Blog Tesla Releases Data on Utah Autopilot Crash

Multiple factors involved. Removing any one of them perhaps wld have avoided the accident.

But even sober attentive drivers might not perceive a partially blocked lane until they were right up on it.

Fire trucks shouldn't partially block lanes. If the lane needs to be blocked they should clearly block the entire lane

 

Over reliance on a hands on driver aid - it is not an autonomous system. It would be interesting to see if this education problem happened when cruise control was first introduced for lack of experience with a new driving aid

 

Agreed, except that an emergency vehicle has its flashing lights on ... I give them extra room even if they are taking up 100% of the next lane ... so I don't have any sympathy with anyone running into one.

Unless the flashing lights are found to be ineffective? I think you have red flashing lights in USA (whereas Blue here). Maybe Blue is more effective (because not used for other aspects of lighting, e.g. [and what I consider to be utterly daft] having flashing-red-tail-light as turning indicators, instead of mandating use of orange for that.

I'm dyslexic, I have "slow processing", and whilst I'm not thick the plethora of marketing-department-originating designs and styles for rear-lights on cars worries me that I might be slow to interpret a situation in an emergency and "more conventional and more consistent" use of lights would help [people like me].

Early cruise-control that I had was not traffic-aware, so if approaching a slower vehicle I needed to disengage or brake (which also disengaged). Thus attention was needed more often, and of course I still had to steer so unlike stay-in-lane I couldn't take my eyes of the road [for long], so maybe that meant that early Cruisce Control had no / little impact [sorry!] on accidents? I don't remember negative news-headings at the time ...

I have a very strongly self-imposed rule for AP driving that I do not allow any distraction. It would be very easy to think that it is OK to just quickly read a text when my phone Beeps. I don't. And I always have a hand on the wheel - to the point where I don't understand people complaining about Nags and driving no AP with their hands on their lap. AP is so good, more than 99% of the time, that I think folk are lulled into false-sense-of-security. Maybe the car is only going to drive into a fatal gore-point/barrier once in 100/1,000 car ownerships ... I do not want to be that statistic, and I firmly believe that in such an extremely rare incident that the difference between hands-on-lap and hands-on-wheel might make the difference that keeps me alive. But my car has now done over 50,000 miles and mile after mile on AP without any scary moments, and that must make some people complacent enough to allow themselves to become distracted.

 

As I posted in another thread: Tesla sued: Woman wants $300k for crashing on Autopilot while reading phone | ZDNet. It seems to be the same woman.

 

you've pointed out many of my "suspected" reasons for why US auto deaths are 3 to 4 times higher by most any metric compared to other industrialized nations.

 

Interesting, thanks. Had never come across that, it was just my observation.

 

I too was previously unaware of that statistic. 3-4x seems a little overstated, though. Looking at List of countries by traffic-related death rate - Wikipedia and just comparing the US and the UK (because they happen to conveniently sort together alphabetically) I see that the US is indeed in that range per capita (3.7x). It's higher than the UK, but not by as much (2.5x) per vehicle. It's higher but by still less (2x) per mile driven. Looking at a few other industrialized nations, France and Germany are closer to the US. Japan actually has a greater death rate per mile driven, as does Spain.