No it is not a solved problem _for drones_
Heck, it worked for the LLRV trainer which was similar if not worse deathtrap than a multirotor, just had a massive jet engine instead of props.
Has any drone manufacturer integrated an ejection seat? Start talking about that and watch the masses lose interest in flying around in a drone real fast. Training or not that isn't an option.
I know, that's why I have said it isn't going to fly. But strictly speaking there is a solution to that problem.
Ballistic parachutes are exactly what I meant when I said parachute Still harder than it looks, as you said, and only part of the solution.
Well, parachute could also mean some hapless guy opening the door and jumping out with a backpack. Ballistic parachute is a lot faster to open than that + you rely on deformation zones of the craft's body to cushion the impact. If you jumped at 400' with a normal chute, you likely wouldn't survive it.
So the original statement stands. No humans in drones for the masses until the low altitude survivability issue is solved (properly).
Agreed to a point. It is not necessarily the low altitude survivability that is the main problem - planes and helis will also not end up well if something goes wrong at a low altitude (you need altitude to enter autorotation in a heli and speed + altitude to glide in a plane) and that didn't prevent their use, even in passenger transport.
I think the main issue is the fact that should anything go cactus with the rotors or engines, the thing will just flip over and crash, regardless of altitude. You can either make each pylon have two motors + rotors (like the one that is supposed to fly in Dubai) or you have more than 4 pylons so that the rest can take over for a failed engine/rotor and land it safely. However, the more things you put in that craft, the higher the chances that something is going to break.
All of that is going to be insanely expensive and complex, not to mention the noise. No advanced AI mumbo jumbo can beat basic physics there. At the end it will most likely be much cheaper to take a normal helicopter (most have only a single engine, only the larger ones have two) and either modify it for autonomous/remote operation (that exists already) or simply put a pilot behind the controls.