Re: OT: Ladies and gentlemen, welcome to Hamburg, we will try landing again shortly.

Jan Panteltje wrote:
On a sunny day (Mon, 03 Mar 2008 10:31:14 -0800) it happened Tim Wescott
<tim@xxxxxxxxxxxxxxxx> wrote in
<hbadnd9NS6R-3lHanZ2dnUVZ_jidnZ2d@xxxxxxxxxxxx>:

Jan Panteltje wrote:
Ladies and gentlemen, welcome to Hamburg, we hope you also enjoy our next landing attempt.
http://edition.cnn.com/2008/WORLD/europe/03/03/germany.plane/index.html

Like I wrote: there was a bit of a hurricane.
Windshear, WHAT windshear?

Do you think a robot could do better?

A robot could certainly do a better job of scraping the wing...

Given what I see in the video, it looks like some excellent flying. If the pilot did anything wrong it was to line up with the runway to early -- but I can't criticize, because if I tried it you'd just have a big wad of aluminum, carbon fiber, and body parts.

Well I do not understand the last sentence, my view is that they should not
fly in this weather, but my view is also that robots (or robotics) have
a better attention span, can make many corrections per second, and possibly
could have kept the plane more level (he is wobbling all the way).
Depends a bit how those robotics were designed of course :-)

Modern fighters have instable flight, and correct settings many times a second,
so the technology is there.

My point in the last sentence is that since _I_ couldn't do nearly that well, I'm not going to chew the pilot's ass. At the very least I'm not going to do so unless I hear from more than one transport pilot who's really looked into it and who points out specific violations of recognized conduct.

I suppose I could criticize the decision to land at all, but (a) I'm still not a pilot, and (b) I've seen videos from other airports (Singapore, I think) where crossed-up landings like that are _routine_, because of the prevailing winds and the direction of the runway.

There are limits to what an automatic control system (your 'robot') can do. Regardless of the capability of the pilot or controller, the aircraft will still fly within the capabilities of the airframe and control surfaces; some wobbling in gusty conditions will always be there, and unexpected gusts will cause unexpected aircraft behavior.

In fact the X-29 fighter, which was intentionally designed from day one to be dynamically unstable in subsonic flight, wouldn't have had acceptable control margins for a 'real' fighter. Why? Because the between the speed of the instability and the bandwidth of the actuators and sensors, the control system _could not_ achieve acceptable phase and gain margins. The control surfaces simply could not be moved fast enough to maintain the stability that would be required for anything but an experimental fighter.

The only advantage that I could see to a robotic pilot is that with the right sensors it may be able to sense gusts quicker than a human with his *** glued to a chair. On the other hand, when stuff happens to the aircraft that are beyond the range that was predicted by it's designers, it's nice to have someone in the hot seat who can re-think the control rules on the fly (as it were).

Modern jet liners have fly-by-wire, although the civilian transport industry currently avoids stability augmentation in favor of guaranteed-stable airframes. You may be able to do something with shear detectors on wingtips that are coupled to the flight control system -- this sort of thing gives you an even chance of combining the strengths of pilot and the control system vs. combining the weaknesses, so it's not something you want to just jump into.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html
.

Loading