Cities to Untangle Traffic Snarls, with Help from Alphabet Unit | Hacker News Reader

Cities to Untangle Traffic Snarls, with Help from Alphabet Unit

nytimes.com

38 pointsph0rque20 comments

First time I've seen a headline mention Alphabet, and it feels forced -- like either Alphabet PR demanded it in exchange for the exclusive, or NYT editors felt they couldn't get through the lede without referencing the connection to Google.

Sounds like total nonsense.

More capacity ⇛ more traffic.

Example reference, if one is even needed http://www.citylab.com/commute/2015/11/californias-dot-admit...

Not hearing that that is part of the thought process here.

Thing is, there is a sense of entitlement which is just not realistic with billions of peoples.

(And down-voting doesn't change that ...)

Spot on. The constant in the equation isn't the number of cars on the road, but the tolerance people have for time spent in traffic.

Hey that's a good way to put it!

If effectively used, this seems like a fantastic use for this data. I think everyone's sat at a light and gone "damn, why don't they do X to make this more efficient?". The reason why might just be that they don't know about that particular problem :-)

>> I think everyone's sat at a light and gone "damn, why don't they do X to make this more efficient?"

After doing that countless times, I'm convinced that our roadways are laid out all wrong. The average speed on the EPA city drive cycle is just shy of 20mph. The reason for that is in large part due to stopping or waiting.

Think about this, to be efficient you probably want easy access to the routes you're going to be using soon. The most obvious road you'd want is the one in front of you, followed by branches to the left or right. The LAST direction you're going to want to go is back the way you came from, and yet we place traffic in that direction right next to us. Now if you want to make a left turn (right in the UK) you have to stop and wait to cross that traffic. Yes, this is an argument for one-way streets.

I've been pondering ways to cover large areas with one-way streets that are still efficient. If you can design things with no conventional intersections (or roundabouts) but with lots of branching and merging type things I think it can be a win. Even if your route is 10-20 percent longer, if you can drive more than twice as fast on average it's going to be a win. And with higher average speeds you'll spend less time on the road, and that means fewer cars on the road at any given time and a reduction in the number of lanes needed to support traffic.

Unfortunately we'd really need to design a city from scratch to get the best efficiency. And there would need to be rules enforced - like not putting shopping centers right at the intersections - that may be unpopular.

That would totally suck for pedestrians and cyclists. Optimizing for cars and not for pedestrians and cyclists is what gave us the awful suburbs of America. No thanks.

Maybe, but the reality is that the vast majority of street use is by automobiles. Optimizing for that massive use case is only logical.

>> That would totally suck for pedestrians and cyclists. Optimizing for cars and not for pedestrians and cyclists is what gave us the awful suburbs of America. No thanks.

One thing I keep coming back to is a separation of pedestrian and vehicle traffic. The idea of having sidewalks right next to the roads is just as stupid as having opposite direction traffic side by side.

Imagine a long high density urban area with pedestrian areas down the middle, and roadways around the outside. This is speculative of course, but the idea is to start with high level concepts like: keep the people away from the cars, make sure traffic doesn't cross, use only high density construction (because it makes public transit work). You get lots of conflicts, but I think we could do a lot better than what exists today.

> separation of pedestrian and vehicle traffic

I have an anecdote to support this. The bike path to my place of work runs along a river, far from traffic. There is a wide area for pedestrians as well. The difference between riding on this (think peaceful) and riding in a bike lane on the shoulder (life threatening) couldn't be any larger.

Some cities in the Netherland does it like this. There's high speed ring roads around the city, but in the city it's a maze of one way streets and dead ends. Basically, if you are in the city and want to drive to another part of the city you have to drive out on to the ring road and in again from another angle. Pedestrians, bicyclists and public transport can move freely inside the city.

This discourages driving within city limits, and encourages bicycling, walking and public transport.

This would never happen from a commercial real estate standpoint. Commercial RE development shapes a lot (all?) of urban areas and the most valuable real estate is almost always in the highest traffic areas. Restaurants want to be at the busiest most congested intersections.

Why? Traffic drives away pedestrians. Restaurants should want to be close to public transport hubs and other spots with high pedestrian throughput. More people will see your restaurant, and it's easier for a pedestrian to enter a restaurant then a driver in a car three lanes over ...

And my local Walmart has a driveway not 100 yards from a freeway exit. They want easy access in high traffic areas, but that doesn't mean we have to give it to them. The congestion caused by this is really bad. One concept I throw around is that all roads have a numeric classification, and we don't allow intersections between things that differ in classification by more than 1 level. This puts all driveways (business included) on side streets. And then disallow intersections being closer that a certain distance along a road based on its classification. This forces a hierarchy that at least separates things to reduce congestion.

Every morning I sit at a few red lights where I watch zero cars cross the intersection. It's infuriating.

Related: Strava (a popular mobile app used by fitness enthusiasts, especially bicyclists) also tries to do something similar with their Metro program: http://metro.strava.com/

That's a very cool system. I suspect that Vanhawks (http://www.vanhawks.com) will attempt to do something similar, though they're starting with a smaller data set and just started shipping bikes a couple months ago.

This is really coincidental.. last night I was sitting at a traffic light looking at my phone and started thinking that google maps data about car traffic would be really good to use to fine tune traffic lights.

I predict in about 15 years drones will cover the cities for various purposes including real time traffic and accident monitoring -- smart phone data is helpful but drones get you deeper.

Reply on news.ycombinator.com