Ask me about being a Traffic Engineer!

The Something Awful Forums > Discussion > Ask/Tell > Ask me about being a Traffic Engineer!

DTurtle: Apr 10, 2011

kapinga posted:

Jeez, where do you live that they have so many power outages to actually need signs? And that traffic is basically unfazed by the loss of signals? I live in hurricane country (where extended power outages can happen), and we definitely don't bother with fold-out signs.

Here in Germany every intersection with traffic lights also has normal right of way traffic signs. At night, main routes have the traffic light turned off, while the side routes have flashing yellow. This is just to add warning for the side routes. What counts are the signs.

# ¿ Aug 14, 2012 23:36

Adbot: ADBOT LOVES YOU

# ¿ Apr 30, 2024 07:40

DTurtle: Apr 10, 2011

Golbez posted:

Word is the reconstruction of the I-80/I-380 interchange in Iowa will cost $270 million and last 5 years. They're apparently convering it from a nasty cloverleaf into a turbine.

The cost, I get. But why the time? What is taking up the bulk of that time? I imagine part of it is "you can't shut down an interchange while you replace it all at once" but ... if they DID shut it down completely, and let's say the traffic magically disappeared into detours, how long WOULD it take to build it? Is it due to the time to cast the concrete, and they can't pre-work that stuff off-site?

I guess I just don't get road work timelines in general. vv

When they reworked the A40 here in Germany (one of the busiest highways here in Germany) in 2012 they actually went and closed it entirely for 3 months. Doing it conventionally would have probably taken 2 years. It went quite well and saved a lot of money.

Obviously, a lot of work was done to make sure everyone knew what would happen and what the alternative routes were and so on.

# ¿ Oct 29, 2017 14:40

DTurtle: Apr 10, 2011

fishmech posted:

Taking away the steering wheel does change that. How, exactly, are you proposing that human occupant operate the steering of the car without it, when the self-driving functionality is unable to run? Like you seem to really not be getting that if a couple important sensors break or get blocked off by something (road grime, snow, whatever) the car can no longer self drive, but a human could drive it with the appropriate controls.

...

There are no cars that are actually self-driving . Tesla's Model 3 can barely be manufactured and their supposed self-driving functionality literally has killed people in the past, as it's the same package as on the Model S and X. You need far more than the sensors used for driver assist technology to actually create a safe and working self-driving car that can handle conditions beyond "perfectly sunny day on those 50 blocks of Mountain View we've tested on for 5 years".

https://arstechnica.com/cars/2017/11/fully-driverless-cars-are-here/

quote:

Waymo is now confident enough in its technology to dispense with a safety driver. The company has released a video showing Waymo cars driving around the Phoenix area with no one in the driver's seat

...

At first, most of Waymo's driverless cars will have an employee in the back observing the vehicle's behavior. If something goes really wrong, they'll be able to push the "pull over" button to stop the car.

...

Last month, I asked�Ryan Harding, a spokesman for the Arizona Department of Transportation, if there were any legal barriers to launching a fully driverless self-driving taxi service in the state. "I'm not aware of any current law that would prohibit" fully driverless taxis on public streets, Harding said. "We don't have�a problem with that."

...

The leftmost button initiates a call to Waymo's customer service center. The second button locks and unlocks the doors. The third button causes the car to pull over�though Waymo says the car won't stop in an unsafe place, like in the middle of an intersection. The rightmost button tells the car to start the ride.

Waymo is already testing in Detroit.

It seems like reality has already solved those unsolvable problems you've brought up.

# ¿ Nov 22, 2017 19:35

DTurtle: Apr 10, 2011

fishmech posted:

Testing is way the gently caress away from being something that actually works in practice in all the conditions a normal driver will expect to experience. I'm not sure why this is so difficult for people to grasp.

Except that for now it does not have to work in all conditions a normal driver can experience. It has to be good enough to recognize when it is getting into a situation that it can't handle (ice, extreme rain, etc.) and be able to react safely - ie. pull off to the side of the road and call in.

Does Waymo currently have a product and solution that can completely replace owning a car? No. But they are able to provide a taxi service without a driver. It is currently location locked and probably weather dependent, but they have a working system that can be incrementally expanded and improved. That is a system that is becoming a reality right now. There are autonomous cars - without a driver at the wheel, or able to grab the wheel - driving on public roads right this second.

# ¿ Nov 22, 2017 23:50

DTurtle: Apr 10, 2011

Hippie Hedgehog posted:

Any opinions?

I'm not sure if this is a real product or just a pipe dream.

The circular saw spinning in a circle making a circular hole gave it away to me that it can't be real.

# ¿ Apr 29, 2018 22:37

DTurtle: Apr 10, 2011

pun pundit posted:

Being allowed to turn right on a red signal seems deeply wrong to me. In Norway it is not allowed. How common is it outside the US?

Here in Germany it is not generally allowed. However there are two ways that it is explicitly allowed.

The first one is with an additional green arrow traffic light. With this one, if you have a green arrow, there aren't any cars from other directions that can come into conflict with turning right.

The second one is with an additional green arrow sign. With this one, you have to make sure no cars are coming, but you can turn on red.

There are more and more intersections that have signs of traffic lights like this, but the vast majority don't have that. And obviously they are only added where it is deemed safe.

# ¿ Mar 31, 2020 11:34

DTurtle: Apr 10, 2011

I�m so, so happy that speed bumps aren�t reaaly a thing here in Germany. If a certain spot needs slowing down, then you narrow the road or build a chicane. Much more effective and mostly independent of the type of car. It can probably even be cheaper, easier, faster, and nicer if you for example do the narrowing with large concrete planters.

DTurtle fucked around with this message at 06:33 on Apr 12, 2024

# ¿ Apr 12, 2024 01:32

DTurtle: Apr 10, 2011

TooMuchAbstraction posted:

For the record, this is a suburban street, 1 lane each direction, no shoulder (going directly onto either peoples' yards or forest, depending). I can't imagine that a speedbump would be more expensive than any of the options you described, especially in the case that you need to retrofit the road to slow down traffic, instead of designing it in from the get go.

A speedbump is never going to slow down traffic in general on a whole street, but only at a certain spot on a street.

This is the type of thing I�m talking about :

quote:

Which isn't to defend the roads around here, to be clear, they're dumb as poo poo. drat near everything is a residential road with lots of hidden driveways and terrible visibility (curves, hills, trees), and people are doing 40MPH on these things because they're the only way to get anywhere. If we were to go back to the drawing board and make a concerted effort to make the way people get around here make sense, then things would look vastly different.

Well, do you really need a two lane street? Is it two lanes + space for on street parking? Narrowing it down to a single lane + on street parking switching from one side to the other every 50 or so yards could help.

PittTheElder posted:

How do y'all handle widths required for ambulances and firetrucks? Because yea those things are vastly more effective (assuming you don't do the idiotic American thing and narrow the road with flexiposts), but curious if first responders need to have a database of where they are or something.

Although for all I know you could put a narrowing large enough for those things and it'd still intimidate people into slowing down, driving skills over here are... poor.

A firetruck or ambulance isn�t as wide as two cars. You narrow it down so that two cars don�t fit past each other, but rescue vehicles can fit.

Blue Footed Booby posted:

I'm curious how much you think a speed bump costs vs reengineering the entire road with chicanes.

See the picture above.

Also, I have to correct myself: there are speedbumps in Germany, but they are usually quite rare.

DTurtle fucked around with this message at 06:34 on Apr 12, 2024

# ¿ Apr 12, 2024 06:31

DTurtle: Apr 10, 2011

TooMuchAbstraction posted:

It is one lane each way, no shoulder. I found it on Google Maps:

There's some scope for making it narrower without pushing vehicles over the centerline, I guess. I think with the way local traffic behaves, though, you'd end up with people giving your obstructions a wide berth in favor of going over the centerline at high speed. Which would usually work fine, as it's a low-traffic road...buuuuut if there is a car coming the other way...

That is two lanes. You only need one. No middle line. Just wide enough that two cars can barely pass each other.

No one is driving 40 MPH on that street.

That is just a random street in a purely residential area of a small town

Here's another one:

DTurtle fucked around with this message at 12:37 on Apr 12, 2024

# ¿ Apr 12, 2024 12:31

Adbot: ADBOT LOVES YOU

# ¿ Apr 30, 2024 07:40

DTurtle: Apr 10, 2011

TooMuchAbstraction posted:

Ahh, I see, thank you for the detailed explanation.

The big thing to understand is that you want to design the street/road in such a way that people inherently drive the speed you want them to.

However, that requires you to think about what speeds you want people to drive. Is the priority getting people from A to B as quickly as possible? Or is the emphasis on making the street safe for cars, bicycles, children playing on it?

There are tons of different ways to achieve that goal. Some of them very easy and simple to implement, others requiring a lot more investment. The thing is that you need to recognize that some improvement, without being a perfect solution, is still better than the status quo. Perfect is the enemy pf good enough. Hell, Germany is still a very car centric society, it is quite far away from really being a good example - but still better than most of the US. You don�t need Dutch street design as a first step.

Strong Towns is an American/Canadian NGO focussing on implementing those kinds of things. You might want to look into them if you are really interested in actively changing stuff in your town.

DTurtle fucked around with this message at 20:21 on Apr 12, 2024

# ¿ Apr 12, 2024 18:48

The Something Awful Forums > Discussion > Ask/Tell > Ask me about being a Traffic Engineer!