Quote:
Originally Posted by Dadhawk
The Summon was a hypothetical more than anything. I just have an issue reconciling the logic that says "our car is a L2 only driver assistance car" and "our car can drive itself around a parking lot without you in it". Which one is it, because to me, either the car can drive itself or it can't. I have a similar concern about cars that just pull in and out of a parking space.
Super-Cruise is limited to measured and tested highways. And, yes, I do think that is a better system at least as a start down this path. It also is much more proactive in its monitoring of the driver, even in inadvertent distraction (looking down too long, falling asleep, etc).
As you, unsure about the legal cases.
|
The Summon feature is interesting because the car doesn't have a driver like a Level 4/5 car, yet it is clearly not a level 4/5 vehicle by definition. Assessing guilt in an accident is never black and white, so I don't know if there
needs to be clear lines who or what is at fault, even if we
want there to be clear lines who or what is at fault in any given situation. Sometimes it is a matter of assessing the unique nature of the situation. In the Summon example, clearly this creates the potential for confusion as to who is at fault, potentially, but maybe it could be teased out with vehicle data from video and information logs, or maybe it is the case that it can't, and responsibility is shared, or maybe both parties are exonerated. I feel like such things happen all the time in many different areas of life besides driving. I also think adding Tesla to the potential list of parties involved is no different than a three car crash where 0-3 drivers could be at fault, as well as, the manufacture if the problem stemmed from a faulty system like an airbag randomly going off or an accelerator pedal sticking or whatever. My point is this isn't that new or different.
This article is interesting. They tested hands-on reaction times with an engaged driver issued a warning compared to hands-off reaction times with a poorly engaged or typical driver. Times were worse. I actually rarely drive with cruise control on because of this very fact. If I am normally driving, and I need to stop suddenly, then as I lift off the accelerator the car starts decelerating and then it decelerates faster, as I press the brake, but during basic cruise control, my foot is on the floor like it would be on the accelerator, and I have to move it to the brake. The reaction time is similar, yet it is typically longer and clumsy because my foot is on the floor and not poised next to the brake, but the effect is that there is a more abrupt transition from throttle input to immediate braking, and there is slightly more distance traveled by the car. This is compounded if I was driving with my knee while on cruise control, or if I had a hand on the radio or shift knob, and the evasive maneuver requires two hands. I don't know if the delay is as bad as an extra two seconds, but it is definitely greater than normal. This article suggests that all systems will inherently be worse when the driver has their hands off the wheel, and especially if they are disengaged from facing the road. I don't know how sensitive Supercruise or Subaru's driver monitor system is at warning the driver to look at the road. With these systems coming around for driver assistance packages, it makes me wonder if they will use them all the time, even when the system isn't activated, so it can reduce distracted driving, or if these systems will be limited to hands-off driving or only when assistances are on.
https://www.caradvice.com.au/699131/...actions-study/