Bacon's Rebellion

Democracy Thrives in Sunlight

Invest More in Smart Traffic Lights

—

by

James A. Bacon
in

Traffic lights are stupid things. We’ve all encountered thoroughfares that create stop-and-go traffic by throwing up one red light after another. We’ve all sat at a late-night intersection, waiting for a red light to change, irritation mounting as we observe that no cars whatsoever are using the cross street. Surely, we’ve all thought, there has to be a better way.

It turns out that the National Transportation Operations Coalition agrees. A new study, the “National Traffic Signal Report Card” concludes that improper signal timing accounts for 5 to 10 percent of all traffic delays, and could be significantly improved for an investment of less than $1 billion annually. (Examiner.com has a brief story here, which I’ve based this blog post on. I could not access the Report Card itself, available here, this morning, probably due to heavy traffic.) Reports Examiner.com:

Some of the biggest problems cited are those that each of us experience on a regular basis, including: (1) signal sequences where drivers pass through a green light at one intersection only to find a red light at the next intersection. (2) making drivers stop at intersections where there are no vehicles and no pedestrians at the cross street. (3) intersections where drivers must sit through more than one green cycle of lights. …

The coalition says that management of traffic signals on a national level rates at “D-,” the operation of individual signals gets a “C” but the coordination among traffic signal systems gets a “D.” The worst grade of all goes to traffic monitoring and data collection, which gets an “F.”

The key to creating “smart” traffic lights is setting up sensors and monitors that can track traffic speeds and the number of cars backed up behind a light, and then adusting the length of traffic signals dynamically — either through human operators at a central station or through artificial intelligence. It’s certainly not a cure-all. But, then, neither is building building more roads. The problem, of course, is that the traffic light lobby isn’t as powerful as the construction lobby, so only pennies are spent where dollars are needed.

Share this article

ADVERTISEMENT

(comments below)

ADVERTISEMENT

(comments below)

Comments

8 responses to “Invest More in Smart Traffic Lights”

  1. Anonymous Avatar
    Anonymous

    “The worst grade of all goes to traffic monitoring and data collection, which gets an “F.””

    How can we ever solve a problem we can’t quantify or understand?

    RH

  2. Larry Gross Avatar
    Larry Gross

    several folks have suggested tolling existing roads rather than raising taxes.

    I would suggest that tolling of existing roads might well be acceptable – if VDOT promised to use the money for “smart” uses – like real implementations of smart signals and direct attention to known bottlenecks…

    AND … they ACTUALLY DID DO IT!

    and… as RH points out.. data collection and monitoring are/would be part and parcel of those efforts.

    as I’ve often advocated:

    quid-pro-quo… you pay a toll and you get something in return for it.

  3. Anonymous Avatar
    Anonymous

    I don’t think there is a very big difference from universal tolling and raising taxes, except universal tolling will cost a lot more.

    Universal tolling does set up a situation in which, if people situate themselves and their jobs carefully, they can reduce the tax they pay. Practically speaking we won’t avoid much because moving is so expensive, jobs change, and everything we buy will cost more.

    Paying a toll gets us right back to the other conversation: what is the right value? We usually like to deceive ourselves that user feees are based on cost of service provided. But in the case of tolls, especially congestion tolls we are intending to charge based on demand. This puts the government in the perverse cabbage patch effect in which it is to governments interest NOT to provide more service so that they can charge more according to demand.

    There is a reason we keep government and privateenterprise separate.

    On rural or exurban roads you may not be able to charge much based on demand, but you could charge more based on cost, and having a route monopoly. You may have different fee structures and different benefits from paying them, based on where you live.

    It is going to be a mess.

    Nobody beleives that you are going to pay a toll and get something in return for it. The evidence that you will is really poor. It is going to be intercepted and redirected just like any black box money.

    Just as with environmental problems the issue is whether you really get the benefits that were promised when the fees were invoked. Without unbiased measurements and agreed standards, we’ll never know.

    We are never going to agree on the measures and standards because they affect our notions of what the outcome SHOULD be.

    You still have the same problems: who pays for benefits, who receives benefits, when, and where.

    Most importantly, after you equalize between those that pay and those that recieve, do you still have a true net positive value? Or do you have income redistribution parading as transit policy?

    Better start collecting that data.

    RH

  4. Anonymous Avatar
    Anonymous

    Don’t invest more in traffic light technology, just get rid of the things PERIOD. There are way too many unnecessary traffic signals in this state, it’s almost beyond comprehension.

    There is a reason that other countries use traffic circles and it isn’t because you can put pretty fountains and gardens in the middle of them, it is because they effectively work to move traffic through an intersection.

    When you put a traffic circle to connect 2 2-lane roads (1 lane each way) you essentially can keep most traffic moving with occasional slowdowns, and I would guess the throughput is in the range of 80-90%, with a max of 100%.

    On the other hand when you put a light in you’ve instantly reduced the throughput to 50% both ways assuming even size roads. So what happens next is traffic starts to pile up and our traffic engineers and pols decide the roads need to be widened to 4 lanes each, just to get back to where we started.
    But it doesn’t end there; drivers start complaining because it’s too difficult to make a left and nearby crossroads aren’t able to get on and off the road. Left turn lights and additional crossroad lights are added reducing the through green time to about 1/3rd of the time meaning traffic stacks up again, and we now have to widen the road to 6 lanes just to move the same traffic as the original 2 lane road. Of course the 6 lane road with turning lanes makes the area unwalkable, increasing the need to drive and forcing the building of lights at every crossing since it’s near impossible to cross without then.
    The moral is it’s all about the throughput, not how big the pipe is.

    Next time you are driving and it seems like you are hitting every red light, it’s because you pretty much have anywhere from a 50-75% chance of hitting one; the law of averages will catch up with you real quick.

    If VDOT wants to make a real dent in congestion, a 50% reduction in traffic signals within 5 years should be a minimum goal. The first ones to come down should be every traffic signal that goes into a subdivision or commercial lot. There are plenty of other alternatives as well: Boulevard style U-turns, small ramps (not semi-based), and circles built specifically for turning around.

    ZS

  5. Larry Gross Avatar
    Larry Gross

    re: traffic circles/roundabouts

    I agree completely but my understanding is that when roads reach and exceed about 25K vehicles per day that circles/roundabouts no longer are effective without signalization.

    Further, trying to “backfit” them in developed areas is very expensive as their “footprint” requires more land.

    Comprhensive strategies to optimize traffic flow and remove bottlenecks should be a major part of what VDOT is about in my view.

    It appears though, thatVDOT is oriented primarily to road-building as the favored solution and most engineers find the difficult process of trying to optimize existing roadways “not” engineering work.

    And they are right, it’s another one of those multi-discipline jobs that requires transportation expertise as well as computer/software modelling and deploying real-time computer control of signals – as “timing” is a real misnomer.

    In order for a group of signals to be able to deal with traffic flows, the lights need to not only adapt to the flow at it’s location but it needs to be able to communicate with the other lights in the area – so that a single light.. optimization.. does not result in adverse impacts to the other lights.

    The reason that I suspect that this kind of thing is not institutionalized is that even for lights that have been “timed”, a new light, often is not planned and integrated with the others and it seems like VDOT put’s the light in.. hears the complaints and responds with a “schedule” when it ought to be standard operating procedure as part of the planning of the new light – to start with.

    This is why I think work like this should be done on a contract basis and the contract can be stopped for non-performance whereas with VDOT .. you get your basic shoulder-shrug from folks whose jobs are not dependent on performance at the customer level.

  6. Anonymous Avatar
    Anonymous

    I was fourth in line for a light today, and it turned yellow before the third car reached the intersection. Maybe there is some reason for it, but ??????

  7. Anonymous Avatar
    Anonymous

    “I agree completely but my understanding is that when roads reach and exceed about 25K vehicles per day that circles/roundabouts no longer are effective without signalization.”

    The 25k is for single lane roundabouts, but it gets up to 40-50k for double lane. There are more complex designs that can go higher than that; still much better than our traditional light every 1/4 mile system. I’m not advocating for traffic circles at major intersections, those need to be separated grade, though you can use a circle in conjunction with that to save the space that a normal cloverleaf takes and not have to build so many lanes.

    “Further, trying to “backfit” them in developed areas is very expensive as their “footprint” requires more land.”

    You can put compact circles that don’t require any more space than a standard intersection. The best example I can think of in the DC area is on Rhode Island Ave up in Mt Rainier. It should be the rule for intersections of that traffic requirements rather than traditional lights.

    Totally agree that VDOT only cares about building it bigger and not actually thinking through solutions. Just amazes me when you look at all the creative solutions for traffic management in other areas and countries and we have VDOT pretty much doing the dumbest designs possible.

    ZS

  8. Anonymous Avatar
    Anonymous

    ZS is correct about throughput.

    It is a well documented fact that the maximum throughput occors at around 25 to 35 MPH.

    If we designed our roads With that in mind, we could have alot more throughput for a lot less money.

    But, We also know that most people will not travel much more than an hour to work. At least not for long periods of time.

    That would imply that for maximum efficiency we need work centers on 25 mile radiuses, or less.

    And, those work centers should be no larger than the carrying capacity of the roads leading to them.

    In addition, multiple work centers will need higher capacity roads leading between them.

    I have located a scientific paper that lays all of this out, with relatively simple equations, and graphics to prove the point.

    The facts are out there. Allwe have to do is find them, recognize thme for what they are, and set aside or preconceived notions.

    Maybe the paper is wrong, but I don’t think so. So far as I know, it is the only paper that explicitly consider job density as part of the land use situation. After I have digested all the equations, I’ll report back.

    RH

Leave a Reply

Read More