Now its getting really dangerous

But NOT what you need !NOW!.

why do you think that?

they probably have hundreds of people going through all imaginable and unimaginable scenarios, plus every case FAA etc. ever documented

-Lasse

Yeah, as *if* you could design a menu system that is optimal for every situation... If nothing else, how can a person navigate a menu structure that is constantly changing based on external stimuli?

--

Rick

you could handle the 99% the rest is why there is still a pilot

you wouldn't have to change the menus, just something like;

engine is on fire!, do you want to show the engine controls and the engine on fire checklist?

When someone crashed because they were disturbed and missed steps on the preflight checklist, they came up with an electronic checklist so that you cannot skip steps.

-Lasse

This system is the interface between the pilot and the airplane. The pilot can't do anything to make up for a poor interface.

We'll let you work on this.

--

Rick

On Wed, 05 Nov 2014 16:04:48 -0800, Robert Baer Gave us:

You have to say it in Russian. (or is it Cambodian?)

Yep. However, how many pilots out there purchase commercial airliners? The people that sign the checks for the big birds are airline upper management and government functionaries. Many probably don't have any flying experience. Oddly Virgin was shopping for pilots with no flying experience even though Richard Branson is a pilot. Strange: Anyway, the sales droid in the video is selling to management, which doesn't respond well to acronyms, buzzwords, and technobabble. He's pushing the warm fuzzy feel of the latest technology, not the hard reality that it just might not work or be unusable.

Keep going. You're on the right track. By automating just about everything, the next step is to put the pilot on the ground, and fly the airplane like a giant drone or UAV. The technology is here today, but on a small scale:

I don't believe that the number of buttons is really a problem. The

747 flew with more clutter and buttons than your worst nightmare: What changed things was the lack of panel space for instruments. Sensors were being stuffing into everything possible which required threshold settings, alarms, indicators, etc which added instruments, not buttons. More instrument space justified the original EFIS displays, which had multiple pages of virtual instruments and thus saved panel space. The buttons are mostly still there, but the displays had shrunk. The next step will be to virtualize the buttons, which is what the video is about. After that, virtualize the pilot, and we have a UAV.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

That doesn't even make sense. Automation is not required to fly a plane remotely. The two are unrelated. I flew a plane remotely when I had a hand control via string for a 0.049 cuin, the original fly by wire. No automation at all.

Again I don't follow. If you need more space you either need a bigger panel or you need fewer buttons and instruments. Even if you have the space, it becomes a problem remembering where all of the many dozens of controls are. There is a limit at some point.

Every instrument needs controls.

--

Rick

No imagination needed. "Without warning; a lightning bolt struck the aircraft and concurrently encountered severe turbulence. All EFIS screens went blank; static in cockpit so loud that intercom to backseater unusable; generator fell off line; landing gear unsafe lights on; VHF comrdo's overcome by static; autoplt unusable.

I have an LCD radio panel somewhere that took a static electricity blast through the front panel. It looks like miniature black lightning bolts, complete with radiating fingers. Unfortunately, the lightning also killed the drivers, so the panel won't light up.

Push buttons are fairly rugged. Virtual push buttons, less rugged.

I run Flight Simulator X, which offers voice control with an add-on program: It also works nicely with a touch screen addon: Todays games, tomorrows reality.

The problem with technology is that anything that you can imagine, will surely be tried by some enterprising inventor or suicidal maniac. The difficult part is separating the good, bad, stupid, and dangerous ideas as the distinction is usually not evident until a few people die trying.

Another problem is that people have strange ideas of how things should work. For example, a friends Samsung S4 smartphone includes S-Voice, which sucks and was replaced with Dragon Mobile Assistant. He showed me how he makes a phone call by merely saying "hello dragon" followed by "call jeff". While talking to me on my cell phone, I asked him to tell it to hang up. It didn't work. He tried a few other commands, which also didn't work. Gee... he made the call using voice and he should have been able to end the call the same way. Nope. There's too much risk in having the program interpret the words of the conversation as commands, so Dragon just disabled it. I can imagine similar "mode sensitive" situations in an airplane. (Think state machine with undefined states).

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Like this?

True, automation isn't required, but it sure makes things easier. For example, synchronizing the rudder, ailerons, and elevator makes a coordinated turn instead of a clumsy sideslip. Monitoring flight parameters to prevent stalls, excessive g forces, and generally preventing damage to the aircraft are all automation. You don't need them to fly the airplane. You need them to fly the airplane without breaking it.

However, I'll admit that such automation isn't always effective. The dumbest example is the gear-up landing alarm, which warns the pilot that he's about to do a belly landing and ruin his insurance policy. Still, despite the automation and alarms, we still see gear-up landings: My favorite was the story where the pilot proclaimed that he couldn't hear the tower yelling at him to abort the landing because his landing gear was up because the gear-up alarm was making too much noise.

Sure. Just compare the panel space for the A310 with what Thales proposes. The Thales console as maybe 20% of the space used in the A310. However, not all the controls, switches, buttons, and indicator need to visible at all times. They only need to be available when needed, as in the multiple pages of the EFIS display. So, a much smaller panel can be used without loss of functionality. Digging through pages of virtual switches and indicators may not be as convenient as reaching directly for a control, but it will work (given sufficient training and experience).

Well, maybe I should say that. There are heater and A/C controls in my Subaru that I still don't understand how to use them or how they work.

Maybe, but instead of controls, how about a supervisor circuit (or perhaps a hypervisor)? The current mentality is that the solution to unreliable electronics or software is to add more unreliable supervisory electronics or software. Whether that will work is largely a function of the implementation. Watchdog circuitry does have it's place but I've seen places where it didn't work. Cut-n-pasted from my posting to rec.radio.amateur.antenna:

...I had much the same problem except that the self test function was aptly named BITE (Built In Test Equipment). All too often, the radio would be fully functional, but the system would be down because the BITE was complaining about something spurious. Someone suggested a second BITE to monitor the first. That was prototyped by some other engineer, which further reduced the reliability of the system. Eventually, I was handed the mess, with instructions not to spend any money, not to change anything, and get it done before my boss returned from vacation (so he couldn't be blamed if anything went wrong). My solution was to redesign the interface between the radio and the BITE so that the BITE boards could be removed without affecting the operation of the radio. The result was predictable. All the BITE boards were immediately removed on arrival by the customer, thrown into an unmarked cardboard box, and left to rot while the radios continued to operate normally without them.

Morals: Complexity does not improve reliability and the first thing to fail is usually the fail safe.

Somewhere, there's a min-max solution that offers the best compromise between electronics supervisory control overkill, and system usability and reliability. Too many controls and too much supervision (as in my BITE example), don't work well. Oversimplification hides too many important indicators and diagnostics. Too much complexity is an opportunity for additional failures. Manual controls and overrides are great for knowledgeable users, but useless when nobody understands how things work. Often, human intervention just makes things worse (TMI). So, we do the balancing act between automation and control, and adjust for failures as we blunder forward. If you have a better way, I'm all ears.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

...only AFTER a solid RFQ or sale. Betcha it is not complete yet.

The international language of air-traffic control is English.

Presumably all pilots can be relied on to speak enough if it communicate with the plane as well.

--
Bill Sloman, Sydney

On Sat, 08 Nov 2014 14:13:41 +1100, Bill Sloman Gave us:

It was a Firefox reference.

It was a Firefox AND a technology reference. The iPads are made over in the ring of fire somewhere. Cambodia was just one guess.

We could always insist they bring back the Flight Engineer position.... That moves 1/3rd of the instruments to the back and provides someone to watch the pilots. :-)

Just kidding....

Steve