[pricechild]

Howdy! FI(H) here. Looks very cool, I watched the video too.

Many students get the hang of hovering within 10 hours - "learning to fly" really isn't that hard? I see the rest of the course as "learning to become a pilot." That means the background knowledge and practice to allow decision making & dealing with emergencies etc.

I think if you could improve just a single one of the problems you touch on (difficult controls, navigation, mass and balance, fly by wire etc) you could make a lot of money...

... But improving all of them in one shot... that sounds unreasonable to me? Are they really connected?

> Our system makes it impossible to lose control of the airplane, potentially solving 80% of today’s fatal accidents in general aviation.

I'd love to see details on how you achieved this... or at least your definition of "impossible"...

[n_ermosh]

You are definitely correct that "learning to become a pilot" is the meat of training. Our belief is that we want to make the continued act of "aviating" as easy as possible so that pilots can focus on the decision making aspect of being a pilot, and when doing so, not let the airplane get ahead of them and cause an accident.

Basically, you can't command the plane to do anything it can't safely do. If you pull the stick full back, you'll just climb at your maximum safe angle of attack. Hold the stick full right, you'll be turning at a 45 degree bank. Compared to todays planes, which you have to actively be on top of and "ahead of" the entire them they are being flown.

[pricechild]

I get the idea... but what happens when a pitot tube is blocked or an aoa sensor gets stuck or ... ?

With "traditional" controls and an autopilot failure, you still have control over all the control surfaces.

With your solution, you don't have enough hand axis to fall back to manual flying? How can the computer possibly guarantee safe flight in all conditions?

In the case of an e.g. air speed sensor failure, how do you get on the ground safely? Is the answer "Just BRS"? Or multiple sensors etc. ?

[n_ermosh]

Multiple sensors, actively modeling the flight dynamics to detect anomalous sensors/behaviors (i.e. is what we are measuring what we expect to measure), and BRS as a final backup.