Or to take things to their logical conclusion as per Cirrus and Airbus, a really big microcontroller ought to be able to prevent 100% of accidents.
Well, I sure as hell wouldn't subscribe to that view in case of the Scarebus !
Or to take things to their logical conclusion as per Cirrus and Airbus, a really big microcontroller ought to be able to prevent 100% of accidents.
What have Cirrus done in terms of microcontrollers?
They have just removed the mixture lever and replaced it with a crude throttle linkage. The rest of the plane is standard, except for the modern looks.
The parachute is a relatively novel thing (on a full-size certified plane) but I don't see how that helps to prevent bad decisions.
Also the extra Airbus automation doesn't deliver a better safety than what say Boeing have. It comes down to good or bad pilot training, with the highest profile Airbus crashes simply showing that there is some truly crap pilot training, in both the 1st World and the 3rd World.
... But it is important not to confuse pilot mistakes with mechanical failure.
Today's mechanical failure most probably was yesterday's pilots error. Or the error of a maintenance mechanic or a design egineer or quality inspector with the manufacturer. Properly designed and maintained machines do not fail on their own, it always needs a human being to damage it. Engine failure is engine failure, no matter who caused it and why.
... a really big microcontroller
I would rather prefer to have all the superfluous switches and levers taken away. Fuel selector? There is no fuel selctor in the jet I'm flying at work, so why on earth does a Pa28 need one? Carb heat, mixtrure control, cowl flaps, electric fuel pumps, gyro slaving switch, .... 50 percent of all controls in a light aircraft could easily be replaced with simple, mechanical automatic devices, perhaps with a warning light showing if something is wrong and should be fixed before the next flight.
As much as I like to play with microcontrollers - their problem is that they need sensors to work. So every switch and lever in the aeroplane would need an extra microswitch (or similar device) for telling it's position to the "watchdog". What a cabling nightmare and endless source of trouble by itself...
Ok, let me change the question. What are the 3 items, in your opinion that manufactures could incorporate to make the most significant reduction in fatal accidents (I know we can't stop ourselves taxing into sheds!).
Let me start...
now... the pilot shouldn't stall/spin, fly into a mountain or mismanage fuel. If we eliminated just these three items, a big chunk of accidents might just disappear.
Today's mechanical failure most probably was yesterday's pilots error. Or the error of a maintenance mechanic or a design egineer or quality inspector with the manufacturer. Properly designed and maintained machines do not fail on their own, it always needs a human being to damage it. Engine failure is engine failure, no matter who caused it and why.
Yeah... that's true in the stricly logical sense, but one has to draw the line somewhere if looking at it in the context of somebody buying a plane and flying it, hopefully carefully.
If, upon buying the plane, you had the engine taken apart and every part x-rayed for cracks, then reassembled by an expert crew, with duplicate inspection, and no financial pressure, then you have a good start there. But who will do that?
If I bought a used plane, I would probably not fly it but have the engine done by a really good engine shop. Unless it is say over 1.5k hrs, an "overhaul" should not be needed so the cost should be under $10k. Obviously there would be a due penetration NDT of all parts, but that won't reveal subsurface cracks. Only an x-ray has any chance of finding those. But how many people are going to have the engine gone over at the beginning?
Prop strikes (which are vastly more common than most people hear about) can result in subsurface cracks in the crank, the crankcases, the gears, etc. It's going to be very hard to find those (economically). A lot of prop strikes are not shock load inspected and are hidden...
Fortunately sudden catastrophic engine stoppages are rare. If they weren't, GA would be dead and most pilots would end up dead or crippled within a few years.
It is bad that we don't have FADEC on our engines, but does not having it cause a lot of crashes? The worst thing people do is fly full rich, burning some 30% extra fuel. It isn't going to stop the engine running until one climbs to something like 10000ft which almost nobody does, and those who do tend to due some research on how to make things work right up there.
What causes most bad crashes is
The last one is a whole spectrum of course and even the best people get bitten by it.
I can't see any way to really attack these short of delivering exotic stuff like an auto throttle, and some form of "auto land". That is how e.g. Ryanair do it.
But one would need to dramatically revise PPL training too, which isn't going to happen.
The last-1 item is nicely addressed with a fuel totaliser, but it works only if the pilot programs the route, sticks to the route, and has not cut it so close that a change of wind takes him to the brink.
Just seen DMEarc's post. Yes... but that isn't going to happen on 99% of the GA fleet for many years. I also don't think SV has such a huge impact (no pun intended) if one sticks to "classical IFR"; it is marvellous for DIY approaches in zero-zero conditions of course 
And great anytime in IMC especially if you have a second pilot watching it.
Another way to approach this is to ask yourself: if you were training somebody you care about to fly, what would you teach them? The standard PPL syllabus? HARDLY!
I would teach "flying", VMC and IMC combined without a distinction, visual-reference and instrument flight combined without the present clear legal boundary, and definitely landing with an ILS. Anything less is just pretending that clouds and haze will get out of the way when they see a pilot with a PPL...
When I was talking to my flying school about choices of new aircraft, one of the reasons they were trying to avoid Rotax engines, was that a large proportion of pilots will go on to fly older aircraft that have carb heat levers and mixture controls. They figured that it would be a disservice to train pilots on aircraft without these controls, and I tend to agree - it's easier to move from an aircraft with a carb heat lever to one without, than the other way round. Most of us will have flown C152s at one point or another.
Fuel selector? There is no fuel selctor in the jet I'm flying at work, so why on earth does a Pa28 need one?
One argument I've heard is that people flying close to the margins (e.g. bush pilots) like to swap tanks every half hour. That way you get to practice your engine failure drill every flight, then get another half hour to find the airstrip or somewhere reasonable to perform a precautionary landing. I'm guessing the real reason is that it's a heck of a lot simpler to have a switch than a fuel system that can be guaranteed to draw fuel out of both tanks equally. Secondly, when people crash because they forgot to swap tanks, it's pilot error. If something goes wrong with a clever fuel management system, it's Piper error.
Now, my personal feeling is that I prefer to have electronic systems sitting on my shoulder reminding me what I might have forgotten to do, than to have something performing a function for which I then relinquish control. GA SEP aircraft aren't so complex. I'd rather have an alarm warning me of carb ice or that I'd forgotten to lean, than a system that pulls on carb heat and leans for me.
I have a similar philosophy towards navigation. I navigate traditionally, but have an airspace aware yelling at me when I am about to get too close to airspace.
Ok, let me change the question. What are the 3 items, in your opinion that manufactures could incorporate to make the most significant reduction in fatal accidents (I know we can't stop ourselves taxing into sheds!).
Autopilot Autopilot Autopilot
I know most of you on here fly a/c with APs, but many (most?) people in the GA scene do not. An AP hugely decreases workload and frees up mental capacity when things start to head South. Of the two a/c I fly regularly, only one has an AP. I am always amazed, how much lower my workload is when I get into the one that has it.
Another nice thing to have would of course be accurate fuel gauges or a totalizer, but then you'll find people cutting it so fine they still crash....
now... the pilot shouldn't stall/spin, fly into a mountain or mismanage fuel. If we eliminated just these three items, a big chunk of accidents might just disappear.
Yes, but...
Envelope protection: Airbuses have had that since quite some time. Despite it, some crews still managed to exceed the envelope and crash. Every certified lightplane has a stall warner of some sort, most of them even work most of the time. Yet a lot of stall/spin accidents continue to happen. Even aeroplanes equipped with stick pushers have been stalled.
Synthetic vision. For me just another one of those tools to assist the bold ones to push the limit a little further. Minimum? What do I care, I have synthetic vision... And as we recently saw in San Francisco, even on a CAVOK day with perfect natural vision pilots are able to crash land perfectly functioning airliners 100 yards short of a perfectly visible and marked runway.
Fuel management. The biggest part of fuel related accidents were caused by pilots simply running out of fuel. No system in the world can save anybody from continuing to fly when the tanks are empty. And regarding fuel management: A simple ON-OFF tap is all that is required. With a large red warning light in the middle of the panel when it is in the OFF position.
I think the only real remedy against accidents is training. I 100 percent agree with the slogan of a well known training organisation: "The best safety device in any aircraft is a well-trained crew."
I'm guessing the real reason is that it's a heck of a lot simpler to have a switch than a fuel system that can be guaranteed to draw fuel out of both tanks equally
Absolutely so, and also it enables a solo pilot to balance the aircraft in flight, by running down the left tank. On the TB20, which is not exactly a tiny plane (1400kg MTOW) there is a big change in the roll force needed to keep it level if you don't do that. Presumably less so on a Cessna, which doesn't have the fuel selector, due to the wings being high up so the hull "hangs down" from them.
One really bad thing in GA are the crappy fuel gauges. Again, there is a solution but it costs money.
Autopilot Autopilot Autopilot
Yes 
But that is more money...
Presumably less so on a Cessna, which doesn't have the fuel selector, due to the wings being high up so the hull "hangs down" from them.
Peter, Cessnas DO have fuel selectors and some people do exactly what you describe. That said, Mr. Cessna has thought of the issue and due to the position of the vent line, the left tank draws fuel before the right one, even on crossfeed. You see, early automation at work ;-)
