To be fair, even the photo sensors should have been able to pick it up in this case

At least pick up the CLOSED barrier.

At least for my elderly brain - in low-visibility conditions, the red flashing lights of an "active" railroad crossing signal are far more distinctive and attention-grabbing than the very thin, kinda-fog-colored gates.

I believe it to be more common in the EU to have dangly bits (metal fencing?) that hangs from the barrier to the floor. when the barrier lifts up they 'fold' vertical. This presents a proper barrier (that stops kids, cyclists, pedestrians from getting under the barrier when deployed), the 'struts' are painted red and white so they're very visible. *can't speak for absolutely everywhere, but my ride to work there are two sets of tracks I pass that have barriers that are designed to stop the stupid. Oh, I guess self-driving vehicles, too, now.

Lidar can't handle fog or snow. It's a research area for improvements: https://www.cts.umn.edu/news/2023/april/lidar

Humans can't handle fog or snow either.

watch the video, a person can easily see the train, gate and lights before they disengage. Also a person would have heard the crossing noises, and the train horn.

From what we see in the video the driver was paying attention: he hit the brakes and turned the car to avoid hitting the train.

Now of course he did do it "too late" to avoid any collision, but at what point are we expecting an attentive human driver to realise that their self-driving care has decided not to stop at an oncoming train and make the split-second decision to intervene?

If we expect drivers to mistrust their allegedly self-driving vehicles to the point that they must be ready to anticipate the car failing to react to an oncoming obstacle and immediately intervene, then this is not self-driving technology — it is human driving with extra steps.

>If we expect drivers to mistrust their allegedly self-driving vehicles to the point that they must be ready to anticipate the car failing to react to an oncoming obstacle and immediately intervene, then this is not self-driving technology — it is human driving with extra steps.

Well, it is not self driving. Anyone who uses FSD Beta is beta-testing it and should be paying the same amount of attention as if they were driving themselves. It is human driving with extra steps.

(I use Tesla autopilot, don't have FSD.)

> Well, it is not self driving. Anyone who uses FSD Beta is beta-testing it and should be paying the same amount of attention as if they were driving themselves.

Putting aside that the initials "FSD" stand for "Full Self-Driving", the level of attention of ordinary human driving is not enough. If that is the case, FSD Beta is accidents waiting to happen and should be removed from public roads.

Human driving goes beyond the factual realisation that there is a train crossing the road and the driver reasoning that they must push the brake. Rather, it is the driver having a constant feedback loop of information followed by intuitive reactions as to the action to take. Case in point, the cliché of having driven all they way from home to work and vice-versa and realising that one cannot remember the journey, even though the driver paid full attention to all rules and reacted appropriately to each obstacle. An attentive driver may even brake their vehicle when an object has flown into vision before having actually processed what the object was.

Compare that to the type of attention that would be necessary for "FSD Beta". The speed at which the car is travelling is disconnected from any immediate feedback loop in the driver's nervous system. The driver sees an obstacle in the road, one of many over the past few weeks which required no intervention given that the car reacted appropriately. Here is another obstacle. Has the car seen this one? If not, will the car see it soon? Will it react appropriately in time? Should the driver intervene now? That requires much greater attention and decision making than the intuitive process of human driving, all the more so considering that the majority of the time the occupant of the driver's seat in the "FSD" car is becoming accustomed that there's no need to intervene.

> the level of attention of ordinary human driving is not enough. If that is the case, FSD Beta is accidents waiting to happen and should be removed from public roads.

Respectfully, that's the point. People have been stating this clearly and unambiguously for years. People are not designed for that kind of split-second monitoring for hours on-end.

It felt like the driver should have hit the brakes a lot earlier, though.

I imagine driving a self driving car is like watching someone else driving as a passenger:

"oh look, a railroad crossing with a train crossing"

"we're getting close to the crossing, surely he noticed the train"

"why aren't we slowing down? oh crap he falls asleep at the wheel"

*grab the wheel and steer off the road*

As long as self driving cars require constant supervision, I think this kind of accidents will keep happening. Drivers will have reduced reaction time when they're not actively driving the car.

Indeed some silly people are really thinking "self driving" means "self driving", I only wonder why.

Maybe because users are being actively lied to about it, and not all are insiders to be able to see through the lies?

that's not the point. Commenter said a human couldn't handle fog/snow either, which, in this case, is objectively false. FSD failed when human vision and hearing did not. Maybe we need more then cameras afterall?

I dunno if lidar would have helped here much since it was so foggy (unless fog is transparent to IR or something?).

Radar would, but radar is also a bit crap - in particular angular resolution is very very poor.

Radar would probably not, in this case. At the point where it crossed the road the train was not moving toward or away from the car, so radar would have to filter it out as a stationary object (stationary objects have to be aggressively filtered with radar to avoid false positives from metal signs or stopped cars near the road). Even if it somehow was detected as a moving object, sometimes trains stop in crossings, at which point it would definitely be a stationary object and radar would be useless.

Correct me if I'm wrong but one shouldn't crash their car into stationary objects either, right? Or I'm missing the point of your comment maybe?

You are correct. The point is radar in cars has really bad angular resolution so it can't distinguish stuff that's directly in front of you (like a stopped train on a crossing) from stuff that's just to the side (like signs). Since both signals appear identical to the radar, and you don't want your car to emergency brake every time you go past a sign, cars filter basically ignore stationary objects detected by radar.

That isn't what you want, but it's a limit of current commercially viable radar technology.

Fog (small water droplets suspended in air) is transparent to light but it can reflect it if the wavelength of light is smaller than the water droplet diameter. Since there are so many droplets it scatters light making lidar useless.

The usual diameter of fog droplets is in the micrometer range, but is highly variable.

So microwave/milimeterwave lidar could theoretically work.

A better rail crossing design would have helped. At such fog even human will miss this. Put a barrier further away, brighter lights, etc.