Fatal autopilot crash, NHTSA investigating...

[stopcrazypp]

If the sensors see an object in the road that is at windshield level height, Tesla should not create software that tells the vehicle to ignore what it sees. Plain and simple.

Read up on how the Mobileye system works and how radar guided ACC works in general. Your statement shows a fundamental ignorance on how these systems work. Long story short, such software does need to ignore certain things the sensor sees because of false positives (objects, parked cars, walls on the side of the road, tunnels, hills, etc). Otherwise it'll be constantly slamming on the brakes in every road that is not relatively flat and straight!

Exclusive: The Tesla AutoPilot - An In-Depth Look At The Technology Behind the Engineering Marvel
AutoSpeed - Technology of Adaptive Cruise Control

 

[josephhearne]

It seems quite simple, a trunk cuts in front of a car, the car hits the truck. It doesn't matter who or what is operating the car, this accident was the truck drivers fault.

 

[Rashomon]

On the Tesla airbag not deploying: it wouldn't be expected to. Most cars have either one deceleration sensor at the airbag module, usually near the center of the car, or that sensor and another one or two up front on the radiator support structure. The airbag module has to take the information from those sensors and determine whether to deploy in the first 20 milliseconds or so of a crash. In an under-ride such as this one that shears the upper structure of the car away, the module is going to see a very long, relatively soft g-pulse from the sensors. If the pulse is below a deployment threshold for too long, the module is programmed to never deploy the airbag, because the relatively long deceleration is likely to have driver and front seat passenger out of position, and too close to the airbags. The worst possible case is if the airbags don't fill the space between occupants and structure and then collapse in a controlled manner, but instead push too-close occupants back, adding an additional acceleration to the occupants. You also see this occur at times with multiple impacts, as when a vehicle plows through some middling-large saplings before hitting a large tree; that may also result in a non-deployment.

 

[SR22pilot]

I've never seen an overhead highway sign that is 8-12' above the road. They're all substantially higher, because, you know, semi trucks have to fit beneath them.



Many of the fanboys on here are trying to give Tesla both ways.

If the sensors see an object in the road that is at windshield level height, Tesla should not create software that tells the vehicle to ignore what it sees. Plain and simple.

Based on the wording in you post and your join date I'll make a wild guess that you are a short who joined TMC solely to trash Tesla. You certainly haven't read my earlier posts in this thread where I discussed beam spread vs. detection distance. Go back and read my posts. Your question is answered in detail. This is NOT unique to Tesla so if you are truly concerned over this then please rant on Mercedes, BMW and other forums too. As things like scanning LIDAR and/or phased array radar become cheaper and more prevalent this will be less of an issue. Better camera systems will also help. MobilEye says they will have lateral intrusion capability around 2018.

 

[CuriousG]

And, for the record, McDonalds didn't change the temp of the coffee even after the lawsuit. Just like autopilot will (and should) continue to be available for people to use - with caution. If you don't use caution then unfortunate things can happen.

Are you sure about that? I'm not a coffee drinker but I'm using my own empirical evidence based on McDonald's apple pies. As a kid I remember the filling used to be scalding hot. Somewhere along the line (maybe they changed recipes) the apple pies are just warm now. Granted not exactly scientific but my Spidey sense tells me the McDonald's coffee lawsuit outcome played a role.

 

[Pdub2015]

The comparison is not very useful and even possibly misleading.

For it to be more useful, it should compare AP-driven miles to ONLY miles that AP can safely handle instead of comparing it to all vehicle miles traveled. As most know here, there are many limitations of AP making it unsafe to use in those areas/conditions.

If AP was always turned on w/its current limitations, it would surely have much worse accident rate.

The Tesla press release states that this is the only fatality in 130 million miles driven with AP enabled. It is NOT comparing it to "all vehicle miles traveled" as you state. How is that misleading?

 

[Nicoletta]

This accident would not have happened if the truck was also on autopilot. 99% of accidents are the result of human distraction, carelessness and stupidity.

 

[InTheShadows]

Never mind Tesla vs Mercedes AEB systems, It would be interesting to see if any manufacturers at all have an AEB system that can recognize a big semi trailer like this straddled across your freeway lane. I suspect none of them as the forward radar is likely set at the correct height to detect upcoming obstacles at car/SUV bumper height, and the air gap under the trailers is enormous - (probably up to 10m wide and over 1m tall?) - radar or lidar could just pass right through umderneath the trailer body

As I have said before NHTSA needs to mandate proper side underride protection on all these trailers.Would potentially save hundreds of lives each year
View attachment 183697

I think the problem with that is the low clearance when going over large humps.

I love the idea though, and the first thing that comes to mind is the side fairings that many trucks use now to increase their aerodynamics. Maybe they should be a requirement. They would help improve fuel efficiency and protect other drivers with advanced braking systems. Since the fairings are much closer to the ground, the braking systems should detect them.

 

[cwerdna]

The Tesla press release states that this is the only fatality in 130 million miles driven with AP enabled. It is NOT comparing it to "all vehicle miles traveled" as you state. How is that misleading?

Incorrect.

A Tragic Loss says
"This is the first known fatality in just over 130 million miles where Autopilot was activated. Among all vehicles in the US, there is a fatality every 94 million miles. Worldwide, there is a fatality approximately every 60 million miles."

I see no separation of the latter 2 stats to ONLY include miles driven where AP can be safely used.

 

[Pdub2015]

@cwerdna : Okay, I finally see what you're saying. Sorry about that. Thanks for your patience.

 

[Mike Smith]

It turns out the Tesla driver had received 8 speeding tickets in the past 6 years, and was knows as a speed demon to his friends. The owner of the truck (who owned his own company and was the sole driver), had receieved 7 violations in the past 2 years.

source:
The Latest: Truck firm in Tesla crash had safety violations

It's starting to look more and more likely that speed and bad driving led to an accident that no human or machine could have prevented.

 

[Brian May]

Two points on the topic of Harry Potter and the DVD player:

• He could have been playing Harry Potter Music on the sound system, and the witness assumed it was a DVD. Alternatively it could have been an audio only version of Harry Potter, or a radio station discussing Harry Potter, etc.
• Even if he was playing Harry Potter on the DVD player, that does not necessarily lead to the conclusion he was watching it at the time of the accident. It is possible he was listening only.

Sure listening could be a distraction too, however I don't hear many people trying to ban music systems from cars.

Some people in this thread seem to assume that the car driver didn't have enough time to slow down, and as such it was the truck driver's fault. Other people seem to assume the car had plenty of time to observe a large truck and slow down, and as such it was proof that he wasn't paying attention. However all of this seems to be pure speculation, we don't know for sure based on the information we have.

If he was speeding this could explain why he didn't have enough time to slow down even though he may have had enough distance to slow down (if he was doing the speed limit).

 

[aus]

It seems quite simple, a trunk cuts in front of a car, the car hits the truck. It doesn't matter who or what is operating the car, this accident was the truck drivers fault.

Tell that to the MS drive who's head is stuck in the side of the truck trailer. ;-)

Two points on the topic of Harry Potter and the DVD player:

• He could have been playing Harry Potter Music on the sound system, and the witness assumed it was a DVD. Alternatively it could have been an audio only version of Harry Potter, or a radio station discussing Harry Potter, etc.
• Even if he was playing Harry Potter on the DVD player, that does not necessarily lead to the conclusion he was watching it at the time of the accident. It is possible he was listening only.

Sure listening could be a distraction too, however I don't hear many people trying to ban music systems from cars.

Some people in this thread seem to assume that the car driver didn't have enough time to slow down, and as such it was the truck driver's fault. Other people seem to assume the car had plenty of time to observe a large truck and slow down, and as such it was proof that he wasn't paying attention. However all of this seems to be pure speculation, we don't know for sure based on the information we have.

If he was speeding this could explain why he didn't have enough time to slow down even though he may have had enough distance to slow down (if he was doing the speed limit).

If he was paying attention, do you really think he wouldn't see an 18 wheel truck across the road from 1/2 mile away? It's not like the truck could quickly cut across the road, and then suddenly come to a stop across the road... unless he ran into something strong enough to stop a truck.

My GUESS is, he was watching the movie on a phone or tablet and runing the sound through thespeakers via BT since the sound from a phone or tablet is pretty weak.
.

 

[Tam]

I...A Tragic Loss says
"This is the first known fatality in just over 130 million miles where Autopilot was activated. Among all vehicles in the US, there is a fatality every 94 million miles. Worldwide, there is a fatality approximately every 60 million miles."

I see no separation of the latter 2 stats to ONLY include miles driven where AP can be safely used.

I am not sure it's possible to compare Tesla Autopilot miles to other non-Tesla Autopilot cars.

How do other cars track their Autopilot mileage when they don't have over-the-air logging capacity?

When did other manufacturers offer Autopilot even?

I am not sure whether they would offer Autopilot if Tesla didn't offer the current beta program.

I just look up Volvo and they will offer IntelliSafe Autopilot but not now.

 

[GoTslaGo]

What about Tesla's total fatality for total miles (AP activated and non-ap activated)?

Still wouldn't answer @cwerdna's valid point, but it would be an interesting comparison index to total fatalities in the US (1/94 million miles). It also would be interesting to see what AP usage is in Tesla's fleet.

 

[Mike Smith]

There's a big difference between causing a crash and failing to prevent a crash. It's not as if the Tesla swerved into the truck. The description of the accident suggests this was driver error and it would have been physically impossible for the Tesla to avoid the collision. The fact that the Tesla driver had received 8 speeding tickets in 6 years, and the truck driver 7 violations in 2 years lends support to this being true.
The Latest: Truck firm in Tesla crash had safety violations

 

[Canuck]

And, for the record, McDonalds didn't change the temp of the coffee even after the lawsuit.

Are you sure about that? I'm not a coffee drinker but I'm using my own empirical evidence based on McDonald's apple pies. As a kid I remember the filling used to be scalding hot. Somewhere along the line (maybe they changed recipes) the apple pies are just warm now. Granted not exactly scientific but my Spidey sense tells me the McDonald's coffee lawsuit outcome played a role.

It depends on the franchisee... "the company now tells franchisees to serve coffee 10 degrees cooler than it did at the time Liebeck was burned." If franchisees fail to follow that direction, they do so at their own peril. McDonald's corporate is trying to protect itself from future liability with its direction.

But that's not the point. What bothers me is that it seems the McDonald's hot coffee case comes up every few months in a thread here in an attempt to establish some point that really has nothing to do with the case at all (right Bonnie?). I'd much prefer that if people don't want to live in a society that guides itself by the Donoghue v Stevenson principle of tort law, that people just say that. Please also tell us why people and corporations owing a reasonable duty of care to others in our society is such a bad thing?

Of course, the law in not perfect, but travel in many third world countries where no one owes anyone a duty of care, and see if you'd prefer to live in that type of society. Our common law duty of care has established itself for a reason. It may not be perfect but we need to draw lines somewhere and I, for one, like where we have drawn them, and how our law has evolved to the point it is today.

It wasn't all that long ago that we had to huddle around fires to keep warm and lawlessness was the rule. Maybe we can evolve further and develop better principles than the seminal Donahue case, but rather than trash the ones we have established with false commentary on case-law, tell us how you propose to make us better.

I'm all ears.

 

[wdolson]

If he was paying attention, do you really think he wouldn't see an 18 wheel truck across the road from 1/2 mile away? It's not like the truck could quickly cut across the road, and then suddenly come to a stop across the road... unless he ran into something strong enough to stop a truck.

My GUESS is, he was watching the movie on a phone or tablet and runing the sound through thespeakers via BT since the sound from a phone or tablet is pretty weak.
.

It's also quite possible the driver fell asleep. It appears he was quite used to AP and he may have perfected a way to doze when on AP and just wake up enough to touch the wheel when the car chimed.

 

[Ben W]

Year 2018 seems like a long time. A lot of life can be saved (for all car manufacturers) if LTAP can be implemented earlier.

There are already millions of cars on the road that have non-LTAP TACC that cannot be realistically retrofitted. However, existing trucks with trailers could easily be retrofitted. (with side guards and skirts.) I'll take easy over unrealistic anyday.

 

[Ben W]

I look at it this way: taking your eyes off the road while using AP is sort of like taking your seatbelt off in a car with airbags. Sure, you can get away with it now and then, even most of the time, but it's still a dumb and dangerous thing to do. (even once.) Just as the airbag enhances safety but doesn't remove the need to stay buckled in, Autopilot enhances safety but doesn't remove the need to keep eyes on the road. Drivers ignore these respective mindsets at their peril.