20 MPH Bills Gain Support in Albany, But Will Need Help From Key Senators

State bills to set the default speed limit in NYC at 20 miles per hour picked up several cosponsors Tuesday, including a Senate Republican, but key Senate leaders have not signed on.

Senators Jeff Klein and Marty Golden
Jeff Klein and Marty Golden could get 20 mph legislation through the State Senate. They have yet to sign on.

As of this morning, bills from Assembly Member Dan O’Donnell and State Senator Martin Malave Dilan have at least 20 and 11 cosponsors, respectively. Those totals include some whose names were added yesterday, when Families for Safe Streets traveled to Albany to meet legislators face to face. At the end of the day, the group said, 15 lawmakers who were not cosponsors agreed to back the bills.

One of them was safe streets champion Joe Lentol, who spoke with grieving family members in his office Tuesday. Assembly Member Lentol was not aware he was not a cosponsor, but attached his name to the bill after the meeting. Speaking of his campaign to make deadly McGuinness Boulevard an arterial slow zone, Lentol said, “That’s just the beginning. We need to do more.”

In the Senate, city Democrats Michael Gianaris, Bill Perkins, and Jose Serrano signed on, as did Nassau Republican Carl Marcellino. Marcellino belongs to the Senate majority, but to clear the chamber the bill will probably need to pick up the support of either Marty Golden or Jeff Klein. Klein, of the Bronx, is Senate co-majority leader and heads the Independent Democratic Conference, and Golden is the leader of NYC’s Republican delegation.

At other meetings I attended, Assembly Member Barbara Clark, Democrat from Queens, and Republican David McDonough of Nassau County committed their support, but are not yet listed as cosponsors.

Members of Families for Safe Streets stressed yesterday the importance of getting the life-saving 20 mph bills passed this year. We will follow developments as the legislative session draws to a close in the coming weeks.

  • Peter Engel

    I’m not sure about Klein, but the only way to get Marty Golden to do the right thing is to find compromising pix of him w/farm animals.

  • Joe R.

    In lieu of this, I would rather just have a bill which lets NYC set its own speed limits without needing to ask for Albany’s approval. I’m also finding that there are some issues here which this bill, and the others, don’t really address. “Reduces the speed limit in the city of New York to twenty miles per hour, except where a different speed is determined appropriate…”

    Determined by whom? Qualified traffic engineers or laypeople? The bill should specify that qualified traffic engineers can and should be able to set the speed limit of a street to any number they deem appropriate by standard engineering practices. On the flip side, NYC DOT should get serious funding for reengineering streets to make slower speeds self-enforcing. Ideally, the 85th percentile speed on every local street should end up at or less than the speed limit so that we don’t depend heavily on enforcement for safety.

    The other problem I’m having here is I’m not seeing anything which requires uncontrolled intersections in 20 mph slow zones. This is a key factor in their success overseas because uncontrolled intersections by their nature keep traffic speeds to about 20 mph or less for the simple reason that it’s suicidal to go any faster than ~20 mph on on such streets. NYC should be no different. If you have a 20 mph zone then every single intersection within the zone must be uncontrolled unless there are good reasons for not doing so. Those reasons might include poor lines of sight which can’t be corrected (i.e. bridge abutments and the like). Poor lines of sight due to allowing parking close to the intersection is NOT a reason here, but rather something which should be corrected when a slow zone is implemented. Anyway, I think slow zones will benefit NYC but we need to think things through a bit more as we implement them to emulate what we know works elsewhere.

  • nycbikecommuter

    We have 30mph speed limit now and people drive 40+. We will have 20mph speed limit and people will drive 40+. What’s the difference?

  • Joe R.

    That’s exactly why I said we need to think things through a little better than we are. As it stands now, it seems the only thing this bill will do is change the numbers on the signs from 30 to 20. Streets need to be engineered appropriately if you want traffic to move at 20 mph or 30 mph instead of 40 to 50 mph.

  • lop

    Cameras that only enforce limit+10 will keep people from going more than 30.

  • lop

    There’s no reason not to ask people to drive slower than the fastest they feel comfortable driving. Unless you want to wait thirty years to see half the streets rebuilt before there’s progress.

  • Joe R.

    It won’t take major street rebuilding to get people to slow down. Note what happens when the traffic signals go out after a power outage. The same thing would happen if we removed them permanently. In fact, we don’t even have to go through the bother of removing them. Just turn them off, or have them flashing yellow (i.e. yield) all the time.

    Uncontrolled intersections are generally what makes slow zones self-enforcing everywhere else.

  • Andrew

    Note what happens when the traffic signals go out after a power outage.

    In my experience: drivers on wide streets go as fast as they like and don’t stop anywhere. Pedestrians who need to cross the street are out of luck.
    I’m not sure where you get the idea that maniacal NYC drivers suddenly gain sanity in the absence of traffic signals, but it’s nonsense.

  • Joe R.

    The problem is traffic signals are rarely out long enough to provide an adjustment period. It’s sort of like when the first heavy snow hits in early winter. Drivers slide all over the place because they’ve forgotten how to drive in snow. By the second or third snowstorm, they know when they need to slow down (usually anyway although some minority of morons still continue to drive like it’s summer).

    Once traffic signals were off for a week or two, drivers would learn to slow down. It may take being in or witnessing a few collisions to drive the point home but it will happen. It’s happened everywhere else uncontrolled intersections have been used. I’m more than a little tired of this idea that NYC is “special”, so what works everywhere else won’t work here.

    Uncontrolled intersections are a key part of slow zones everywhere else. They incidentally get more people to buy into the concept because very often average travel speeds are improved by allowing traffic to move at a slower, but more steady, pace. If NYC refuses to go along with this, the concept will be doomed to failure. Remember it’s strings of green lights where a motorist knows there’s virtually no chance of a collision which tend to encourage speeding in the first place. Same thing with trying to make lights.

    I’ll grant there may well be an adjustment period where things are “messy”, but bent sheet metal will drive the point home loud and clear that the days of going 50 mph on local streets are over.

    Do you have any ideas (other than saturation enforcement which NYC doesn’t have the money or manpower for on a citiwide basis) to get drivers to slow down? Albany has already tied our hands on speed cameras. That pretty much leaves street redesign as the only viable venue open to us.

  • pol

    Uncontrolled intersections do nothing to help pedestrians on arterials, which are responsible for 60% of pedestrian fatalities.

    Even on narrower roads (one lane each direction plus parking) if there is high traffic uncontrolled intersections don’t help pedestrians on their own.

    Where traffic lights are removed or never existed in the first place the smaller roads generally yield to the larger road, where traffic tends not to stop.

    Removing traffic lights can help some intersections, but is no cure all. You’re really overselling it. Increasing enforcement the same. Automated enforcement isn’t going to do much either. There are no quick fixes.

    What’s required is expensive road reconstructions over the next few decades, improved surface transit and rapid transit options, decreased accommodations for drivers etc…

  • Joe R.

    What’s required is expensive road reconstructions over the next few decades, improved surface transit and rapid transit options, decreased accommodations for drivers etc…

    No kidding. If you dig through my profile you’ll find that I’ve alluded to drastically reducing motor vehicle traffic levels as the only real way to reduce traffic carnage. Of course, as an aside once you reduce traffic levels enough then traffic signals really aren’t needed for pedestrians to cross the streets.

    Everyone here is looking for a quick fix when the real answer is a major change in our culture away from private motor vehicles.

  • qrt145

    I saw it after Sandy in parts of Manhattan that lost power, and it worked. Due to the combination of reduced traffic and nonfunctioning lights, drivers were indeed driving more slowly and taking turns.

    The problem I think is that once trafic reaches a level enough for cars to form a relatively uninterrupted stream at high speed, then they just never yield. (At least until it reaches a highly congested level at very low speeds, in which people may start taking turns again.)

  • sbauman

    Both Golden and Klein have fairly similar districts insofar as unlinked trips that originate or terminate within their districts. Unlinked street trips are roughly 50/50 between motor vehicle and non-motorized modes. For NYC as a whole, 63% of all unlinked street trips are non-motorized.

    The next question is how much would motor vehicle users be affected by a 20 mph speed limit on streets (assuming they obey the new and current laws). That depends on how long the street portion for each trip is. Long trips are likely to take highways which are not affected. Also, trips that cross the City Line are also likely to take highways. I’ve looked at unlinked trips that start and end in NYC. I’ve broken these trips origin or destination by senate districts.

    These statistics for Mr. Golden’s district are: 80% of the trips that start or end within his district are 4 miles or less; 60% are under 2 miles.

    These statistics for Mr. Klein’s district are similar. The 80% decile is closer to 3 miles. The 60% decile is also under 2 miles.

    These are distances are as the crow flies. NYMTC’s 2010-2011 Household Travel Survey lists only the census tracts for trip origins and destinations. The grid pattern means that the actual distance is at most 41% more than the straight line distance.

    The bottom line is that lowering the speed limit to 20 mph, will cost the law abiding motorists an extra 3 minutes for 60% of their unlinked trips that start or end withing these senate districts.

  • qrt145

    I think the time cost of lowering the speed limit is even less if you consider that the actual average speed for a trip is much lower than the speed limit, especially for short trips and when there is congestion (I don’t know how congested those districts are, though). The time stopped at red lights or looking for parking really kills your average speed. One often reads of average speeds close to 12 mph in the CBD, for example. And this includes illegal speeding when the cars are actually moving!

  • Joe R.

    If we do slow zones like everyone else, meaning we get rid of at least traffic signals, even if we still keep stop or yield signs, then travel times may well decrease. In fact, all my reading on the matter seriously suggests that using traffic lights to allow 30 or 40 mph cruising speeds instead of 15-20 mph in practice gains us little to nothing in average speed when you consider that half the time you’re stopped at red lights. All traffic signals really accomplish is to allow cars to drive at speeds which are dangerous to people on foot or bikes.

  • sbauman

    I was careful to define my “cost” for lower speed limits in terms of extra minutes spent moving per trip. If a trip is 3 miles long and the driver obeys the 30 mph speed limit, the minimum time spent actually moving is 6 minutes (3 miles x 2 min/mile). If the speed limit is lowered to 20 mph, the minimum time spent actually moving is 9 minutes (3 miles x 3 min/mile).

    It’s difficult to reliably quantify waiting time during a trip. However, moving time is pretty easy because it’s related to the distance and speed.

  • sbauman

    If you read contemporary literature, traffic lights were not installed in NYC as a safety measure. Cars, unlike horse drawn vehicles, could not negotiate intersections without causing traffic jams. Police officers in booths were first used to establish rights of way for the cars. Traffic signals were installed because there were not enough cops to go around. Traffic signals were installed to decrease motor vehicle congestion.

    The problem with traffic signals is they reduce average speed by half. A 30 mph speed limit is equivalent to 15 mph without stopping for red lights. Traffic signals probably would not be necessary, if motor vehicles had governors that limited their top speed to 15 mph. Ford, Durant, and Chrysler would not have been able to sell cars had their top speed been limited to what was practical in cities.

    The higher speed did not get their customers to their urban destination any quicker. Rather than recognize the cars’ inherent speed limitations within an urban environment, other scapegoats were sought. Streetcars were banished, streets were made one-way, elevated railroads were removed and expressways were built or proposed to eliminate congestion. Nothing has worked.

    The justification for lower a lower maximum speed is safety for pedestrians and cyclists. The first study linking fatality rates to impact speed appeared 15 years after the current 30 mph limit was enacted. The author suggested cars be redesigned to permit survival at higher impact speeds. That was 35 years ago. Automobile manufacturers have not incorporated safety for pedestrians into their designs.

  • pol

    ‘Cars, unlike horse drawn vehicles, could not negotiate intersections without causing traffic jams.’

    Why? Was the issue just how many cars there were?

    I had thought that uncontrolled intersections can work in low volume areas well enough with sufficient accommodations for pedestrians (raised medians, crosswalks etc…) but few signalized intersections in NYC are sufficiently low volume.

  • Joe R.

    Actually, it’s not true that few signalized intersections in NYC have sufficiently low volume to be changed to uncontrolled intersections. Community boards in this city have a bad habit of requesting traffic signals even when traffic volumes don’t remotely justify them. NYC DOT has an equally bad habit of often granting these requests. I’ve seen the number of traffic signals in my local neighborhood in Eastern Queens triple or quadruple in the last decade, despite the fact that traffic volumes haven’t increased significantly, if at all. None of the extra traffic signals are needed. In fact, some of the ones which were here prior to the city going traffic light crazy could have been removed as well. NYC covers a smaller area than Chicago, but has more than ten times as many signalized intersections. All these extra traffic signals have done is to cause drivers to take ALL traffic controls far less seriously. The more traffic lights you put up, the more drivers will try to do things like make lights for the simple reason that doing so saves more time the more traffic signals you encounter. The city has also either inadvertently or intentionally set up many “double stops” where once the light on one block goes green, the one a block away goes red. The end result can be a few minutes to travel 2 or 3 blocks.

    None of this improves safety. Traffic lights don’t mitigate traffic jams, either, because they drastically cut the capacity of a street. Traffic jams should be dealt with directly-by taking increasingly draconian steps to discourage driving until traffic levels fall to such a level that there are natural gaps in traffic for pedestrians/motor vehicles to get across a street.

  • sbauman

    “Why? Was the issue just how many cars there were?”

    There are several factors. Number of cars is important, when one considers their footprint. A pedestrian occupies 5 sq. ft., a bicycle about 14 sq. ft. and a car about 120 sq. ft. A car’s footprint is 24 times greater than that of a pedestrian. If a car’s footprint is to have the same congestion component as a pedestrian, it has to be moving at least 24 times faster or around 72 mph. They don’t so cars contribute far more to congestion than pedestrians. Bicycles do move leisurely at 15 mph, so they are congestion neutral vis-a-vis pedestrians.

    A second factor is maneuverability. Cars are clumsy compared to horse drawn vehicles. They require a bigger turning radius and are as slow as horse carriages when making a turn. They are also wider, requiring more room to bypass when making a turn. The result is every vehicle behind a turning car must travel around 8 mph. That’s below horse and buggy speed.

    A third factor at urban intersections is visibility. Vehicle lanes are approximately 20 feet from the building line (sidewalk plus parking lane). The MUCTD assumes drivers have a +/- 20 degree horizontal field of view and a 1.0 second reaction time. The building line and limited horizontal field of view means that two drivers approaching the intersection from different directions, equidistant from the intersection and travelling at the same speed, will not see one another until they are 55 feet from the intersection. If each is has a 1.0 second reaction time and is travelling at 30 mph, they will first realize each other’s presence when each is 10 feet from the point of impact.

    Slower approach speeds mean a greater distance when they realize each other’s presence. The distance is 25 feet at 20 mph and 32 feet at 15 mph.

    That 32 feet at 15 mph was sufficient to negotiate the right of way without need for a third party in the horse and buggy days. 15 mph is also the average speed obtained with a moving speed of 30 mph, when traffic signals are factored in. Any traffic commissioner who can increase average NYC speeds to 15 mph would be hailed a genius.

  • Joe R.

    Don’t forget that we didn’t have the means to externally limit motor vehicle speeds to 15-20 mph only on city streets back in the horse and buggy days. If we could have, then traffic signals in cities might have been very rare, perhaps only used at very busy intersections, or those with uncorrectable poor lines of sight. Nowadays we can use GPS to determine the local speed limit, then use that information to govern the maximum speed. We don’t even need to do this on every vehicle. Just do it on commercial vehicles to mitigate “privacy” concerns if it were done on all vehicles. In a city like New York with a large percentage of commercial vehicles, the net effect will be the same as if all vehicles were governed.

    I have a feeling in the not too distant future when self-driving cars become the norm, average travel speeds in NYC will indeed rise to 15 mph and beyond. There will no longer be any need for traffic controls. Vehicles will be able to negotiate among themselves and pass intersections at speed. They can also be programmed to automatically stop whenever people are crossing.

  • sbauman

    There has always been the capability of governing a motor vehicles top speed. All it takes is a pushing a lever. What’s new is that this can now be done without the driver’s intervention. It was never implemented because automobile manufacturers believe that the illusion of speed sells. There’s OnStar and similar systems today. They don’t need a governor – they can sound an alarm when a speed limit is exceeded. Has anybody implemented it?

    The self-driving cars are self-contained. They do not communicate with other vehicles either directly or indirectly. Their sensors are visual, radar and sonar. These sensors do not see through buildings. The same limitations of negotiating intersections will constrain the self-driving cars currently under development.

  • sbauman

    ” NYC covers a smaller area than Chicago, but has more than ten times as many signalized intersections.”

    I have no basis for corroborating or denying this assertion. The important parameter isn’t area; it’s the number of intersections and traffic volume. I do know that Chicago has fewer intersections per mile. Their downtown is based on 8 blocks per mile vs. NYC’s 20.

    “Traffic jams should be dealt with directly-by taking increasingly
    draconian steps to discourage driving until traffic levels fall to such a
    level that there are natural gaps in traffic for pedestrians/motor
    vehicles to get across a street.”

    The purpose for lowering speed limits is to give pedestrians a fighting chance when a car hits them. Limiting the number of vehicles does not address this problem. I’d be happy to allow higher speed limits, if cars were designed so that pedestrian survival remained 5% (survival rate at 20 mph) for higher impact speeds.

    The irony of lowering vehicle speed is that it would make congestion no worse and might actually reduce it.

  • Joe R.

    I’m totally on board for lowering vehicle speeds, but only if the majority of intersections have traffic controls removed at the same time. That’s really the primary way slow zones are self-enforcing elsewhere-mainly because it’s suicidal to drive more than 15-20 mph given the lines of sight and typical reaction times.

    And designing cars to allow greater pedestrian survival rates should be done regardless of whether speed limits are lowered or not.

  • Joe R.

    What I’m trying to say is you didn’t have the capability to limit a vehicle to 15-20 mph only in the city, but allow it to run at full speed elsewhere. The car companies probably would have been OK with that.

    For now self-driving cars are self-contained but long-term plans call for networking them together. Still, given typical lines of sight at intersections, it would be possible for self-driving cars to negotiate intersections at 30 to 40 mph even without networking. Cars will realize the presence of other cars, then react, in milliseconds. The primary constraint will be stopping distance. At 30 mph you only need about 30 feet to stop under dry conditions.

  • lop

    ‘What I’m trying to say is you didn’t have the capability to limit a vehicle to 15-20 mph only in the city, but allow it to run at full speed elsewhere. The car companies probably would have been OK with that.’

    Then why isn’t there a lever or button or switch I can hit in my car to limit it to 15 mph? Maybe add a light that goes off outside the car if I don’t have it engaged too.

    ‘Cars will realize the presence of other cars, then react, in milliseconds. The primary constraint will be stopping distance. At 30 mph you only need about 30 feet to stop under dry conditions.’

    You want your car to give you whiplash everyday?

    ‘it would be possible for self-driving cars to negotiate intersections at 30 to 40 mph even without networking’

    Let’s assume you’re right, and that even if not all cars are self driving, it would still apply. Still makes it a hostile environment to anyone on foot or on a bike to have cars moving that fast, so they won’t be able to do that in urban areas. Doesn’t do all that much for NYC.

  • sbauman

    ” At 30 mph you only need about 30 feet to stop under dry conditions.”

    You’re off by a lot.

    Yellow interval duration is based on a braking rate of 10 fps/sec (6.8 mph/sec). That’s roughly half the maximum braking rate on dry pavement without losing control. The stopping distance from 30 mph at 10 fps/sec is 101 feet. The stopping distance from 30 mph at 20 fps/sec is 50 feet. Besides, the self-driving cars don’t have sensors for oil or gasoline spills on the road.

  • Joe R.

    My calculation is based on 0.75 g deceleration ( 35 fps/sec ). Many modern cars can manage that. If we required stickier tires they all could.

  • lop

    NYC is ~305 square miles of land.

    Chicago is ~227 square miles of land.

    https://en.wikipedia.org/wiki/New_York_City
    https://en.wikipedia.org/wiki/Chicago

    NYC has 12460 intersections with traffic signals.

    Chicago has more than 3000.

    http://www.nyc.gov/html/dot/html/infrastructure/signals.shtml
    http://www.cityofchicago.org/city/en/depts/cdot/provdrs/traffic_signals_andstreetlights.html

    But that’s not a real meaningful way to compare them.

  • Joe R.

    Didn’t I also mention having the cars stop for crossing pedestrians? That’s a better environment than today in that you don’t have to wait for a walk signal to cross the street. You just walk into the street and the vehicles stop.

    Your idea of a lever or switch or button still requires manual intervention. The technology didn’t exist for external ways to selectively limit car speed until maybe a decade ago. Now that the technology exists, it should be implemented, at least on commercial vehicles.

    Deceleration rates of even a few g are quite tolerable by the majority of the population with proper restraints. The primary limit here is tire traction, not human ability to tolerate the deceleration rate. 30 feet to stop from 30 mph is hardly whiplash territory. I’ve done it on my bike by modulating the front brake just enough to keep the rear wheel slightly off the ground. A car can do better because there isn’t the possibility of doing an endover. Don’t forget also with AI getting better all the time it’s quite possible to anticipate potential hazards in advance, like maybe a child who looks like they may run into the street. When these potential hazards are noted, the vehicle then slows down just enough so it doesn’t need full braking capability to stop if the hazard pans out.

  • Joe R.

    The numbers get even more skewed when you compare NYC to cities overseas. London has a similar population to NYC, about twice the area, and still only 6,000 sets of traffic signals ( http://www.dailymail.co.uk/news/article-1361295/Traffic-lights-numbers-increase-30.html ). This was with an increase of nearly a third in 8 years.

    There are lots of metrics to compare, but anyway you look at it NYC has grossly overused traffic signals. There is no need to have signals every 250 feet, even from a pedestrian crossing perspective. One signal every ten blocks are so would provide enough of a ripple effect to allow gaps in traffic for people crossing at the unsignalized blocks, for example.

    We should also seriously reexamine whether we really need a grid based on 20 blocks per mile. This grid was designed back in the days of horse-drawn carriages. One block more or less represented about half a minute of travel at the typical 6 mph of a horse and cart. A grid with 4 blocks per mile would be equivalent to that today with 30 mph travel speeds. You can of course still have a pedestrian/cycling grid of 20 blocks per mile. By making 4 out of every 5 blocks off limits to motor vehicles, you eliminate the need for traffic controls at those intersections. Pedestrians/cyclists would just cross when there’s a gap in traffic. This gap would conveniently be provided by the signalized intersections spaced 1/4 mile apart.

  • lop

    Sorry. Hostile was the wrong word. Unwelcoming or unpleasant perhaps. At 40 mph even reasonably aerodynamic small cars (eg. prius) are unpleasant for nearby pedestrians. The noise, the whoosh of air as they pass. It’s something that should never have been introduced to city streets. Larger cars, SUVs and trucks are of course much worse.

  • sbauman

    “My calculation is based on 0.75 g deceleration ( 35 fps/sec ).”

    Not on this planet :=)

    http://en.wikipedia.org/wiki/Gravitational_acceleration

    Gravitational acceleration is 32.174 fps/sec. It’s rounded down to 32 fps/sec for convenience. 0.75 g is 24 fps/sec, which is close to the 20 fps/sec figure I gave for a controlled panic stop (without stickum).

    This means that stopping distance (after applying the brake) is between 50 and 100 ft. You had better make that 100 feet @ 10 fps/sec because the car immediately behind will not be self-driving nor have stickum on the tires. Its myopic driver will require his full 1.0 seconds of reaction time.

    Any surface treatment, like the sticky substance, will wear off with time. It’s not good engineering practice to base stopping distance on the best conditions.

    A 30 ft. stopping distance requires a constant braking rate of 33.75 fps/sec. That’s 75% of 45 fps/sec.

  • lop

    ‘A grid with 4 blocks per mile would be equivalent to that today with 30 mph travel speeds. You can of course still have a pedestrian/cycling grid of 20 blocks per mile. By making 4 out of every 5 blocks off limits to motor vehicles, you eliminate the need for traffic controls at those intersections. ‘

    ‘One signal every ten blocks are so would provide enough of a ripple effect to allow gaps in traffic for people crossing at the unsignalized blocks, for example.’

    We aren’t clearing out NYC neighborhoods and rebuilding from scratch with a new grid. So how about trying concrete examples with what’s on the ground today. Can you point to a specific neighborhood with specific traffic lights spaced every 250 feet where it would be appropriate to remove them?

  • Joe R.

    I don’t take issue with that. My whole post was just to point out that it would be theoretically possible to have average travel speeds well above 15 mph in NYC once you took the human out of the loop. Whether or not it would be desirable is of course another subject.

  • Joe R.

    Can you point to a specific neighborhood with specific traffic lights spaced every 250 feet where it would be appropriate to remove them?

    All of the Manhattan Avenues for one. I base this on local arterials here in eastern Queens which sometimes have similar levels of traffic to the Manhattan Avenues during peak hours. Even in the places where lights are spaced 4 or 5 blocks apart, you can usually easily find a gap in traffic on the blocks with no lights once the signals turn red. Sure, you’ll get some traffic consisting of cars turning onto the arterial from side streets, but you won’t get a continuous traffic stream which prevents you from crossing.

    You can also eliminate traffic signals at any intersections where you can go from one side of the street to the our via the subway mezzanine (i.e. without entering the controlled fare zone). For example, these function nicely on Queens Boulevard as defacto underpasses.

    In my example we don’t need to build the grid from scratch. The grid would stay in place, but four out of every five cross streets would have bollards to keep cars out. That’s it-just bollards. The city can certainly find the budget for that. They’ll more than make up for it in the money saved operating/maintaining the traffic signals are removed.

  • lop

    ‘All of the Manhattan Avenues for one.’

    High traffic streets have a poor rate of compliance with non signalized pedestrian crossings and are a danger to those not in cars. If only 1 out of 5 cross streets has cars then those streets would often bring a near continuous stream of turning drivers. This would not increase safety, it would do just the opposite, and instead of being able to cross every 250 feet with a modicum of safety you would have that only every 1250 feet, and with the concentration of turning cars those crossings would be dangerous as well. At which point you would need to shut all lights red to give a dedicated pedestrian phase, which would lower average traffic speeds that you want to increase. To get away with uncontrolled intersections on high traffic streets, or on midblock crossings on high traffic streets, some places install simple crosswalks for pedestrians. This gives them pedestrians a false sense of safety and has been shown to increase the danger to pedestrians. So additional measures are needed. Constant blinking yellow lights are used in a lot of places, but don’t have a good track record for compliance. People get desensitized to the lights and ignore them. Some places have pedestrian actuated ones, which are better, but still have a lower compliance rate than standard traffic lights. Not to mention, if you allow them to be used when the next light is green then you’d get gridlock, as there would be enough pedestrians to keep the cars from moving. So you’re in essence giving a pedestrian walk phase not be default, but only when a pedestrian pushes a button. Hardly ideal, and would be a decrease in pedestrian facilities in the city, especially since they are no safer than a pedestrian walk phase when pedestrian traffic levels are high. You can add more signs to the road telling people to watch out for pedestrians, markings saying where to yield if you see one, but it’s still not safer than a controlled intersection. The only other options are shortening crossing distances by removing lanes, introducing raised crosswalks etc…neither of which is cheap, and the latter of which would serve to slow travel speeds, which seems to work against your goal.

    ‘You can also eliminate traffic signals at any intersections where you can go from one side of the street to the our via the subway mezzanine’

    No you can’t. Too many are not handicap accessible. And sending pedestrians into a dark subway mezzanine is off putting and dangerous. You’d need to build elevators, fix them up a lot, put in a cop in many neighborhoods for people to feel safe etc…

    That’s a lot of money to let cars keep moving. You’d be turning QB and other roads into highways essentially.

    ‘but four out of every five cross streets would have bollards to keep cars out.’

    Few would tolerate that. It means your car is far away, and for there to be room for the cars then the ‘car streets’ would have to be lined with parking garages. Perhaps what you meant was line bollards, install tire shredders, put in concrete dividers etc… to keep cars from using them as through streets.

    Where in eastern Queens are there lights every 250 feet? You can pick your neighborhood if you want, I’d assume you know it well. Which would be car through streets, and which would be car local streets?

    You have to assume a pretty slow crossing speed to accommodate elderly and youth. That means a large gap in traffic. Much larger than is needed for someone who is rather fit, as I take it you are.

    You want to remove traffic lights and replace them with uncontrolled intersections to slow cars down but then remove most streets from car use? Wouldn’t that defeat the purpose?

  • Joe R.

    I ultimately think we should drastically decrease traffic levels so we don’t have to rely on kludges like traffic signals to let pedestrians get across streets. Read my post at the very bottom of this thread. We’re looking for quick fixes here but none exist. My problem with traffic signals doesn’t stem from the fact that they slow down motor vehicles. Rather, it stems from the fact that they dramatically slow down walking and biking if obeyed. That in turn results is extremely low compliance rates by cyclists and pedestrians, which in turn defeats the purpose of traffic signals in the first place.

    The heart of the problem here is you really can’t efficiently accommodate high levels of motor and non-motor traffic no matter what you do. What you end up with, which is what we have now, is a slow, inefficient, dangerous clusterf*ck. Of course, you could put cars on one level, pedestrians on another, and bikes on a third, but that’s not happening in any scenario I can see. At best, *maybe* we’ll eventually be lucky to get a few grade-separated bike lanes in the most crowded places. Therefore, the only viable option left is to drastically reduce motor vehicle volume to the point traffic signals are largely no longer necessary. The lower traffic volumes would tend to encourage speeding, but using uncontrolled intersections would counteract that. Or if traffic volumes drop enough you can close many streets to cars altogether to keep traffic density up (and hence speeds lower), signalize only the car streets, and have the unsignalized streets for bicycles/pedestrians.

    Anyway, there’s little argument we need to do *something*, but I don’t feel anything being proposed is going to make much difference. Actively discouraging motor vehicle use is the only thing which will because in the final analysis what we have is a traffic volume problem. Everything else is just a symptom of this.

    Incidentally, since you mentioned pedestrian activated lights, I tend to agree they might cause issues when lots of pedestrians are around but off-peak NYC should put all traffic signals on pedestrian and vehicle sensors. There’s no excuse for dumb-timed signals at midnight which force vehicles to stop 99% of the time for thin air. That’s just dumb, lazy engineering. For that matter, even during the day, if nothing is crossing a street then that particular traffic signal should never go red. Traffic signals should work as follows: If the pedestrian or vehicle sensor detects something on a cross street during the day, then the light on the main street goes red on the next timed cycle. It stays red for the entire normal red light cycle, even if the intersection clears. If nothing is detected on the cross street before the red light cycle would begin, the light on the main street remains green. At night is a different story. If the sensors detect something on the cross street, then the light goes yellow and red immediately on the main street to give the cross street the green. As soon as the person or vehicle crosses, the light on the main street goes green again.

    In practice what will likely happen is during the day things will work more or less like they do now. Traffic signals will operate in their usual phases, waiting times will be exactly the same. At night however the main arteries will seldom see red lights. Indeed, after midnight you may well be able to drive miles without needing to stop. However, occasionally you may need to stop, but just for a few seconds until a passing vehicle clears the street. This would be a boon to everyone, cyclists and pedestrians in particular. On main arterials they’ll seldom have a red light. On cross streets the light will typically go green for them within a few seconds.

  • Joe R.

    Sorry, you’re right. I inadvertently did my calculations for 1g deceleration. Most cars are incapable of decelerating at 1g. 0.6 to 0.75 g seems to be the upper range for most vehicles on dry pavement. 0.75g yields 40 feet, not 30 feet. That’s what comes from playing with numbers when you’re tired/aggravated (long story of some personal issues recently).

    I kind of disagree on the 1 second reaction time. I’ve been able to avoid hitting car doors while cycling which were opened less than ten feet in front of me. At my normal ~20 mph cruise speed that implies a reaction time of no more than 0.3 second.

    I still recall stopping in 30 feet from 30 mph on my bike. I was practicing panic stops on my block. I hit the brakes right when I passed the back of a particular parked car, and noted where I finally stopped. Then again, a bike has a lot stickier tires than your typical car. And doubtless the wind drag helped a lot also. Wind drag alone adds about 0.05g to a cyclist’s deceleration rate at 30 mph.

  • qrt145

    Sure, someone who is REALLY paying attention and has good reflexes and already has the foot on the brake might have a reaction time of even 0.2 s. But we have to design for the real world, consider typical drivers, and then add some slack on top of that. I think 1 s is much closer to the mark.

  • lop

    Standards for perception-reaction time used to call for 0.75 seconds before brakes are applied, typically that is seen as outdated and an unsafe assumption. At least one second is usually used, more often 1.5 seconds, or higher to include elderly/intoxicated/distracted individuals.

    Unless you are going to start taking licenses away from a substantial fraction of the population you can’t get that below 1 second as your safety margin without a lot of crashing.

    What works 99% of the time isn’t good enough here either. You pass through a lot of intersections. You can’t have such a low bar for safety. Which means you have to take into account one in a million freak events that increase stopping distance to even 150 feet at 30 mph.

    And if you plan on putting these intersections in cities then go to hell. I don’t want cars screeching to a halt every few seconds going through an intersection every time I go outside. And who’s going to pay to sound proof everything everywhere so it doesn’t bother people who are inside too?

    I don’t care what’s driving the car, 40 mph on a city street is too damn fast. 30 mph on narrower streets too. It doesn’t belong in a city. Some little kid or dog wanders out from behind a car or other obstruction and then gets hit and dies all so some hick can live 50 miles away in the exurbs but still benefit from living in a city and all without ever having to be near another person, say on a train. And even if you think you can engineer around ever little safety issue, and replace every single internal combustion engine with something clean like an electric motor you still have to deal with the noise of an object moving at excessive speeds. Self driving cars aren’t the savior of cities. They are being pushed as the savior of the exurbs, designed to reduce the biggest flaw with them – a commute that is just god awful. If you want better cities, that means eliminating cars, not making them easier to use.

    To improve intersection throughput the general call hasn’t been to have the cars communicate with each other, but rather with the intersection itself, which would instruct them to accelerate or decelerate to eliminate potential conflicts.

  • Joe R.

    To improve intersection throughput the general call hasn’t been to have the cars communicate with each other, but rather with the intersection itself, which would instruct them to accelerate or decelerate to eliminate potential conflicts.

    Whatever. My point is whether the vehicles communicate with each other, or with the intersection, with self-driving cars lines of sight eventually won’t matter. You can have vehicles coordinate movements through an intersection so screeching panic stops aren’t needed. Maximum speeds are another matter entirely. In theory it would be possible to have self-driving cars moving through intersections at highway speeds without colliding. In practice we probably don’t want to do that if pedestrians or cyclists are around, or there would be noise issues from doing so.

    Self driving cars aren’t the savior of cities. They are being pushed as the savior of the exurbs, designed to reduce the biggest flaw with them – a commute that is just god awful. If you want better cities, that means eliminating cars, not making them easier to use.

    I’m a big advocate of getting rid of personal automobiles altogether in cities, or at least restricting them to highways. Read my other posts if you don’t believe me. Unfortunately, the political reality is that will be difficult to do in the short term. In the long term, perhaps, but not in the short term. Given that reality, I’d much rather have clean electric self-driven vehicles in my city than polluting human-driven ones. And unlike human-driven vehicles, these ones will drive at or under whatever speed limit we set, whether it’s 40 mph, 30 mph, or 10 mph.

    My second point is with self-driven vehicles, we can in a manner of speaking have our cake and eat it too. We can have higher average speeds than today because you don’t need time-wasting conventional traffic controls at intersections, and we can have increased pedestrian safety, plus a far less hostile environment, by reducing peak speeds from the 40 to 50 mph of today down to 25 mph or less (except on highways or streets where 25+ mph doesn’t cause sound/safety issues).

    On the issue of why higher average vehicle speeds would be beneficial, it’s mainly not so some hick can live 50 miles away and work/play in the city. A person like that can go to hell as far as I’m concerned. Rather, it’s for efficient movement of goods/services. Delivery vehicles will make their rounds faster (and therefore deliveries will cost less), buses won’t be delayed by red lights, and emergency vehicle response times will dramatically decrease. Cyclists and pedestrians won’t be delayed by red lights, either. These are all things which primarily benefit city residents.

  • lop

    ‘Cyclists and pedestrians won’t be delayed by red lights’

    How exactly? Unless you ban cars, trucks, buses, emergency vehicles etc…that can’t happen without causing gridlock in any area with a lot of pedestrians.

    ‘Rather, it’s for efficient movement of goods/services.’

    Then instead of waiting 50+ years for enough cars to be fully automated to be able to ban the remaining human operated cars just put bus lanes on all multi lane roads in the city, give buses and emergency vehicles at least basic light priority (keeping lights green). Either install bollards or use cameras to keep non bus/emergency vehicles out of the bus/emergency lane. Require all businesses to have a freight plan, and accommodate them by renting out street parking to them for at least part of the day, some days of the week as needed. Raise license and registration costs to reduce the number of drivers, require all cars operated in NYC to have speed governors to limit the speed of the car, controlled by the operator is fine, with some light outside the car go on to make clear to police if it’s turned off on city streets. You can get 90% of what you think automated cars can bring right now. But you won’t get all of that now. And you won’t get all of that in 50+ years when automated cars are widely available. Technology doesn’t solve your problem. Because it’s not a technological problem cities face but rather a political problem. Or a geometry problem if you prefer. Automated cars won’t take up less space to park than regular cars. And won’t use the roads orders of magnitudes more efficiently. And won’t be able to reduce the number of cars on the road all that much if everyone still wants to be in their own private vehicle and drive to work themselves at the same time.

    ‘I’m a big advocate of getting rid of personal automobiles altogether in cities, or at least restricting them to highways. Read my other posts if you don’t believe me.’

    You seem to be a big advocate of automated cars and streets without traffic lights. Neither works in a dense city.

    You say you want to get rid of (most) personal cars though. That means living in dense apartment blocks for most people. The rich can keep their houses. And those who will tolerate a long train ride, with most of those needing too a long walk or drive to the train station. Everyone else has to be crammed in with a bunch of other people, otherwise transit cannot run often enough to allow people to tolerate living without a car. And even then it will usually need to be surface transit. Which means you still need red lights to keep pedestrians and cyclists out of the way. Or else the transit vehicles would be too slow.

  • Joe R.

    ‘Cyclists and pedestrians won’t be delayed by red lights’

    How exactly? Unless you ban cars, trucks, buses, emergency vehicles etc…that can’t happen without causing gridlock in any area with a lot of pedestrians.

    First off, I think we can agree on a conceptual level that traffic lights would no longer be necessary if vehicles were automated. The primary purpose of traffic signals is to keep cars driving at more than ~20 mph from colliding with each other. If they can do that on their own, or if you can reliably keep speeds under about 20 mph (even with human-driven vehicles) then you don’t need traffic signals, at least for the purpose of keeping vehicles from colliding.

    Allowing pedestrians to cross the street has always been a secondary purpose of traffic signals. It seems to me anyway that someone figured “hey, since we have to stop traffic periodically to allow cross traffic to go, that might also be a good time to let people cross the street”. It was cost-free in terms of time in the sense that the vehicles were stopping anyway for other vehicles.

    This secondary purpose of traffic lights has always come with a caveat-namely that it’s considered necessary to signalize an intersection, even if the cross street has little traffic, if there are seldom gaps in traffic long enough to allow people to cross. Here again automation neatly solves that issue. For one thing, with automated vehicles traffic in general will decrease even if the number of car trips in the city doesn’t. Why? Automated vehicles can park anywhere after dropping off their passenger. About 50% of the traffic in many areas is vehicles looking for parking. Automated cars will also alter the paradigm of vehicle ownership. Many people will opt to use automated taxis rather than own their vehicles. This decreases parking needs and traffic in general.

    Next, I’ve little doubt whether we automate vehicles sooner or later that the long term trend in traffic levels in cities will be down. We may implement congestion pricing, reduce parking, etc. That means less traffic, and more likelihood of gaps in traffic long enough to cross a street.

    The real kicker here though is based on everything I’m reading, there will be smart intersections and sensors which will basically allow a pedestrian or cyclist to get cross a street confident that the vehicles will stop or yield for them. Lower general traffic levels, plus the ability to determine less congested routes on the fly, would mean even streets with heavy pedestrian traffic won’t cause inordinate delays. If there’s a steady stream of people crossing a given street at a given time, in a few minutes cars will stop going down that street. Those that are already there may back away slowly and find a different route. We can’t do any of this right now, but we can certainly do it in less than 50 years. We can do large parts of it in under ten. We can do most of it in 20.

    Finally, in some cases there may indeed be dense pedestrian and vehicle traffic. In that case nothing works well except grade separation, as costly as that may be. That’s why subways were built. I’m confident though that the need for grade separation will be very little.

    You seem to be a big advocate of automated cars and streets without traffic lights. Neither works in a dense city.

    Neither do traffic lights, especially for pedestrians and cyclists. Traffic lights can often reduce average cycling speeds to walking speeds, rendering cyclist pointless. They can easily cut average walking speeds in half. I’m fine with delaying private cars in cities to give others priority, but delaying pedestrians and cyclists for cars is frankly insulting.

    And even then it will usually need to be surface transit. Which means you still need red lights to keep pedestrians and cyclists out of the way. Or else the transit vehicles would be too slow.

    Fine, so you make an exception to my philosophy of not delaying cyclists or pedestrians with red lights except for transit vehicles. I’m fine with that because on average the delays will be both very rare, and very short (how long does it take a bus to pass?). That’s the only reason then red lights might exist-to give transit/emergency vehicles priority over everything else. The rest of the vehicles could negotiate amongst themselves at intersections, and give pedestrians/cyclists priority all the time.

    I would be on board for any and all of your other ideas as well but most of those also face political obstacles. Just getting businesses to have freight deliveries at night has been a nonstarter, for example. You’re 100% right it’s a political problem mostly, but it’s also a technological one. If technology increases average vehicle productivity, which it would, then that means we can have fewer vehicles/lower peak speeds with no downsides. People are receptive to new ideas when it means they don’t have to give up anything. That’s the reality. The minute some have to give up convenience, particularly the rich, vocal driving class, it becomes an uphill battle.

  • lop

    ‘We can do large parts of it in under ten. We can do most of it in 20.’

    No. The announcements about the google car and others are overblown. All of the existing ‘automated cars’ require a person to take over at some places, such as complicated intersections. Even the recent stories in the papers about how the google car has mastered city driving is ridiculous. It hasn’t. They have it playing nice on a single relatively calm street in a single suburb, and it has driven too few miles on it to know . In ten or twenty years there may be a technical capability to allow cars to drive themselves 99.999% of the time, but that isn’t enough, and even that may be wildly optimistic, and these cars will cost at least an order of magnitude more than a comparable human operated car. You can’t get rid of 90% of cars to make the automated ones affordable unless people stop having to go to work at the same time and are willing to tolerate not having a car on demand. Which isn’t acceptable to too many people.

    ‘First off, I think we can agree on a conceptual level that traffic lights would no longer be necessary if vehicles were automated.’

    No. Delays would be too great for cars in dense environments if pedestrians and cars moving in cross directions aren’t sufficiently bunched together, which there is no reason to expect they would be able to be in dense cities. At best you increase the traffic levels you need to have for traffic lights to be justified.

    ‘Many people will opt to use automated taxis rather than own their vehicles.’

    A world in which you can press a button or make a call or send a text and pretty soon there’s a vehicle outside waiting for you. Sounds great doesn’t it? We already have that. People will still want their own cars.

    What you still don’t get is that cars are big. Too big for cities, unless you bulldoze half of your downtown to put in parking garages, and entire neighborhoods on the way to put in highways. And guess what, lots of places did this, or weren’t built until cars came around. And all the talk of how great automated cars are going to be are for those places. Not for cities. Because cars aren’t magically getting smaller. They are still too big and they still don’t fit in cities. Automating them helps on the margins, but doesn’t get you half as much as putting someone on a bus. Which you can’t do if they live on an acre on a cul-de-sac a mile and a half walk from the nearest arterial, with so few people nearby that you can never run buses for them at acceptable frequency. Cars can never be more than a plaything of the rich in an urban area, or else they take over and destroy the city. Nothing changes that.

    More locally, destroy suburbia, including much of eastern Queens, putting everyone in apartment blocks, or make the ones who don’t want that sit through a long train ride/trip to the train station, or else you can’t have a pleasant city with low car volumes. Everything you push about how automated cars fix everything works only if almost nobody uses them and everyone else is on a bus or train, or the area they operate in is too sparsely populated to be called a city. We can have all that now. But people like their yards.

    ‘but delaying pedestrians and cyclists for cars is frankly insulting.’

    If you delay cars for pedestrians and cyclists, it doesn’t matter if they are automated or not, you still have to deal with the political opposition from suburbanites. There’s no working around that. You can’t have this many people taking cars into a city. It doesn’t matter who or what is driving them. And it doesn’t matter if they drop someone off like a taxi and then go give someone else a ride. It doesn’t reduce traffic volumes anywhere near as much as is necessary to get to the point where you can have cars stop for every single person who wants to cross the street without forcing the pedestrians and bikes to bunch together with traffic lights.

    ‘ it’s also a technological one. If technology increases average vehicle productivity, which it would, then that means we can have fewer vehicles/lower peak speeds with no downsides.’

    The blight of cars in cities is emphatically not a technological problem. Not at all. Not one little bit. You automate cars. What happens? You increase road capacity. More people can start driving in to the city and live further away because they can just sleep or read on the car trip. And all that’s happened is you’ve increased the size of the group pushing for more accommodations for cars at the expense of people who live in the city. Automated cars helps the suburbs and the exurbs. Not cities.

  • Joe R.

    So what’s your answer here? Fewer cars in cities? I said I’m already on board for that, indeed even prefer it to any other present or future solution.

    It really sounds like you’re saying people like their space, which implies they need cars because alternate solutions don’t work well in less dense areas. They’ll continue to have the political clout to be allowed to drive into cities, and therefore pedestrians/cyclists will continue to have to tolerate ridiculous delays at traffic signals (or risk getting ticketed when unsurprisingly ignore them). If so, then there are no solutions. We’re stuck with what we have and that’s it.

    All I’m saying is assuming there’s nothing we can do to reduce the desire of people to live in suburbs and drive into cities, then automated vehicles can mitigate the problems we have somewhat. They were never a panacea. The best solution is indeed to get rid of private autos in cities but how do we get there from here? I’ve heard talk about banning private autos in cities from the time I was born (1962), and yet if anything traffic levels have increased since then, not the other way around.

    The bottom line is from my perspective what we’re doing now doesn’t seem to be working out all that well. Now we’re thinking just lowering the speed limit is going to magically make things better. It may increase the survival rate when collisions occur if we can really get drivers to drive at or under the speed limit. But if we do nothing to change the infrastructure that isn’t happening. Saturation enforcement in the entire city isn’t possible. We don’t have the money or manpower for it. Besides that, the police chasing down speeders will create more problems than it solves. Albany already tied our hands on automated enforcement. The only way you can get drivers to slow down then is to increase the negative consequences if they don’t. And I feel uncontrolled intersections will do exactly that. NYC can do that right now, without Albany’s permission. It may even improve average travel speed, although that’s a secondary goal, not a primary one. If average travel speed is improved, or at least not reduced, then slow zones will henceforth be an easy sell politically. That sounds like a win-win situation for everyone, although of course slow zones do absolutely nothing to solve the traffic volume problem. That’s an even bigger problem than the direct carnage cars cause if you ask me. Motor vehicle emissions kill ten times as many people as motor vehicle collisions. You also have the opportunity cost of land used for roads/parking instead of something else.

    Let me ask you this-every other place which has slow zones expressly removes some or all traffic controls within them, so why shouldn’t NYC be doing exactly the same? That’s the primary reason they become slow zones. There’s no script people can follow and know they’ll be safe driving at high speeds. Rather, because almost anything can happen at any time, you can’t drive fast, period. Why won’t this work in NYC? I’m really sick and tired of NYC thinking it’s special. It isn’t.

  • lop

    ‘Why won’t this work in NYC’

    Traffic lights are frustrating for drivers. Knowing a pedestrian might dart out at any moment and you have to yield isn’t an improvement. Knowing that a car can be coming around the corner and you don’t know whether or not they will yield is even worse, because they might do more than ding your car a bit. ‘Naked intersections’ work in small areas, where you only had a few traffic lights, and you take them all out, and after passing through the stressful point car drivers can speed away happy. For pedestrians you are completely giving up on the idea that the area will be welcoming, and the intersections are no less stressful, and play a larger role in dividing neighborhoods than they do today. The stress is a good thing for safety though. People pay attention more. As you put this in place over a larger area though, the additional stress takes a toll. Compliance drops.

    The pedestrian who darted in front of a car a few blocks back left the driver a bit ticked off about how hard he had to slam on the brakes, so he speeds up and honks the next time he sees a pedestrian stepping out trying to chase the guy back to the curb. Or just plows through an intersection because he slowed down a bit and is in a hurry and surely the other guy will stop like he’s supposed to…

    Traffic lights have compliance issues with turning vehicles mostly. Running red lights is much less of an issue. Other traffic controls designed to give pedestrians the ROW have a much lower compliance rate. Cars might slow enough to not crash into each other, but not enough to let pedestrians pass, unless there is a critical mass of them. And even then it is far less pleasant than what we have today.

    Does that look like fun?

    What makes NYC different? Too dense. Too large. Look at how naked intersections play out in similarly large dense cities. Cars get by well enough, pedestrians and bikes do not, and to the extent they do it happens by destroying what’s left of the street as a public space as anything other than a hellscape to be avoided when not in a metal bubble – they become car drivers as soon as they can. And for what it’s worth, when traffic levels are high if you just turn off traffic lights road capacity drops, so travel times increase. If you put in a roundabout that probably goes away though. They only decrease if most of the time the light is red nobody is moving, pedestrian, car, or bike. Which won’t apply to all that many of the 12 thousand intersections in the city during peak hours. Turning traffic lights to just flash yellow at night in some parts of the city wouldn’t be the worst idea though.

    ‘All I’m saying is assuming there’s nothing we can do to reduce the desire of people to live in suburbs and drive into cities’

    Why would you assume that? The city can do much to make driving into NYC less desirable.

  • Joe R.

    All your arguments, and those videos, point to essentially the same thing I’ve been saying-there’s no safe, efficient way for large numbers of motor vehicles AND large numbers of pedestrians to coexist in the space. All of the solutions we’ve tried either don’t make things much safer, or create unacceptable delays for everyone, or both.

    The real answer in cities then is to reduce traffic to the point there are always large gaps in traffic so people can cross streets. Or to just put everything on its own level (my preferred solution but that essentially requires rebuilding the city from scratch). I don’t assume we can’t reduce driving and the need to drive. Indeed, the trends are all pointing in that direction with cities becoming more desirable while the suburbs are waning. However, that’s a long term goal. We can do better than that right now.

    What makes NYC different? Too dense. Too large. Look at how naked intersections play out in similarly large dense cities. Cars get by well enough, pedestrians and bikes do not…

    I won’t take issue that naked intersections are lousy for pedestrians/cyclists when motor traffic levels are very high. However, so are signalized intersections with dumb, timed traffic signals. You’re basically trading the variable, possibly zero crossing delay at naked intersections for the certain delay at signalized intersections. I fail to see how that’s any better. The end result, even with signalized intersections, is people often cross anyway as in those Mumbai videos just to avoid waiting at a red light. I know I do. I prefer to cross in the middle because I avoid turning cars. Often that means running through gaps in traffic, but to me it seems both safer/faster than crossing with the signal at corners.

    That said, the problem isn’t waiting at red lights at the occasional very busy crossing. Rather, it’s the cumulative delays encountering red light after red light which make getting around the city miserable for everyone. Arguably in my experience the majority of the time a traffic signal is red I can safely walk or bike across an intersection. This may not be so in Manhattan during peak hours, but it’s largely true in much of the outer boroughs during most of the day. It’s also true in Manhattan during certain times. And yet the infrastructure and laws would essentially have us stopping for thin air much of the time. That to me is unacceptable.

    So I ask you, if traffic lights may indeed be a kludge solution which works passably well, but not great, when things are very crowded, why not have more intelligent traffic lights which recognize when stopping is not necessary? That technology has existed since the 1960s. Railroad signals have never gone red except when there would be a conflict if they didn’t (the only exception is when dispatchers hold trains in stations with red signals). Why wouldn’t it make sense to apply the same logic to road signals, at least during non-peak times? That could give us the best of both worlds for cyclists/pedestrians. They can get across crowded roads relatively safely, albeit at the cost of a delay waiting for a green signal. When roads are uncrowded, they can proceed more or less unfettered.

    Since you mentioned roundabouts, they’re not a panacea but NYC should definitely use them more often. There are quite a few intersections between 3 streets in the outer boroughs where they would make sense. Right now instead we often have very complex signal patterns (hence long delays at red lights) when a roundabout would seem the ready solution. This holds especially true in areas of the outer boroughs which don’t see much pedestrian traffic.

  • nyctuber

    That’s just great. Way to pummel anyone who drives for a living in NYC into bankruptcy. 30 is not an unsafe speed, 20 is beyond absurd.

  • Ian Turner

    I’m pretty sure the ones who are getting pummeled are not the drivers…

    http://www.streetsblog.org/category/the-weekly-carnage/

ALSO ON STREETSBLOG