Live data from Hacker News

USS McCain collision ultimately caused by UI confusion

arstechnica.co.uk

171–180 of 369 posts

Re: USS McCain collision ultimately caused by UI confusion

#171
post #13

Earlier quoted context omitted.

Hearing (one of many) alarms is one thing. Feeling the control wheel or stick fighting you is far more direct. AF 447 was absolutely a UX failure.

If you're going to go against the conclusions of the accident report, which is very thorough, you'll have to do more than just assert that the side sticks were responsible. Check it out here: https://www.bea.aero/docspa/2009/f-cp090601.en/pdf/f-cp09060...

First, I did not say the side-sticks are the fully responsible factor for the accident, but a contributing factor.

Second (and as noted in the parallel subthread), the BEA report does not include cockpit transcripts in section 2, 2.1, and relevant subsections, in describing the cockpit situation. It mentions specific utterances of the crew, but not in the specific context and sequence they occurred. Other reports do include these transcripts and from them it's highly probable that multiple inputs were being supplied, or that even if that was not the case it could not be ruled out readily.

There are multiple references in the BEA report of the level of confusion and complexity of the situation. The merits of scheduling the Captain's relief in the ITCZ crossing, the flight path itself relative to other aircraft which had diverted around the region, the loss of instrument readings, autopilot disconnects, multiple audio alerts, the disabling of a master alarm that wasn't clearly deliberate, the stated confusion of both co-pilots, the switching to and from "alternate law" of aircraft controls, and more.

Adding ambiguity of flight controls, feedback, and possible multiple inputs in the context of all of this again strikes me as, well, a disaster that actually happened.

In a description based on the full transcript:

02:11:21 (Robert) On a pourtant les moteurs! Qu'est-ce qui se passe bordel? Je ne comprends pas ce que se passe.

We still have the engines! What the hell is happening? I don't understand what's happening.

Unlike the control yokes of a Boeing jetliner, the side sticks on an Airbus are "asynchronous"—that is, they move independently. "If the person in the right seat is pulling back on the joystick, the person in the left seat doesn't feel it," says Dr. David Esser, a professor of aeronautical science at Embry-Riddle Aeronautical University. "Their stick doesn't move just because the other one does, unlike the old-fashioned mechanical systems like you find in small planes, where if you turn one, the [other] one turns the same way." Robert has no idea that, despite their conversation about descending, Bonin has continued to pull back on the side stick.

http://www.popularmechanics.com/flight/a3115/what-really-hap...

Re: USS McCain collision ultimately caused by UI confusion

#172

Earlier quoted context omitted.

Is there a purpose to the display being so colorful? If AP were the only green thing and the others were monochrome it would stand out more.

Most of the colors on that display have a rationale. For example, on the speed tape on the left, the white, green, and yellow colors all map to colors of a standard airspeed indicator. That the course (CRS) is green instantly tells me I'm navigating by a ground-based VOR, not GPS (which would be magenta). Same with the green diamond on the right by the altitude indicator. The waypoints at the top are all GPS-calculat…

I'm impressed. But how did you figure all that out in the first place?

Re: USS McCain collision ultimately caused by UI confusion

#173
post #11

Earlier quoted context omitted.

I'd call it an interpretation, not a theory. But clearly the conflicting inputs was a UI problem, and while there may be tradeoffs, a better UI could have prevented the AF 447 crash.

Basic airmanship could've prevented the crash too. The pilot-in-command got a stall warning and pulled back on the stick, climbing to 38,000 feet. That's not a UI issue, it's a forgetting one of the most fundamental rules of flying issue. https://aviation.stackexchange.com/questions/1418/what-happe...

UI issues contribute to situational complexity, stress, ambiguity, and failures to respond appropriately to circumstances.

And whilst the BEA report ... rather inexplicably, frankly ... fails to mention the inputs issues, it does address all the other factors I've described here.

Re: USS McCain collision ultimately caused by UI confusion

#174
post #17

The real cause of the collision was poor seamanship and failure to follow the COLREGS on the part of the bridge crew of the destroyer. Most specifically they were seriously in violation of Rule 6: the Safe Speed Rule http://navruleshandbook.com/Rule6.html Every vessel shall at all times proceed at a safe speed so that she can take proper and effective action to avoid collision and be stopped within a distance appropr…

When will the first ship crash because of a BSOD? You think I am joking? No, unfortunately several types run on general purpose OS, and the possibility is there.

http://www.popularmechanics.com/military/navy-ships/news/a27...

Re: USS McCain collision ultimately caused by UI confusion

#175
post #147

Earlier quoted context omitted.

> The reason such TVs don't exist is simply engineering laziness. If your answer to the question "why can't a billion dollar industry provide me with instantaneous channel switching?" is "because they are lazy", you are severely misguided (at the very least!). There are hundreds of extremely talented engineers working full-time on these problems. If the solution were as easy as you propose, the issues would be solved…

The distance to the satellite has nothing to do with it. You're using big numbers to exaggerate the problem. Decrypting and demodulating have be as fast as the data is coming, otherwise the delay would grow until you never saw the end of the show. So that's not the reason either.

Yes, the MPEG-2 decoder in your TV needs to be fast enough to keep up with the transmitted data rate, but that doesn't mean that it has instantaneous stream startup times.

Your TV doesn't have the information it needs to start displaying MPEG-2+ATSC A/53 video for as long as 700ms. It needs to buffer quite a bit of timing information and at least an I-picture frame before it can start displaying anything: http://www.bretl.com/mpeghtml/startbuf.HTM

Not only that - the overall standard has tight timing tolerances to keep video arriving at the encoder/transmitter in-sync with audio/video leaving the decoder in your TV. It's not like MP3 streaming where every client has its own buffer size and time to start playing depending on network conditions, bitrate, etc.

Having a reliable buffering process to keep receivers in-sync is a feature, not a bug.

Re: USS McCain collision ultimately caused by UI confusion

#176
post #158
post #147

Earlier quoted context omitted.

> The reason such TVs don't exist is simply engineering laziness. If your answer to the question "why can't a billion dollar industry provide me with instantaneous channel switching?" is "because they are lazy", you are severely misguided (at the very least!). There are hundreds of extremely talented engineers working full-time on these problems. If the solution were as easy as you propose, the issues would be solved…

> If your answer to the question "why can't a billion dollar industry provide me with instantaneous channel switching?" is "because they are lazy", you are severely misguided (at the very least!). There are hundreds of extremely talented engineers working full-time on these problems. If the solution were as easy as you propose, the issues would be solved. Simple solution: just put 10x decoders/tuners for most recentl…

Exactly. Just having two decoders instead of one would drastically improve the 80% use case of consecutive channel surfing.

Re: USS McCain collision ultimately caused by UI confusion

#177

Earlier quoted context omitted.

User interfaces have gone backwards decades in quality. Look at your car radio in a 90's car. You didn't need to look at it to figure out how to push the radio button, turn the volume, anything. You could even adjust the bass on the fly with a bass knob instead of navigating to your "eq" menu to do it. Now, you've got to look away from the road, realize what "mode" your in, you can move the menu around to the "audio"…

Don't get me started about TVs. When I grew up our family TV turned on in seconds, and pressing channel up/down reacted in tens of milliseconds.

We have some recent TVs. They have no buttons. Including power buttons. If you want to turn them on, you need the remote control.

Re: USS McCain collision ultimately caused by UI confusion

#178
post #73

There was a ferry accident involving a similar transfer of control problem in New York harbor.[1] Ferries have a lot of maneuvering modes and directional thrusters, because they do so many dockings, and so they have a more complex control problem. The pilot put the system into a backup dumb mode while in cruise, then changed stations to where he could best see the dock while still in the wrong mode. Rammed the dock a…

> The brass hated it. The person in the pilot's seat was really the one in charge. The officers were just back-seat drivers. To this day, the Navy sets up nuclear submarines so that different people steer, control power, and control buoyancy.

After reading Ben Rich's Skunk Works memoire, and coming across some other references to the Navy's intransigence in some areas, this fits right in.

Re: USS McCain collision ultimately caused by UI confusion

#179

Earlier quoted context omitted.

The distance to the satellite has nothing to do with it. You're using big numbers to exaggerate the problem. Decrypting and demodulating have be as fast as the data is coming, otherwise the delay would grow until you never saw the end of the show. So that's not the reason either.

Yes, the MPEG-2 decoder in your TV needs to be fast enough to keep up with the transmitted data rate, but that doesn't mean that it has instantaneous stream startup times. Your TV doesn't have the information it needs to start displaying MPEG-2+ATSC A/53 video for as long as 700ms. It needs to buffer quite a bit of timing information and at least an I-picture frame before it can start displaying anything: http://www.…

And the way to solve this -- at least for the case of consecutive channel surfing -- is to have multiple decoders and multiple buffers. It's simply precaching.

Re: USS McCain collision ultimately caused by UI confusion

#180
post #172

Earlier quoted context omitted.

Most of the colors on that display have a rationale. For example, on the speed tape on the left, the white, green, and yellow colors all map to colors of a standard airspeed indicator. That the course (CRS) is green instantly tells me I'm navigating by a ground-based VOR, not GPS (which would be magenta). Same with the green diamond on the right by the altitude indicator. The waypoints at the top are all GPS-calculat…

I'm impressed. But how did you figure all that out in the first place?

RTM and training.
Post reply on HN