Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

>If a pedestrian suddenly rushes out into the street or a cyclist swerves into your path, the deciding factor is often simply the coefficient of friction between your tyres and the road.

This only true for the uninitiated, never let it get to that. I drove for Uber/Lyft/Via in New York city so I can experience and study these situation. These sort of accidents are preventable. The following is the basic:

1.) Drive much slower in areas where a pedestrian and cyclist can suddenly cross your past.

2.) Know the danger zone. Before people "jump into traffic" or a cyclist swerve in front of you, they have to get into position, this position is the danger zone.

3.) Extra diligence in surveying the area/danger zone to predict a potential accident.

4.) Make up for the reduce speed by using highways and parkways as much as possible.

It helps that Manhattan street traffic tends to be very slow to begin with. Ideally I will like use my knowledge to offer a service to help train autonomous vehicles to deal with these situation. It has to be simulated numerous times in a closed circuit for the machine to learn what I've learn intuitively driving professionally in NYC.




> 1.) Drive much slower in areas where a pedestrian and cyclist can suddenly cross your past.

So essentially everywhere? I have seen pedestrians walk unexpectedly into traffic on high-speed 4-lane divided roads with no crosswalks.

> 2.) Know the danger zone. Before people "jump into traffic" or a cyclist swerve in front of you, they have to get into position, this position is the danger zone.

What "position" are you referring to? There are places where you can account for or predict pedestrians. There are also places where you cannot, such as when someone walks into traffic from behind a tall parked vehicle, where you have no chance to see them in advance.


Simply counting traffic fatalities suggests that crazy pedestrians causing unavoidable accidents cannot be common, even if every single pedestrian accident were both unavoidable and the pedestrians "fault" (although I'd argue that ethically it must be primarily the vehicles fault, but that's another story).

And that'ss with humans behind the wheel: lazy, distracted, slow to react fleshbags that we are.

I'm not sure that a truly unavoidable accident would occur even once a year in a fictional world in which all drivers were perfect and had millisecond reaction speeds.

What is obvious however, is that these situations are so rare as to be irrelevant. In practice accidents are avoidable by the driver of the vehicle - or at least avoidable to such an extent that it's not worth considering the other cases.

Also: although I personally don't object to some rational victim blaming I think it's a little distasteful that we're already speculating about how this must be the victims's fault, when there's simply not enough evidence to make that kind of determination yet. Let's not forget that part of the privilege of being allowed to participate in traffic implies a responsibility not to kill people even when they behave unexpectedly.

For some statistical perspective: if human drivers had as many fatal accidents per mile as uber has, then the average male driver would kill 1 person in his lifetime (men drive more). Clearly: that's absurd; people may cause too many accidents, but not nearly that many - and that's being rather charitable to uber's self-driving vehicles, since they have back-up drivers and thus don't count complicated traffic situations, and safety drivers, and thus may well have caused more accidents by themselves. So going purely by the unusual-ness of such an accident with so few miles, I'd say the initial assumption must be that this is likely a bug in uber's car, even if I'm sure there were contributing factors.

Edit: I guess it's not surprising wikipedia has stats on the influence of alcohol on fatalities, but it tops out at 4 times the legal limit - at which point human drivers are still safer than this (sample size of one...) uber record so far. :-/


> Simply counting traffic fatalities suggests that crazy pedestrians causing unavoidable accidents cannot be common

Of course not. I'm not saying they're common, merely that they exist. I do not agree with the idea that accidents would go away if drivers were just trying harder, though. There are legitimate unavoidable accidents, and there are also limits to practical human driving.

> Also: although I personally don't object to some rational victim blaming I think it's a little distasteful that we're already speculating about how this must be the victims's fault

To be really clear, I am not blaming the victim here. I have no idea what happened. I'm actually very inclined to blame Uber, though I recognize that's just my personal bias against them.


Why can't you slow down when driving past tall parked vehicles that you can't see through? If you are going slower you will have a chance to see them in advance, select a speed where your stopping distance is less than the length of the obstruction and be prepared to brake. You will have no chance of striking all but the most willfully suicidal of pedestrians.

Near my house there is an arterial I often cross where a large bush on the corner of the intersection obscures the view to the left from the stop sign roughly 10 feet back. I don't just stop at the sign then YOLO through the intersection, I stop, then creep forward, first looking for pedestrians or bicyclist that might step out from the bushes, then at the edge of bushes I stop again and look again both left and right for cars and bikes or pedestrians in the far lane before proceeding across. I often have to stop another time between the sign and the bushes for the pedestrian or bicyclist that just emerged, if I was focusing on getting across the arterial and beating cross traffic I would have killed every one of them.


At what speed do tall vehicles suddenly become transparent?

The idea that cars should drive past tall vehicles at 5 mph is slightly ridiculous. I have never ridden in a car with someone who constantly adjusted their driving speed based on the cars parked on the road. Choosing a reasonable speed? Of course. Extra care at obvious obstructions? Absolutely. Slowing traffic to 5mph because a van happens to be parked? No.


It is not ridiculous if you are driving in a narrow lane close to the obstacles. When I am driving on a residential (one lane, parked cars on both sides) street and I am approaching a tall opaque van I start by noticing if anyone is around it as I approach, when I have better sight lines, then as I get near it I absolutely slow down from an average speed of 15-20mph to well under 10mph and yes, depending on the situation sometimes as low as 5mph or slower. When I am on a faster road with two or more lanes there is often room to move laterally in these situations and so my speed reduction is less extreme, but I absolutely do still slow down as I pass these vans/trucks, cover the brakes and also check my blind spot to see if there is room for evasive maneuvers should they be required. If I see people in the area as I approach and have reason to believe they might try to cross, enter or exit a car, or otherwise be near the lane of traffic I will often change lanes to the left if possible.


While what you are saying can reduce the number of occurrences, it does not change what your parent post said.


You're making an observation about what it's like in an urban area. Sadly, in a suburban area like Tempe the presence of any pedestrian anywhere is unexpected.


The ASU campus area around Tempe has diverse traffic conditions. There are spots near the campus where you would expect to stop at every crosswalk, while in some roadways, a pedestrian would be completely unexpected.

Around the bars, people stumble into traffic all the time. Any driver, automated or human, would need to anticipate that.




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: