The Auto-Pilot iPad App

I think the iPad's GPS is just to find a nearby waypoint as a starting point for the automatic approach and landing. Once you have decided upon a route, it's a simple matter of tuning the aircraft's nav radio to the VOR or ILS. From there, you have the full precision that you would normally have on an instrument approach/landing. The innovation is that the passenger doesn't have to know what an instrument approach is.

(I wonder if the code on the iPad gets the audio stream from the radio so that it knows it's tuned to the right channel.)

As an aside, I've read on the X-Plane forums that X-Plane doesn't actually do a full simulation as you say it does. If you try to spin the default C172, for example, you can't do it, because the aerodynamics near the boundary cases are not calculated correctly. (Either that, or the spin recovery for a C172 consists of letting go of the controls and waiting, which is not what the manual says. But I have not tried it.)

But if it connects to the aircraft, it could use the aircrafts instruments, at which point it doesn't use the location based API's :P

FTA: "Using the airplane's onboard satellite navigation data, the software finds airports within gliding range, assessing distance, obstacles, runway lengths and other variables. It issues commands to the plane's ailerons, rudder and elevator to steer toward the best option".

You can legally test and use this on experimental-amateur built aircraft. Which is what Austins Lancair evo is.

I agree with most of what you wrote here. One point I would like to add though is that there were massive amount of military investments that went into plane development during both world wars and the cold war, allowing complete new features.

Also, it's not like no ones tries to bring new technologies on the market. But the number of planes sold is nowhere near the number of cars sold, for instance. So new investments are pretty risky. Look at the Dreamliner or the A380... Both delayed and still with major design issues. In the general aviation, a European motor maker introduced diesel motors for plane, which allows a much lower mileage, and an economic flight of even 2-props planes (see DA-42). There were so much technical issues that the company got bankrupt. It's not a compelling argument for new customers tempted to buy such a diesel plane. When you buy a plane, you don't want the company to go bankrupt in 3 years, so you take a Piper or a Cessna, even if these are massively inefficient.

Finally, look at a C172 glass cockpit. The outside design may still be the same as it was 40 years ago (as is the motor, almost), but the cockpit is totally different, with brand modern software, complex GPS & autopilot capabilities, etc. Flying has become quite different in the last 10 years because of that.

He's counting on the few sales made when people are willing to pay $50,000 for the app.

Have you seen the cirrus aircraft? Unskilled, but rich people are willing to pay > $500K for a plane with a parachute and a button that auto-rights them.

The result is a 6x fatality rate compared with the average for general aviation. That's including some horribly twitchy aircraft in that average.

If the pilot of a small airplane experiences a medical emergency that leaves him incapacitated, this sort of device could mean the difference between life and death for anyone else onboard the plane (assuming their not pilots as well).

You don't get to run apps on the planes EFIS systems. They're sealed, controlled units for /very/ good reason.

Aircraft aren't like hacking on some webapp, you don't move fast and break things. You test, test, validate, prove, test then finally ground test. Then test in parallel in the air. then expand envelope, only then do you get to think about relying on it.