Thursday 5 May 2011

State of the art bike way design - a further look

Please note that this blog post has caused considerable confusion. It does not recommend a design, and the sketch presented is not supposed to be a guide for other countries to use to design their traffic lights. The design shown here has in some circles become known as "a standard Dutch junction" but was not the intention of this blog post and is a myth. Other blog posts show the multitude of different designs of traffic light junctions which are really used in an average Dutch city.

Last month I wrote about the new NACTO designs for cycle infrastructure and held the junction design against Dutch junction design. The video that went with the post was taken out of the context of this blog and discussed on forums and other blogs. Without the context some people completely misunderstood it.

Right: Dutch design of an actual junction with cycle paths in red.
Left: the same ideas used on an American road to see if it could be done.
There were of course comments that you can’t really take seriously: “Yuck. That seems like a parody of over-engineering.”

And if someone dares to criticize anything American there is always this: “could be some anti-American bias there.”

Other comments were more serious but reveal a completely different frame of reference:
“I'm not convinced about the safety aspect. I think that this video exaggerates the danger of crossing at a narrow angle. It's just a lane change.”

‘Just a lane change’? That is an interesting way of seeing it. Indeed the whole object of this design is to eliminate just that action. Because a lane change IS a dangerous thing to have to do.

Some commenters were concerned about the remaining space for motorized traffic.
  • “large vehicles are to move closer to the center of the junction, possibly disturbing traffic flow”
  • [there] “appears to be a sharper turning radius for the Dutch motorist”
The radius is the same as for the conventional US junction. But yes, it appears to be sharper, thus making traffic go slower. One more advantage.

Others were not convinced about the advantages this design has for cyclists. Summarized they:
  • believe the ‘swerve’ to the right and then to the left is slowing the cyclists that want to go straight on down;
  • fear there would be conflict between cyclists going straight on and right turning motorized traffic;
  • fear they would have to wait twice (and long) to make a left turn.

One person on a forum asked if there were Dutch people who have experience with this type of design who could give their view. Others too seem to think this is new and experimental design that should be tested. Well it was… for about 50 years now. Almost all junctions in the Netherlands with separate cycle paths were built exactly like the schematic design in the first video.

Further explanation

This second video shows the design ‘in action’. A number of Dutch junctions showing a number of situations that perhaps shed some light on all the questions. Now you can also see how the various green phases work. Not only do cyclists and pedestrians have their own green phase. Left-turning motorized vehicles get a separate green light as well, when vehicles going straight on -on the same road- have a red light. While this may seem like a bad idea because you would have to wait longer for your own green light, it is in fact the reason for very fast movements on the junctions. Once you get green you can proceed without any waiting for other traffic users that might be crossing your path as is usual in other countries.

More modern solutions

It is interesting that while other people question whether this design could work the Dutch actually have moved on to more modern solutions. The ‘simultaneous green for all directions’ David showed again in his last post is one more modern approach. But an even more radical change is that a lot of the junctions are being transformed into roundabouts. It turns out that these can handle more traffic in a quicker and safer way without even needing traffic lights.

This is still a cross-roads junction in Google Earth...

...but it was already transformed into a more modern roundabout
as can be seen in Google Streetview
Click for many other posts about roundabout design as well as about traffic light junctions, but note that unravelling of cycle-routes from driving routes means that cyclists often avoid such junctions altogether.

A note from David
Please note that while Mark refers to a "standard" junction design in the text above, junctions which exactly resemble this design are not actually common in the Netherlands. Only one junction in Assen closely follows this design and that's already planned to be replaced with a more modern and more convenient (for cyclists) design as used in the rest of the city.

Update 2014: NACTO adopt a similar design
It appears that NACTO took these criticisms on the chin and are adopting a design of junction very similar to Mark's sketch above. Be careful that it is not over-applied. This needs to be one of a range of solutions, amongst better solutions, not the solution or the best solution alongside lesser solutions.


Severin said...

This may supplement your argument. Here is a video I made of a standard American bike lane in Los Angeles. IT shows all the 'benefits' the design offered.

Richard said...

The "Dutch" design works because the volume of cyclists is sufficient that giving them a separate green phase makes the junction more efficient.

CROW guidance (drawing V43) is for a 5m separation between the bicycle and traffic circulation carriageways.

Pjotr320 said...

I recently cycled 600km through Germany. Nice landscape, friendly people, good food, excellent Spezialradmesse.

But, one thing that really annoyed me was their junctions. Slow and complex for cyclist. Involving curbs to cross and several lights for one manoeuvre.

You don't know what you've got 'till it's gone.

Zmapper said...

That was me who posted the video out of context on AAroads. Sorry about that. And the I was trying to see if I could get some of the dutch members that frequent the forum to post their opinion on it, considering that they deal with these intersections daily.

Personally I like that type of treatment and would like to see an American city use it.

BTW this is the thread he was talking about:

Micheal Blue said...

Mark, thanks for the great video. Can you send some of that blue sky to us in Toronto? Well, we have it today, but it's supposed to go back to raining...

NE2 said...

I'll copy my entire response:

Yuck. That seems like a parody of over-engineering.

The main thing to remember when looking at Dutch designs is that the culture is different. If you hit a cyclist, you're automatically presumed to be at fault. Motorists are willing to accept shortening their signal phase length for a reasonable cyclist phase on a sidepath.

If that design were done in the US, where "I didn't see him" is a valid defense, you'd have right hooks out the ass. Try to cross on foot at a busy suburban intersection and see how many right-turners (and left-turners) fail to yield. Now imagine you're on a bike, 2-3 times walking speed, and less maneuverable than a pedestrian.

J.. said...

Thanks Zmapper for the link. Keep spreading the word. I've run into a lot of American cyclists online, who are absolutely committed to "vehicular cycling" and can't even imagine separated infrastructure being safe. It's very frustrating.

Frits B said...

NE2: "The main thing to remember when looking at Dutch designs is that the culture is different. If you hit a cyclist, you're automatically presumed to be at fault." Two myths: the culture is not that different. Until about 1970 we followed the American example, and then realised that there were other road users beside cars. Just a matter of time. And as for hitting a cyclist: drivers are supposed to look out for cyclists, but if the cyclist is at fault the driver goes free. I have had two cyclist deaths on my doorstep this year, both caused by cyclists suddenly swerving out of the bike lane. Dutch solution to the problem: the bike lane will be removed and a separate bike path laid behind a line of trees. Both parties happy (and the road is narrow by American standards).

J.. said...

I think the point NE2 was trying to make, is that given culture/conditions in the US, the Dutch design won't work. I disagree.
For one thing, a driver who fails to yield or isn't looking where he's going, is going to endanger you regardless of intersection design. I don't see how a different lane design or vehicular cycling is going to provide more safety.

Vehicular cyclists are always going on about making yourself visible by taking a proper position on the road. That is exactly what this design does, except it's not on the road and thus minimizes conflict. And saying "I didn't see him" is an increasingly untennable position when the person you just hit was right there in front of you, clearly within your field of vision.

And of course, at the end of the day, the statistics clearly back this up.

Anonymous said...

This blog resolutely promotes the line, that there is no 'cultural' aspect at work in the Netherlands. In fact it is probably the most important factor of all, more important than segregated facilities. You can see how important it is, by comparing cycling within the Netherlands itself, where there are strong regional differences, apparently related to religion. (Those of you who looked closely at the formulas used to predict cycle usage may have seen, that they include religion).

Cycling in Limburg (Catholic, supports right-wing PVV) is not the same as cycling in North Holland. You see less cyclists, and there are fewer cycle paths. You will see road racers: they treat cycling as a sport, the 'normal' pattern in most countries.

If 'Dutch' cycling infrastructure models are in fact regional in nature, and not fully exportable to other regions within the Netherlands, then they are not exportable to other countries either.

I would guess that the best single predictor of cycle usage in the Netherlands, is who you voted for at the last elections. That would probably predict driver attitudes to cyclists as well.

The implication is of course, that cycling infrastructure will not change things, since such attitudes are deep-rooted and personal. For that reason, I think cycling in the USA is a lost cause, and always will be. It is un-American to cycle, that's how Americans think, and no amount of new cycle infrastructure will change that.

Floris said...

@ Anonymous

The difference you notice is a geographical one.

Lage parts of Limburg consist of hills and are considered interesting for road racers from the rest of the Netherlands. They come to cycle parts of the Amstel Gold Race.

A few weeks ago a dutch newspaper published an article about normal Dutch people (highschool students, a mailman) who cycle up and down the Cauberg, a famous climb in that race, on a daily basis.

J.. said...


Firstly, you misunderstand. This blog does not assert that "there's no cultural aspect" at work. Rather, it correctly points out that Dutch cycling policy works regardless of these cultural aspects. Dutch catholics cycle more than other catholics, Dutch conservatives cycle more than other conservatives, Dutch liberals cycle more than other liberals, etc.

Secondly, in your Limburg-argument you confuse correlation and causation. And you imply that catholics are less enthusiastic cyclist than other people. I don't buy that, certainly not without a compelling argument, a component conspicuously absent from your post.

As for politics, it's a fact that the two big rightwing parties are more pro-car, pro-highway, etc. But this is contingent, and hardly an ideological fixture. For instance, small rightwing christian parties have a rather good track record on cycling policy. Plus, being pro-car doesn't mean you're anti-cycling. You can be both, to a large extent.

I can't verify whether political preference is a good predictor for cycle usage, but I doubt it. Maybe David can weigh in on this.
I do know cycle usage actually increases with income (i.e. rich people are more likely to cycle than poor people) and I know that conservative views are more likely found with more affluent people. On the other hand, other rightwing views (racism, zero-tolerance, anti-immigration policy) are more popular with poor people. But then, that would be correlation again, not causation.

J.. said...

Anonymous wrote:

"It is un-American to cycle, that's how Americans think, and no amount of new cycle infrastructure will change that."

I think this statement reveals a profound ignorance of American society. First of all, the US is a very big place with wildly differing cultural roots. California and Alabama are not exactly the same thing, culturaly. They don't all think alike. Furthermore, when you look at what's going on in a lot of American cities, you find that building the infrastructure *does* change that. Car-dominance is not going anywhere any time soon, but it's slowly becoming more and more normal to cycle.

And looming on the horizon there's always the Energy Issue. In 2004, in the thick of the Iraq/Afghanistan wars, television analysts were foaming at the mouth with anger at the prospect of oil prices maybe breaking the 40$/barrel barrier. Today, you hear sighs of relief when it drops below the 100$/barrel mark. This trend isn't going to be reversed any time soon. Americans are going to be driving their cars a lot less, whether they like it or not.

Anonymous said...

I was with you untill it got to the bit about the roundabouts.... A lane around the outer edge of a roundabout is a really bad idea. Some lunkheads here in the UK built us a few of them in the 80s and now they have been mostly removed. Going round that outer edge puts you in very real risk of a collision unless you are turning left. Roundabouts have to be handled as a vehicle OR there needs to be segragation.

David Hembrow said...

Londonneur: Roundabouts, of course, are not all created equal. I have to say that I flinched a little when I first saw a roundabout like that in NL. However, there are several design features which maybe are not immediately obvious and which do make a difference. This is not the same in use as the dire attempts at doing something similar by the lunkheads in the UK.

For a start, the centre of the roundabout is raised. You'll see that the green part in the centre looks small, but around it are light coloured bricks. These are noisy to drive over and also sloped, leading to a feeling of driving around a corner with incorrect camber. Most drivers completely avoid this area, making their corner radii much smaller, but trucks and buses have to use it. Either way, this slows down motor vehicles.

The cycle lane around the outside is actually segregated for much of the distance. You can see this clearer if you look directly on streetview. Again, this has two effects. It makes the radius smaller for drivers at the same time as making it larger for cyclists.

Lastly, the sort of junction where you'll find this is different in character. Note that on streetview, especially if you rotate the view by 360 degrees, you see well over a dozen cyclists of all ages and on all kinds of bikes, but only two cars using this junction at the time that the Google car went through it. This junction design is appropriate for that sort of usage pattern, but perhaps not on high speed roads (whether by design/speed limit or by culture) where there are few cyclists - which would perhaps be most of the places where it might be considered in the UK.

Anonymous said...

"This junction design is appropriate for that sort of usage pattern, but perhaps not on high speed roads (whether by design/speed limit or by culture) where there are few cyclists - which would perhaps be most of the places where it might be considered in the UK."

And there you have put you finger right on it... The cultural issue is the main thing holding us back in the UK. In Holland the drivers are carefull as they turn right across that cycle lane. It's a low speed one lane roundabout. You need the lanes to connect the incomming/outgoing lanes and in any case you could build anything you like because the culture is doing the work. I'm still unconvinced by roundabout lanes though :-)

Love your blog

David Hembrow said...

"And there you have put you finger right on it... The cultural issue is the main thing holding us back in the UK."

Well, actually, not entirely. The roundabout as shown by Mark depends on this, and I'd not like to see an attempt at copying that in the UK at the moment either.

However, neither the style of junction which is the main subject of this post nor that which I covered a few days ago depend much on driver behaviour in order to be safe for cyclists. These designs could be put down almost anywhere and would work.

Infrastructure has an enormous role in altering attitudes. Once cyclists are no longer in conflict with drivers, conditions are much better. In the few places in the Netherlands where conflict is forced by design of infrastructure (i.e. it resembles the UK, USA etc.) then problems for cyclists return.

Cars have a very strange effect on their drivers everywhere. There is nothing magical about Dutch people which makes them particularly different. Just as most motorists stop at red traffic lights everywhere in the world, here most motorists also obey other rules forced on them by the infrastructure, some of which particularly favour cyclists.

There is a reason why the only "road rage" which I've experienced in four years of living in the Netherlands occurred where it did, and it comes down to a break in the usual standard of infrastructure, re-introducing "might is right".

Anonymous said...

I agree...

I should point out though that I was using a much broader meaning for "culture". Of course driver culture is an issue but there is also wider culture that allows this stuff to be built in the first place.

Our shiny new Transport minister, Philip Hammond or that guy "Rubber Knickers " Pickles are never going to build us anything useful. They see cycling as an annoyance rather then the solution you or I see it as. It's cultural.

I'm not dewey eyed about holland. I just know that the govt over there is prepared to use bikes to solve some issues that we seem to just be happy to live with. It's a shame.

Anonymous said...

Notice the AVD Van in the street view picture :)
I've not seen one of these before.

Step-Through said...

There is a cultural difference between the Netherlands and the U.S., but it is neither inherent nor immutable. It is simply the custom that has emerged in response to very different safety legislation and enforcement. In NL (as I understand it), the motorist is always at fault in a crash with a bicycle or pedestrian. I don't know how it is enforced or prosecuted, but that alone is very different from the U.S.

Crashes in the US get sorted out by right of way, and there is often missing or ambiguous infrastructure for bicycles and pedestrians - missing crosswalks, bike lanes in the wrong place. Even in the other picture you posted on here showing right-turning cars merging across the bike lane, it's not really clear who has right of way. But after the crash has happened, a ticket may be issued, and then they go to traffic court. If the arresting officer doesn't come to the court date (and they are often too busy to come) the charge is usually dropped. If the cyclist or pedestrian is no longer alive to testify, the motorist can just say that it was the cyclist at fault, or "I didn't see him", and usually get away with it. Tickets are rarely issued for traffic violations that do not result in damage or injury, and the penalties are usually a minor fine.

This makes many American bicycle riders choose settings where they can have more control over the situation rather than relying on motorists to comply with traffic laws or exercise caution. It is not the best solution, but simply a reaction to the current conditions. However, those conditions could and should be changed with stricter legislation, enforcement, and penalties.

David Hembrow said...

Step-through: You misunderstand the law in the Netherlands. Drivers have the heavier vehicle and are held financially responsible for damage in many cases due to their ability to cause that damage, but (except when it comes to minors) fault actually lies with whoever caused a crash.

This legislation was actually introduced quite recently, and has no real bearing on the rate of cycling in this country.

I think the difference in interest on the part of the law comes down more to the average member of the public's understanding of cycling. You're more likely to get a fair hearing from a policeman / judge / newspaper journalist who is a cyclist than one who is not.

In the Netherlands, almost everyone is "a cyclist", including the police, judges, newspaper journalists etc. Cyclists are not considered to be outlaws just for being cyclists.

r s thompson said...

" However, those conditions could and should be changed with stricter legislation, enforcement, and penalties."

thats stupid.

the penalties and enforcement (its not enforcement, only reaction) dont stop a large object from hitting a smaller object...

when the money could be put into bike lanes creating someting truly safer and not the ridiculous subjective safety the blog owner keeps blabbing about.

r s thompson said...

them a separate green phase makes the junction more efficient.....

running through empty intersections in america when the light is red is efficient.

Reaperexpress said...

Just to clarify, when I said "it's just a lane change", I meant for motorists. Cyclists do not need to make a lane change if the junction is designed properly. said...

For me as an urban planner this Dutch example of a cycle track at an intersection is an old-fashioned model and a dangerous trap for cyclists and pedestrians. First of all pedestrians have a much longer way to cross the intersection, because the way for pedestrians swivels to the right. Many pedestrians on cycle tracks like these use the cycle track on the crossing street and give reason for heavy problems with cyclists.
Second the waiting cyclist at red light obstructs the way for other cyclists crossing his way. This leads to forbidden usage of the sidewalks by cyclists to sidestep, but pedestrians won´t like this. And why building cycle tracks, if you regularly can´t use them due to regularly obstructions of other cyclists. Here an example from Germany with the same model like it is established similar to the US-example on the right picture.

In the long run you will trainee cyclists always riding on sidewalks at any situation, if the cycle track cannot be used clearly.

Third the car driver who stopped in the intersection to let the cyclist passing, who stopped in front of the signal, has even to look backwards to look too for more cyclists coming later. So the accidents between cyclists and car driver will happen here at this moment, if the car driver does not look backwards. That is the reason for many heavy accidents with cyclists in Germany.

At least: The kind of animation is brilliant, the example seems really bad to me.

Anonymous said...

hamburgize, as a person who crosses this type of junctions every day (as a pedestrian and cyclist), I can only say you are wrong on almost all accounts.

Pedestrians always stay on the sidewalk, it just does not occur to anyone to walk on the cyclepath because it's a few meters less walking. This coming from a pedestrian that will take just about any shortcut if it saves me some time (eg crossing roads diagonally).

Cyclists don't put themselves in the way of other cyclist. They will stay back or leave a gap. Noone wants to be seen as "the one blocking everyone's way". You will get scorned if you do this, so it does not happen.

Of course you still need proper design, the german one you link is silly. It puts a button to push on a place where you can *only* obstruct others. The dutch design (and all examples given here) have a buffer behind the light where people can stand without obstructing. Or they put the "stop" markings back behind the crossing with the other path.

The german "design" is only like that because they turn parts of the sidewalk into a bike path overnight. In other words, not a design at all. There is not even a height segregation (in dutch junctions bikes are always at road level and sidewalk always raised or at best partially lowered (for accessibility).

For the third thing you list, I don't have a drivers license so I can't comment on people in cars who drive somewhere without looking first (though as a cyclist this is a strange thing to imagine for me).

But none the less, on large junctions this is an irrelevant argument, as pointed out there will be no right turns when bikes cross in that case. Even when this is not the case, the bikes swerving right before crossing (the very thing your criticize earlier) makes them much more visible compared to when they come from behind you because the bikelane is right next to the road.

I'll point out to you though dutch cyclists have to deal with these junctions routinly, the Netherlands has the second lowest amount of traffic deaths per million inhabitants, despite having a very large share of cyclists.

sunnyvale_trails said...

Looks good, but I am unsure how to make it work ay expressway and freeway entrances where the cars do not have a stop line but a non stop merge and yield often with a porkchop. As the term people cycling vs vehicular cycling. People cycling appeals more, but I see vehicular cycling being the reality in some cases.

Hope more videos come on freeway type entrances