[maxsilver]

The safest infrastructure are ones that split pedestrian / bicyclist / motor traffic as much as possible, for obvious reasons. But that infrastructure has fallen out of favor as of late. "Complete streets" are the new fashion trend, and are significantly more dangerous, by combining these forms of traffic altogether and just sort of hoping it all works out.

But, "complete" streets are super cheap, and give the illusion of improved infrastructure, so the trend will likely continue for the near future -- further increasing accidents as it does.

[admiral_biatch]

In my opinion isolating separate transportation modes is only good if it is complete isolation but that's never feasible in a city. Too expensive and space-inefficient to make every intersection collision-free for cars, bikes and pedestrians.

If you separate cars, bikes and pedestrians most of the time but their paths cross at intersections then you have a problem because drivers might not expect a sudden bike lane out of nowhere. It's better to have the bike lane on the street so that cars see it all the time. This makes them drive slower and more carefully because they expect bikes to show up there.

I don't have data to back this up although I vaguely remember reading about it in "Streetfight" by Janette Sadik-Khan. If I recall correctly introduction of unseparated bike lanes in New York City didn't increase bike fatalities despite increasing the number of bikers and it also decreased number of pedestrian fatalities thanks to cars driving slower because of bikes. Of course the article shows that now the pedestrian deaths increased so it might have been a premature conclusion on Sadik-Khans part.

[jnty]

Of course drivers in cities do expect 'bikes coming out of nowhere' because, unless you literally separate bikes with a 6ft wall, you can in fact still see them despite there being kerb-separation in place. At car/bike junctions with poor visibility the same interventions are available as for car junctions with poor visibility.

If you've ever seen a mangled barrier by the side of a road, you'll understand why encouraging humans into the road as a traffic calming strategy is rather problematic.

Enlightened cities such as those in the Netherlands tend to take a risk-elimination approach - residential streets will be designed to keep speeds low, and neighbourhoods designed so that through-traffic doesn't try to take shortcuts along them. This makes it safe for cyclists to use the roads without special infrastructure. Only busier main roads have infrastructure. This is hard to imagine in the US where many cities in the US seem unfamiliar with the concept of any road not being a busy main road!

[admiral_biatch]

Where I live (Poland) it's not uncommon for a bike lane to be effectively a part of the sidewalk with trees and parked cars between the sidewalk and the road so it's not that obvious if you're visible to drivers or not.

I agree that putting unprotected humans on the road should happen only after other traffic calming measures have been put in place.

Regarding your last remark. There are basically two styles of bike infrastructure. You described the dutch way quite well. But there is also a Copenhagen style of bike infrastructure where it's directly by the road. Sometimes separated by the curb but still on the road. https://goo.gl/maps/QWTfALXSbsj

I won't judge which style is better. It probably depends on the city.

More about differences between Amsterdam and Copenhagen styles https://robertweetman.wordpress.com/2017/11/04/amsterdam-vs-...

[Someone]

”but their paths cross at intersections then you have a problem because drivers might not expect a sudden bike lane out of nowhere.”

If you design your intersections correctly, cyclists do not appear out of nowhere, they always intersect at right angles with the car lane (https://bicycledutch.wordpress.com/2014/02/23/junction-desig...). That hugely increases visibility of cyclists for car drivers and vice versa, and makes eye contact possible.

[maxsilver]

> but that's never feasible in a city.

That's kind of funny to hear, because I was going to use real-world Chicago and Seoul as practical examples of this already existing in cities today.

Chicago Loop has a dual-layer approach (where faster moving cars are on the ground floor, but a "second street level" is directly above them, for pedestrians + buses). It's not an exact match (cars can drive on both levels, those lanes should all exist on one level), but it's pretty close to this idea already in practice.

On the opposite side, Seoul has a "pedestrian highway" slung above 8ish lanes of car traffic below, which is a cheaper (although less effective) version of the same idea.

[lostlogin]

Paris has it pretty good in the central areas. Wide roads with (outside in) footpaths, cycle lane, road, then a central section with tram, trees and garden or parking. Having massively wide roads helps.

[anonymou2]

The current traffic system is the best known way to share the roadways among different types of vehicles, all subject to the same rules. Trying to construct a set of grids for different types of vehicles which interact at certain points is silly, consider what happens when new types of vehicles come into being, like electric scooters, are you going to construct yet another "independent" grid for them?

[admiral_biatch]

I agree with you. I'm not in favor of separation precisely because it's unrealistic in the city.

[rconti]

Plus, separating all of this infrastructure takes space, which in turn makes everything less walkable. In cities, there really need to be fewer car lanes for it to work.

[junga]

>The safest infrastructure are ones that split pedestrian / bicyclist / motor traffic as much as possible, for obvious reasons.

Can you provide a source for this? (especially concerning the separation of bicyclists and motor traffic)

[maxsilver]

> Can you provide a source for this? (especially concerning the separation of bicyclists and motor traffic)

https://rspcb.safety.fhwa.dot.gov/Dashboard/Default.aspx

USDOT routinely finds freeways to be the safest roadway for all involved each year (lowest number of ped/bike/car fatalities), since it separates traffic modes.

Pedestrians and bicyclists go over/under/beside it on dedicated infrastructure, they are never on it. And similarly, cars are never on the dedicated ped/bike infrastructure, so collisions are effectively impossible from either side, due to physics.

[cglace]

If you completely separate two things they can not collide.

[admiral_biatch]

True, but it's not feasible to separate the traffic completely. Every intersection would have to be collision-free for cars, bikes and pedestrians. How much money and space would that require?

[anonymou2]

There will still be bike-bike collisions, also exclusive roads for bicycles are usually of low quality not up to engineering standards for roads, they will be full of dirt, unmaintained, with potholes. Most bicycle accidents are not even collisions with cars. The truth is, it is more convenient for motorists to remove cyclists from the road so they have them all exclusively for them, that is just not fair nor safer.

[loeg]

> There will still be bike-bike collisions

These are both more rare than car-bicycle collisions and less likely to be injurious.

> also exclusive roads for bicycles are usually of low quality not up to engineering standards for roads, they will be full of dirt, unmaintained, with potholes

Citation needed? The main problem I've experienced on bike paths is tree roots. Roads for bikes don't wear out particularly quickly, because bikes don't make anything like the kind of wear and tear on roads that cars or especially larger vehicles (buses and trucks) do. Road wear is proportional to weight squared.

> Most bicycle accidents are not even collisions with cars

What are they collisions with? Do you think bike-bike, bike-ped, and bike-fixed object collisions combined are more frequent than bike-car collisions? I don't have numbers on this one way or another. To provide some color, motor vehicle-fixed object collisions are about a third of all motor vehicle crashes[0].

Additionally, 95% of cyclist collision deaths are definitely car-bike collisions[1]. So even if they are not the majority of collisions, they must be disproportionately deadly and worth consideration anyway.

[0]: https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/... p. 18 (table 5(b)).

[1]: https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/... p. 5-6.

[anonymou2]

No, cyclists fare best when they act and are treated as drivers of vehicles: http://www.johnforester.com/

[mytherin]

No, the safest option is to separate cycle lanes from roads altogether, as is evidenced by the significantly lower death toll per km that e.g. the NL and Denmark have compared to the US [1]. In those countries, cyclists almost never share the road with car drivers.

In roads that are not designed for cycling (where cyclists have to share the road with car drivers), perhaps it is true that they fare best when they act as vehicles. However, the safest option by far is not sharing the road with cars to begin with.

[1] https://www.forbes.com/sites/niallmccarthy/2015/02/24/the-mo...

[anonymou2]

You really can't compare the US and Denmark, you should pick a location and compare before and after segregation. Did it improve safety? There were plenty of bicycle trips in Denmark before cyclists were kicked off the roads for the convenience of motorists.

[lostlogin]

If it’s convenient for motorists that I get a segregated cycle lane, I’d be very happy to go along with it. Are there cyclists who don’t like cycle lanes and prefer to share with cars?

[kevin_thibedeau]

I don't like them when they aren't adjacent to the main lane. You end up outside the zone where drivers are looking for fast vehicles.

An improved shoulder wide enough to not have storm drains or debris to deal with is ideal. You can enter the roadway proper as needed to take the lane or prepare for turns like every other vehicle. You don't get abuse from ignorant drivers who expect you to stay in your designated space.

[loeg]

As a recreational cyclist, I wouldn't like dedicated cycle lanes that had a too-low (and enforced) speed limit, for example.

[anonymou2]

I do, I have been driving my bicycle to work for 12 years now. Not only I prefer regular roads, which are much better, whenever there is a segregated lane I avoid it. It is very obvious what segregation is for when you get hostile behavior on the part of motorists when you ignore the cycle lane and they complain that you don't use it.

[tropo]

Someday you will be old, unable to accelerate like a car and keep up with traffic, and you will instead wish that you could ride on the sidewalk. Part of the reason why bicycle use is unusual is that very few people can ride well enough to be out in the road.

[cr1895]

>cyclists almost never share the road with car drivers

That's not true...there's plenty of shared road. Drivers are more cautious and aware, but it's not all separate pathways.

[mcv]

Shared road is getting increasingly rare, though. Except in 30 kph zones where cars have to go slow anyway.

When I was a kid, there were still plenty of 60-80 kph roads where bikes had to ride on the right side of the car lane, but almost all of those have separate bike lanes or bike paths nowadays.

[ahoy]

This is plainly wrong. The biggest threat to a biker is being hit by an automobile. If that biker is in a lane completely away from automobile traffic, that threat vanishes.

[anonymou2]

Being hit how? What type of accidents are more likely to happen? A bike path with no interaction at all with cars will of course prevent any bike-car collision, there will still be bike-bike collisions and other types of accidents (the majority of bike accidents are not car-bike collisions anyway). That kind of exclusive bicycle path is delusional anyway, specially in a urban settings, where you always have to interact with the regular roadway. And it is there, in those interactions where most bike-car collisions happen, in fact most collisions in general happen in intersections. Segregating by vehicle type only makes intersections, which is were accidents happen, more complicated. It makes no sense to have an exterior lane being a thru lane when the adjacent lane is allowed to turn right, that is a contradiction to traffic engineering principles and it is what many bike lanes do.

[loeg]

> Being hit how? What type of accidents are more likely to happen?

95% of bicycle deaths are a result of motor-vehicle bike collision, and the vast majority of collisions are between the front of the motor vehicle and a bicycle.[0]

> there will still be bike-bike collisions and other types of accidents

Sure, but these collisions are far less deadly and injurious. Exchanging car-bike collisions for bike-bike collisions is a good tradeoff from a public health perspective.

[0]: https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/...

[anonymou2]

No, what is good for a public health perspective is to avoid collisions (and falls which is the #1 accident with bikes). It is quite easy to avoid car-bike collisions, in fact car drivers are far more predictable than ignorant cyclists. Of course in order to do that you have to know what the most frequent type of car-bike collisions are, guess what? they are not cars hitting a bicycle from behind, they happen in intersections due to crossing traffic and bike segregation only increases them by making intersections more complicated.

[loeg]

Your argument hinges upon the notion that separated grades increases incidence of car-bike collision per capita. I concede that if that that core fact is true, you're correct — separate grades would be worse from a public health standpoint.

So, do you have evidence for that central claim, or is it just speculation?

[rhinoceraptor]

Segregated bike paths, at some point, have to intersect with roads and driveways. So you can never eliminate the risk of accidents.

I can’t find any hard data on this, but from what I can see online (and my own experience riding in the road), the most common types of accidents are when cars are turning, and fail to yield to a cyclist. Being rear ended or sideswiped seems to be much more rare than accidents where the car is turning and doesn’t see the cyclist.

For example, a cyclist is riding in a bike lane, and a car is turning left into a driveway. The car is looking for oncoming traffic in the road, so the cyclist is most visible in the road or bike lane, where the driver is expecting fast moving traffic to be. If we build a segregated bike path, cyclists will be much less visible, and it’s more likely that a driver will fail to see them when turning.

[Doctor_Fegg]

Forester is widely discredited.

[anonymou2]

Thanks to his book I've been riding safely for 12 years, only those who did not make the effort to learn discredit him.

[wlesieutre]

What, you don't like to share a narrow lane with no shoulder right up against a concrete barrier that you'd better not go over because the other side is a 20 foot drop down to train tracks?

https://i.imgur.com/0cxvaJM.png

At least they put up a no parking sign.

EDIT: Not to mention this photo shows the one and only bike symbol on the block. Given how worn off it is I assume most drivers have no idea this is the "bike route."