[userbinator]

Who cares if he had 5 seconds to see the barrier if he only had 0.5 seconds to realize that the car went into casual murder mode right as it veers into the barrier.

This video is quite relevant:

https://www.youtube.com/watch?v=6QCF8tVqM3I

A fully alert driver trying (and succeeding) in reproducing this behaviour. Observe how long the barrier is visible, how long it takes him to react, and how close the car was to hitting the barrier.

[brandonjm]

This video is horrifying, I assume even the sensors on cars with automatic braking sensors should be able to identify a barrier that large in front of the car. I am way oversimplifying this but I'm sure there is some level of prioritisation in Tesla's Autopilot software to decide what action the car should take. Is the Autopilot software prioritising staying in lane (the car appears to stick to the left white line) over collision avoidance?

[URSpider94]

It can’t, nor can most cars with automatic braking. The Tesla will happily run headfirst into a brick wall all day long. The reason is that the driving world is chock-full of objects with zero velocity relative to terrain - the trivial cases being a rise in the road ahead, or an at-grade bridge where the travel lanes suck below grade to pass. Therefore, autopilot and many auto-braking algorithms filter out completely static objects (see previous story about a Tesla ramming a fire truck stopped on the highway).

Musk has commented publicly before (though somewhat obliquely) about this flaw. He indicated that the company is trying to build a map of reference data so that it can be filtered out automatically, and real hazards can be seen/found.

[skywhopper]

Mapping can never solve this problem. If these cars don’t yet have the ability to detect stationary physical barriers that represent a crash risk, then they are further away from being practical than I thought, and I’m extremely pessimistic. It is clearly a hard problem to solve with naive tech. Unfortunately we have a huge industry and tens of billions of dollars and politicians, corporations, governments, and media who all believe this naive tech is close to perfect. But it can’t see a wall or a fire truck in its path? If this isn’t solved then this whole house of cards will fall apart. The sooner the better, IMO. Let’s start building cars that supplement driver awareness instead of numbing it.

[slivym]

This would be trivial to solve with a vertical mounted LIDAR. You vertically mount it and if the horzontal distance measured level with the bumper of the car are significantly closer than the horizontal distance at a lower angle you can classify that as a stationary object, if it's a continuous lengthening or shortening then it's a ramp. The classification is almost trivial (though the hardware may be expensive or difficult to implement for high speeds).

[azinman2]

But in the path of travel??!

[saalweachter]

Not to exonerate Tesla, but cars are ALWAYS almost about to hit something. The next time you’re driving down a curvy road, pay close attention to all the obstacles that are by the side of the road and how often you are pointed right at them and how small the steering adjustments you make are and how little time there is between you almost hitting something and then not.

This is part of why trying to second guess a driver or autopilot is insane; a crashing and non-crashing car are almost identical, except for the crash.

[axaxs]

True. But these aren't auto self driving cars. A self driving car should know its next move, 10 seconds before it makes it(or something similar). There's 0 excuse for it hitting a barrier in 0 traffic.

[hyperbole]

Tesla has likely built 80-90% of an autonomous car enough to fool those who use it into believing it can be trusted. But as with all engineering endeavors that suffer from the Pareto principal its going to take far longer to get close enough that it's better than a human whose paying attention.

Its foolish to oversell thru marketing the capabilities of AP, as it's fairly easy to conflate reliability with observed behavior.

As long as the system mimics enough of the capabilities of full autonomy and it's oversold as such customers will underestimate the risks and become victims of an unfinished product

[Latteland]

tesla's system is lane following which has a radar that knows when the cars in front of you slow down or speed up. that's pretty much it. it's not an autonomous car in general. mine is very reliable for cruise control adjustment, good but not awesome for lane detection (falls down in bad rain for example, works well in clear weather). it tells you this, via display, it knows if its working. I can't understand these fools who thinks its some super advanced system. it really does little more than follow lanes and adjust speed based on detection of cars in front.

[hyperbole]

Video from 2016 showing exactly that:

https://www.tesla.com/videos/autopilot-self-driving-hardware...

If you use this to market your car without any caveats, don't be dismissal of some people who might read into that your car is capable of self-driving, when in reality its clearly absent of anything of the sort.

[greglindahl]

Some Tesla owners are being dismissive of attempts to paint Tesla owners as being unaware of the actual capabilities of Autopilot, based on Tesla's marketing.

Owners see marketing occasionally. Owners see the reminder that they have to stay alert 100% of the time each time they turn on Autopilot. Owners get the alert that they need to put their hands back on the wheel. And so forth.

If you've got data about owners being confused, please share.

Meanwhile, next time an owner offers anecdata about what they personally have learned about Autopilot's capabilities, perhaps you could respond to that, instead of telling the owner that they're being dismissive.

[Latteland]

well probably fool was too strong when they call it "autopilot". once i explain to people (well, programmers) what the tesla system is, they go oh, that seems possible, tesla made it out like there was some magical ai running on the car.