More time wasted reinventing the wheel. Just take the Dutch CROW manual off the shelf and use it.

The trees are in the wrong place in the last pic.

These design concepts will be put in adjacent to the Forest Hills MBTA station in Boston within the next couple years.

Unfortunately – these “standards” have no teeth in Massachusetts – even if they are officially adopted by municipalities, it usually takes educated citizens to make sure that traffic engineers are meeting these recommendations. It does give some ammunition to activists, though… just FYI, the state had recommendations for bike lanes dating back to the 90s, plus it was in the city of Boston’s own roadway design manual as early as 1997 – but it wasn’t until 10 years later in 2007 when the city finally started striping bike lanes.

Regarding the signal phasing options, it should be standard procedure for state DOTS that red for cars is flashing yellow (i.e. yield) for bikes unless lines of sight make it unsafe to yield on red. No good reason cyclists should be forced to wait at empty intersections. Repeated stops for red in areas with lots of traffic signals seriously impact the time/energy efficiency of cycling. By using separate signals for bikes, such as a flashing yellow when motor vehicles get red, we avoid the need of having to pass controversial Idaho stop laws to give cyclists the same benefit.

The CROW manual was referenced in the design of this guide, but CROW won’t fly with the FHWA so it needs to be tweaked and adopted unfortunately. The design team took a refrence trip to The Netherlands in the design of this guide as well.

My first reaction to these is that they are quite dangerous. The bend-in approach perhaps the worst.

The straight (SK-10) is close but the crossing should be moved farther back from the junction so that the crossing is perpendicular to the car & driver and so brings bicyclists in to the natural field of view of the driver (the 90 degree example in the first image is not within the natural field of view). This will also allow the driver to fully make their turn before encountering the bikeway which has one benefit of allowing them to know that they are out of the way of traffic behind them but perhaps most importantly separates turning from the bikeway in time so that they only deal with one item at a time instead of multiple.

I do commend them on the driveway crossing though. Much better than most we see in the U.S.

I think you’re close, but red for cars should be flashing red for bikes (stop and proceed when clear, same as stop sign), and stop sign should be flashing yellow (yield).

Don’t bikes have much greater visibility and travel at lower speeds? The entire reason traffic lights exist is because motor vehicles can’t safely negotiate intersections at speeds much above 20-25 mph due to visibility, plus inability to stop in time if an obstacle is detected. If we gave motorists a flashing yellow instead of a red when cross traffic had the green they would almost certainly continue across the intersection at the normal traffic speed for that street, whether it was 25 mph or 55 mph. End result is collisions would occur. These factors just don’t exist for cyclists. As a matter of course cyclists slow down enough at red lights to see and stop for anything coming, if for no other reason than self-interest.

There are other advantages to letting cyclists treat reds as yields:

1) They avoid be in the aggressive pack of cars starting out from the light.
2) They avoid breathing exhaust fumes of stopped idling vehicles.
3) They save time and energy.
4) They may avoid hitting as many red lights.
5) They often have the street almost entire to themselves for many blocks while the platoon of motor vehicles is waiting at the red light.

Obviously I feel the best policy with regard to bike infrastructure is to bypass busier intersections altogether with overpasses or underpasses. This may not always be practical. Moreover, for much of the day cross traffic may be light enough that “red as yield” gives practically the same benefits as an overpass or underpass, saving a lot of money.

On another note, the only reason we might even need to consider this idea is due to the failure of traffic engineers to only give red lights when something is actually crossing. In theory, no traffic signal should ever go red when there is no cross traffic or pedestrians. In practice we can probably come close to this ideal with a system of sensors but few places bother even trying this.

What’s interesting to me is that while I’ve had the same thought in the U.S. and occasionally in the UK, I’ve never thought this while riding in The Netherlands.

I think the difference may lay in a few areas:

– Signal cycle times in The Netherlands are massively shorter than in the U.S. and a bit shorter than UK.

– Main routes default to green during non-peak periods and most riding is along main routes.

– NL does have sensors so you will often get a green by the time you get to a crossing if there is no cross traffic.

– When you do have to wait it is nearly always less than 9 seconds (though occasionally 20 or 30 seconds).

Perhaps the biggest difference, and this gets a bit at your underpass comment, is that there are massively fewer traffic signals in The Netherlands. Actually I think any country has many fewer than the U.S. Many fewer stop signs as well. Mostly replaced by roundabouts and sharks teeth.

See: https://bicycledutch.wordpress.com/2014/01/30/ride-five-kilometres-priority/

Stupid stupid stupid. When are people going to wake up and realize that straight traffic to the right of turning traffic is recipe for disaster? The only way to make this remotely safe is with separate signal phasing, and no right on red, assuming cyclists obey the signals. But unless they employ the “all directions green” style of signal phasing used in the Netherlands for cyclists, it will end up creating more wait time for everyone, more congestion, and will lead to more cyclists disobeying the signals because they’re tired of waiting.

It’s clear as day to me. Why don’t these people get it??

One small flaw in the design. It puts the crosswalks further away from the intersection , meaning that pedestrians will have to go out of their way to get to the crosswalk instead of walking in a straight line. In NYC at least, they won’t go for it. They will just walk into the bike lane and cross the the street.

That crossing placement is much safer as it allows cars to complete their turn before encountering the crossing and places the crossing perpendicular to drivers so it is much easier to see people. This also allows cars to clear the traffic lane so that cars behind them can continue while they are waiting on people to cross.

That said, I agree with you about NYC. NYC is a bit of a different world.

A green wave has been suggested but the problem here is any green wave which accommodated the majority of cyclists would be intolerably slow for motor traffic. I’m aware that green waves are sometimes used in the Netherlands but in general they tend to be limited to relatively short (1 km) sections right in the CBD. As such, they don’t severely impact travel times of motor vehicles or faster cyclists. Remember NYC arterials with lights every 250 feet go on for literally miles, sometimes more than ten miles. Under these conditions, a 12 or 13 mph green wave just wouldn’t be viable. A green wave with a speed tolerable for motor traffic (i.e. 20 to 25 mph) would be much too fast for most cyclists.

I’d be really curious exactly what a Dutch traffic engineer would do to allow safe, efficient bicycle travel on typical NYC arterials. My guess is in the end they would probably seriously consider putting the bike route on a viaduct as there are really few other options allowing safe, fast bike travel given the sheer number of other street users.

Remember that when the NACTO guide came out literally two years ago, it was pretty much the best available.

That just goes to show how quickly our knowledge is advancing. Perhaps I should have put time in, then, to learn Dutch so I could read CROW in the original. Haha.

Dutch engineers likely would take a lane or two from them to make room for bikeways then put in trams.

Ah yes, the good ‘ole American exceptionalism at work again.

The bend-in approach isn’t that bad, that’s the design that necessary for all-directions green phasings to really work. It also looks like SK-10 actually is within the range of best distances between bikeway and travel lanes which is between two and five meters. I’d agree that they should err on the safer side, but this isn’t bad at all and would not be uncommon on even new construction in The Netherlands.

Generally speaking, the offset is hardly noticeable and I doubt most pedestrians would even be bothered by it. For disabled individuals, it’s actually an improvement in many ways because the curb cut is oriented toward the direction of travel, not toward the middle of an intersection that is in some cases, literally the size of half of a football field.

I’m going to add a bit of a pause to my comment above. We’ve seen such a consistent series of poor to awful design guidance from AASHTO, NACTO, MUTCD, and others, including from Toole, that my reaction in looking at these was ‘more of the same’. As @marvennorman:disqus pointed out below however, the bend-in approach above may be intended for simultaneous green and if so then this might perhaps be a good solution.

We’ll see when the full guide is available.

May being the really key word since ADG isn’t allowed for bike signals at the moment.

Actually, experience in Chicago suggests that having bicycle-specific signals increases bicyclist compliance, not reduces it. Cycle times could potentially be an issue, but I would hope that the engineers would be able to figure out the appropriate times and detection for smooth operations, something which will certainly be made even easier in CA due to the switch from LOS to VMT being the threshold for traffic impacts under CEQA. Though as you pointed out, ADG is quite effective and can easily handle quite a high number of bicycle movements with two phases per cycle.

OK, you’re two for two. SK-10 looked a lot closer to the junction. When I blew it up a bit it does appear much better than I’d originally thought.

Signal cycle time is sort of a game of balance. If you shorten the time for each phase, you don’t have to wait as long between phases, but fewer cars (or bikes) can make it through on a given phase. Make the time longer, and more cars can go on each phase, but the wait time is also longer between each. In both scenarios, traffic ends up getting backed up farther and farther. And the longer the queue, the longer it takes cars farther back in the queue to make it through the light since space opens up between each car as everyone starts moving.

So now with bike signals you double the number of phases from 2 to 4, plus any left turn arrow phasing. At least with the all directions green you’re only adding a single phase to the equation for a particular intersection.

No! The whole point of separate bike signals is to separate straight-going cyclists from turning motorists! You know, since you have to deal with the violation of the rules of movement by having straight traffic to the right of turning traffic in the first place.

If you have north/south and east/west traffic, one has to wait for the other during the other’s green phase. Add in bike signals and now you have each N/S and E/W direction for EACH MODE TYPE having to wait for THREE other phases (N/S and E/W of the other mode, plus the other direction of your mode). To make the wait time the same as before, the green time for each direction would have to be more or less cut in half.

The only type of bike signal that makes sense is the all directions green for bicyclists, while all motorists wait.