Another driver says Tesla’s autopilot failed to brake; Tesla says otherwise
arstechnica.com
arstechnica.com
Then they go one to essentially disavow autopilot - "...but it does not turn a Tesla into an autonomous vehicle and does not allow the driver to abdicate responsibility."
There are air incidents in which the pilots were unclear about whether their autopilot was in charge or they were. That's extremely dangerous.
I think Tesla needs to impose some very clear indicators when they are in autopilot mode vs manual mode (lights on the A pillars, etc.). They also need to revisit what exactly disables autopilot mode. Tapping the brakes should not disengage auto braking for example, only accelerating should do that.
There was a similar accident in Aeroflot Flight 593 https://en.wikipedia.org/wiki/Aeroflot_Flight_593
> With the autopilot active, Kudrinsky, against regulations, let the children sit at the controls. First his daughter Yana took the pilot's left front seat. Kudrinsky adjusted the autopilot's heading to give her the impression that she was turning the plane, though she actually had no control of the aircraft. Shortly thereafter Eldar occupied the pilot's seat.[4] Unlike his sister, Eldar applied enough force to the control column to contradict the autopilot for 30 seconds. This caused the flight computer to switch the plane's ailerons to manual control while maintaining control over the other flight systems. A silent indicator light came on to alert the pilots to this partial disengagement. The pilots, who had previously flown Russian-designed planes which had audible warning signals, apparently failed to notice it.
Aeroflot Flight 593 Crash Animation + CVR : https://www.youtube.com/watch?v=RrttTR8e8-4
That's just the interim period before full autonomy. Does anyone know how insurance claims are handled in collisions between fully autonomous cars?
What's "unbiased" information? For example if the user (driver) had the wrong model of the new cruise control she may have done what she considered correct. And if she didn't understand, was it her problem for lack of consulting the manual in detail, or the company's for not explaining it properly?
> That's just the interim period before full autonomy. Does anyone know how insurance claims are handled in collisions between fully autonomous cars?
As soon as full autonomy is better than humans, the insurance companies will really push for it. In this context it makes sense for the liability to go with the manufacturer / software developer in the case of an unmodified vehicle, both because it aligns incentives correctly and because any patch can be immediately rolled out to the entire fleet.
(I presume individuals won't buy cars at all)