Cross-posted from streets.mn
Cross-posted from streets.mn
The traffic signal at Oak St. SE and Fulton St. SE in Minneapolis is mis-timed/mis-phased/mis-indicated. It has been so for several years.
Northbound Pedestrians on the west side are given a red indicator (Don’t Walk) even though Southbound traffic has a green light and green left turn arrow. Clearly Northbound Pedestrians on the east side of the intersection should have a red indicator in such a configuration, as they are in conflict, just not those on the west side.
We can imagine why this might have occurred, but basically the walk signals are tied together even though vehicle traffic has a split phase. Obviously the technology exists so this is not necessary. This occurs from the 0:36 to the 0:49 mark in the video below. It’s “only” 13 seconds of course, but we could say the same about vehicle delays. It’s 13 seconds every minute of every day for every pedestrian at the intersection. This is near the University of Minnesota campus so the number of pedestrians is non-zero. In late spring it’s not an unpleasant wait. Talk to me in January.
Video (looking SB on Oak Street, East on the left, West on the right):
How often does this occur? Where else do the traffic engineers not think through the implications for pedestrians?
I cannot comment on what the optimal traffic signal timings are for this intersection, but this is clearly not it.
My favorite five-way intersection that should be a roundabout was on flashing red yesterday, and I concluded that was a time for me to monitor this natural experiment. At 4:30 on May 12 I video of the intersection for 4 minutes. I counted 100 cars and 11 pedestrians and bikes in just over 4 minutes. Simply multiplying 111*15 gives a capacity of 1665 units per hour. This is somewhat off my estimates from a signalized intersection. On the other hand, delays were rather short, and while there were short queues at some of the approaches (which helps ensure the intersection is fully utilized), they were not getting longer over time.
The video is below:
A map of the location.
In places that are, or want to be, walkable, and serve pedestrian traffic, traffic signals should have a default setting of pedestrian scramble (Barnes Dance), and only switch to a green light for motor vehicles from a particular approach (for a short time period) when it is actually actuated by a vehicle. Buses and emergency vehicles would still be able to get priority by signaling from upstream.
Today, in the United States, traffic signals are usually designed with the objective of minimizing motor vehicle delay, yet many policies and plans have a stated aim of reducing the amount of vehicle miles traveled or the automobile mode share. How does lowering the cost of driving and increasing the cost of every other mode help with that objective?
Now we place pedestrians in the supplicant position of begging for a green light. Let’s give walkers some dignity, and instead of making them “scramble” at the intersection, allow them to simply purposefully walk, or even amble. If instead of pedestrians waiting for cars, cars had to wait for pedestrians, vehicle delay would undoubtedly rise. But vehicle counts would fall, and pedestrian demand would rise. Where would the vehicles go, would they disappear or reroute?
Think about places this would work in your community. In the Twin Cities, I think this would be great for Dinkytown and Uptown.
It is a change, it would need to be tested somewhere before it could be done everywhere. There will always be resistance by the stalwarts. But we should experiment.
The measure of success would be change in pedestrian and bike counts, the reduction in vehicle counts (at this location), and maybe the change in sales at nearby businesses.
Sioux City has automatic red-light running detection cameras. These are doing their job. In the Sioux City Journal by Molly Montag article on the topic, the facts are all clear, unfortunately the headline takes the small negative instead of the large positive as the lede: “Sioux City data: Rear-end crashes increased at 5 red-light intersections”
In Sioux City’s case, new police data obtained by the Journal show a 40 percent decrease in crashes from motorists running red lights at intersections with cameras and a 15 percent reduction in all crashes. Iowa Department of Transportation data also show a decline in accidents involving red-light running. “In general, this is a positive result, as rear-end crashes (though not desirable) are not as severe (or dangerous) as right-angle crashes that might otherwise occur,” David Levinson, professor of transportation engineering at the University of Minnesota, said in an email.
University of Missouri-St. Louis transportation studies professor Ray Mundy said rear-end crashes typically increase after systems are installed and drivers slam on the brakes when they see the camera flash. Such accidents usually decrease over time as people get used to driving through camera-controlled intersections. He echoed Levinson in saying more slower-speed rear-end crashes are a tradeoff for reducing higher-speed, T-bone crashes that happen when drivers run red lights.
The University of Minnesota’s Levinson agreed the decrease in red-light crashes showed the systems improved safety.
“That is the important takeaway,” he wrote.
I looked at the data. While in my analysis the total number of crashes did not change, it is clear that automated traffic enforcement reduced the number of “ran traffic signal-involved” crashes and increased the number of “rear-end-involved” crashes. The differences before and after installation by crash-type are statistically significant and meaningful in both cases. This is consistent with general results nationally about the effects of automated traffic enforcement. (And what you would expect if there were more sharp braking at intersections as people strive to avoid fines).
Cost to users is a transfer to the city (and should otherwise reduce some other taxes the city is collecting), and though there is some cost to administering the system, that is outweighed in general by the safety benefit.
Nick Musachio, local inventor in Minnesota, has just been issued a patent (No. 8,711,005) for his Always Green Traffic Control System. (Since this is transportation, we will abbreviate this AGTCS)
Imagine you have an isolated signalized intersection, operating near but below capacity. If vehicles were able to travel at the correct speed when approaching the intersection for a significant distance, they should be able to travel through the intersection without hitting a red light or being delayed by standing queues. If at 45 MPH they would hit a red light, but at 35 MPH would get a green, they should be informed to reduce speed to 35 MPH. This not only reduces driver delay, but should decrease crashes and decrease emissions, both of which are exacerbated by intersection control and braking and acceleration.
How would drivers know which speed to travel? An upstream Variable Message Sign with Dynamic Speed Limits (tied into the traffic signal controller cabinet, or with the pre-programmed traffic signal timings) would tell them the best speed to avoid stopping. If only the first car in a platoon does this (on a 1 lane road), all following cars are controlled by default.
Audi has a similar in-vehicle system. That is only useful if the traffic agencies produce live feeds of traffic signal timings. Comment: it is appalling that such a traffic signal timing live feed doesn’t generally exist (even transit agencies, not historically known for their cutting edge research) have GTFS.
AGTCS is infrastructure based, and works for all vehicles anywhere an agency wants to set it up.
Some videos and simulations below.
Cross-posted from streets.mn: Begging for Simplicity … in which I complain about beg buttons, and general second class treatment for pedestrians.
Pedestrian actuators call for a pedestrian signal at an intersection which is semi-actuated (where the green-time goes to the mainline except when a vehicle is on the side street, subject to a maximum cycle length and a minimum green time for the side street) or fully-actuated (where the green time is allocated to approaches which are actuated subject to a lot of constraints). When they do this, they also tell the controller to extend the green time (and parallel walk signal) given to a phase to be sufficiently long to allow pedestrians to safely cross.
But this has problems. Imagine you are on a side street about to cross a main street and the light turns green for the cars, but the Don’t Walk sign remains (since no Pedestrian actuation was recorded). You did not get to the actuator quickly enough. Should you cross on the Green but against the Don’t Walk, or should you wait almost an entire cycle for this to come around. You may or may not have enough time to make it across.
Second imagine the actuator is broken. It may never give a walk signal. (There aresolutions for that, where the default state of broken is “on” instead of “off”, but that doesn’t seem to be widely deployed). For instance, recently I reported toSeeClickFix a broken pedestrian actuator at Franklin and Seymour Avenue in Southeast Minneapolis, which was corrected within 18 days (i.e. the case was closed within 18 days). I am fairly confident a broken traffic light serving cars would have been corrected sooner.
The next two “beg buttons” (pedestrian actuators for traffic signals) were recently photographed. The one on the University of Minnesota campus (at Beacon and Harvard) recently had a time exemption added, implying that the actuator need not be pressed between 8 am and 6 pm weekdays. (This time exemption seems to have been removed since the photo was taken). This is an improvement over the previous situation (pushing the button in the middle of the day). Still one expects this will, like so many others, become a placebo button, or just break and make it so there is no pedestrian phase.
However, the complexity is still needless. Traffic signals on streets with sidewalks (which implies pedestrian traffic either exists or is desired) should ALWAYS have an automatic walk phase, just as every cycle gives green time to cars from every approach. (This is especially true in pedestrian areas like a college campus which has a plan that aims to prioritize walking.) Actuators are fine if they make the walk signal come sooner, but being unpushed should not be used as an excuse not to have a walk phase at all. Car drivers don’t have to go out of their way to press actuators, why should pedestrians?
(If traffic is so low you are concerned the time devoted to a pedestrian phase (~12 seconds (36 ft at 3 fps)?) is too long (will cause too much vehicle delay) for this two lane roadway, maybe it shouldn’t be a signal but instead a stop sign (which requires no pedestrian signal) or a yield sign. This can be implemented with flashing red lights if you must you electrical gear.)
Fortunately it is not as complex as the last, at Minneapolis’s favorite five-way intersection (Franklin Avenue, East River Road and 27th Ave SE), which gives instructions for something that should be tacit. That it is not tacit indicates it is a flawed design. If I can read the instructions, I already know how to cross a street. It is not like pedestrian actuators are a new technology. While I want more information at bus stops, crossing a street should be straight-forward, and not require an 11 line instruction set with five graphics. Sadly there is more information here than at the nearest bus stop.
Update. I found the tweet below, which seems appropriate.
I found this rendering of 222 Hennepin earlier today, after reading Bill’s post on architecture. The part that I liked best is the traffic light. First it is sideways, a design that is used in some places in the world, but not Minneapolis, for a lot of good reasons.
I grew up in Columbia, Maryland, which for many years had sideways traffic lights, mainly as an architectural distinction, but which were abandoned because of the confusion created.
A major problem of such uncommon lights is that color-blind person might not know if the green is on the left or on the right. In Wisconsin (as everywhere else in the US where it is done, and standardized in the MUTCD, it is red on the left, green on the right.
In the rendering, it is the opposite.
It makes you wonder what other tricks and graphical shortcuts are going on to make the rendering desirable to approve and move into, but won’t really turn out as implied.
Eddie, a traveler in a hurry arrives at a traffic light from the East. He would pay up to $18 to save an hour. Sue, a less-hurried casual traveler arrives from the South, she would only pay $6 to save an hour of travel right now. Who gets the green light, who gets the red? Presently this is decided without consideration of how much Eddie or Sue would be willing to pay to save a few seconds or a few minutes. No one has the ability or the authority to make a transaction occur where Eddie can pay a few cents to Sue and get the green light while Sue waits for the light to change. Until recently, this was because it was technologically infeasible, but in recent years, advances in transportation signal technology and real-time wireless vehicle-infrastructure communications have made this once impossible transaction possible. Now it is institutional constraints that prevent this from happening. Traffic signals are in almost all towns, cities, counties, and states publicly owned and managed. Imagine instead that this was a service that private firms would bid to supply.
A new organization, LightSpeed Traffic, has paid your city $100,000 a year for the privilege of managing traffic signals. Instead of this being a cost center for the city, it is now a revenue generator. Why do they do this? A private operator is able to use traffic signals more efficiently from an economic perspective than a public agency, they can obtain revenues from sources such as:
Presently a few companies operate traffic signals under contract to municipalities, notably in Sandy Springs, Georgia. None yet use signals innovatively as described above.
Obviously this can get quite complex: there may be more than one approaching driver, how do you decide the baseline to estimate vehicle time-saved or time-list by adjusting signal timings, how does this work in networks instead of just isolated intersections.
But scarce resources (like two vehicles seeking to use the same space at the same time) can be allocated in many ways other than arbitrarily or first-come first-serve, to the benefit of all. Sometimes the best solution is a yield sign, sometimes a stop sign, sometimes a roundabout, sometimes a traffic signal, sometimes a grade separation. To be clear, it is not always a traffic signal. When it is a traffic signal, there might be some merits to thinking creatively about the organization and operation of the market that is created by the rationing of time for the benefit of all.
UM News reports on my colleague Henry Liu’s new SMART Signal Technologies startup: University of Minnesota startup to improve traffic flow on congested roads:
“Based on research from the University of Minnesota, SMART Signal Technologies, Inc., will commercialize a system to better predict and manage the flow of traffic on roads controlled by traffic lights. The system could potentially cut down on traffic congestion and help drivers save both time and fuel.
Using data from existing traffic signal equipment, the system accurately calculates queue length at signalized intersections. These data, collected in real time and archived in a database, will allow cities across the state to better mediate the flow of traffic at peak times using real time performance measures provided by the system.”
I have talked about this before. I hope it gets widely deployed, what we don’t know about travel times arterials in real-time is embarrassing.