine behaves. And unless the cockpit has a 360-degree visual ability, you will be missing a lot of very important visual cues and sensations from a real machine - sunlight through an overhead window, sounds of wind. Screaming instructor.
Sims are hugely helpful to learn systems operation and for instrument flying, but the real thing is needed for the real flying.
So, people have been replying about a wide range of simulation... all the way from consumer games to Level D sims. I was thinking more toward the Level D in terms of fidelity. You don't need 360 degrees of visuals - we don't have that in the real aircraft. The Redbird VTO has 200° of horizontal field of view and 70° of vertical field of view. If that's not enough, it's close, and in this day and age simply a matter of adding a few more graphics processors and screens. I would postulate that vertically, you need to see a bit above the rotor disk, and right down to the chin bubble. Horizontally, not sure - I do know the Frasca cockpit simulator I used at Bell in the 90s didn't have nearly enough (it was a single projector), but it's difficult for me to believe you would need much more than 200°... I don't have much more than that in the L3 I fly... BTW, another reference to the Redbird VTO:
https://www.ainonline.com/aviation-n...copter-trainer
As for motion base, I read (but can't find) one of the Redbird execs saying they had to redesign the motion base in order to get the correct feel for a helicopter. My recollection is that it was the responsiveness that the existing fixed wing motion bases didn't supply - the helicopter needed a faster response time to give a correct feel. I also read that the design they arrived at also can be run from a regular electrical plug, i.e. it avoids having to have special electrical wiring installed, so it's probably a cost saving thing as well. I'm a little surprised that they think a motion base is required... I supplied computer gear to Boeing during the design of the 777 and got to fly their sims at Renton. Only one of the engineering simulators had a motion base (the 747 cab) the others were fixed and yet were totally convincing to fly (as well they should be - the test pilots got out of those sims and went and flew the real thing during certification).
As for the simulation needing to be EXACTLY the same as the helicopter, I disagree. If that was true, then learning in the R22 wouldn't be of any advantage to learning to fly a 407 or 412 or... name any other helicopter. But of course it is; there are differences and we train for that during differences training or type ratings, but the amount of negative transfer from learning one helicopter to learning another is recognized and trained for. I would assume that if you allowed 50 hours of sim and 5 hours of dual in the aircraft, that you would be covering exactly that during the 5 hours.
Originally Posted by
WillyPete
But do you assume these guys have their software set up properly?
Every new patch changes the flight profiles in DCS, touted to be one of the most accurate civilian sim products out there.
Who will inspect them?
If a Robbie doesn't meet standards you can end up dead, but who will there be to measure actual cyclic and collective response and travel on every sim?
This is covered by the FAA in
http://rgl.faa.gov/Regulatory_and_Gu...E/Contents.pdf an AC. There are required qualifications for upgrades, and there are requirements for Recurrent Evaluations of the simulator, every 6 months unless you apply for some of the alternative processes.
Originally Posted by
Twist & Shout
I understand the level D SIMs cost circa $20M
Yes, but an interesting question is how much of that is because of the complexity of older motion bases and cockpits? I interviewed with Evans & Sutherland back in the 80s when they sold visual systems for simulators for... it was either $10M or $20M at the time, just for the visual system. Now that performance is dwarfed with a single GPU, and systems like the VTO are running multiple GPUs... you can essentially scale the graphics as much as you want... several sims in recent years are using flat panel displays rather than projectors, so it's just a matter of how many panels you want to install...
Similarly, motion bases are much cheaper since people have started using electrical, rather than hydraulic servos. Also, I would claim that for training in a basic helicopter simulator, you probably don't need heave/sway/surge, but simply roll, yaw, and pitch motions (if that).
Finally, in sims like I contributed to at Boeing and Hughes (for the F/A 18 dome simulator) the cockpits all had all the avionics out of a real aircraft, i.e. hundreds of thousands of dollars worth of avionics. The cost of avionics in a trainer helicopter is in the tens of thousands, and many of them are simulated today with a flat panel display along with some knobs on a thin panel, i.e. the cost of the cockpit is much less than that of the actual aircraft. Some equipment like an approach certified GPS typically will use actual instruments (because it's too costly to simulate, although that has been done with the Garmin 430), but equipment like the basic pitot/static system, engine manifold pressure, attitude indicator... these can all be done quite inexpensively with flat panels today.
So the question is, if you can deliver a system like the VTO for $150,000 (and you assume those visuals and motion base are indeed good enough for primary training), then the question is how much the software to get Level D fidelity would cost? I think it's highly unlikely that you would need to charge hundreds of thousands of dollars for the aircraft modeling software.
Originally Posted by
bigglesbutler
The reason for the limit on sim time is because you can't learn everything in them. Sims are great for the logical and procedural based learning but that comes after you have learnt to hover
Of course, Redbird may be full of ****, but since they specifically designed the VTO for teaching hovering and autorotations, I would think they'll get sued if you can't in fact learn to hover in it. I'm actually not all that skeptical of being able to learn to hover in a simulator, assuming you have wrap-around visuals. Learning to attitude fly is mostly a visual skill, learning how to move the controls so that you see what you want to see out the window. There are other cues that we all use of course, but the visuals are by far the most important.
Originally Posted by
Lonewolf_50
Paul, how much time have you spent as an Instructor?
I've been teaching for 30 years.
Originally Posted by
paco
I have a 206 sim here with proper controls and panels being used with the Oculus Rift (VR thingy) and a leap controller and the results are nothing short of startling, even with the FSX Beaver.
I think I know the sim you are talking about - one of our customers was evaluating it to see whether it would be useful. I look forward to trying it myself, although I think it's a far cry from the sort of motion base sim I was thinking about. Still, it's an example of how much can be done today for modest costs - I can see lots of scenario based training being done in a basic simulator like you are talking about.
Originally Posted by
alphanumeric
Learning to fly in a simulator from zero hours, then transferring to a real aircraft would make a PPL about £60,000
Not sure if you're saying that's a lot or a little! Here in the US, the basic FAA minimums will run you about $13,000 for a PPL, but people usually end up closer to $20,000. If you assume you could rent a capable sim for $150/hr you could save the average guy money, or you could give him 120 hours of dual in the simulator, plus 5 hours in the aircraft which arguably would make him a better pilot. Or, perhaps 90 hours in the sim and 20 hours in the aircraft (which is, probably, more realistic).
Nobody so far has responded to the part about how no junior instructor will want to be teaching in the simulator, because he won't be gaining PIC hours. And, at least in the US, there is a similar problem with the commercial pilot rating because you need 150 hours of "flight time" including 100 hours of Pilot In Command time... so unless the FAA changed those requirements to allow a substantial amount of that in a simulator, you might end up making the rating more expensive for the candidate.