Somebody (Jeff Schneider, I think) commented a day or two ago that many red lights are run by the last car (or cars) trying to make the light. I've been thinking why that seems so. Not only does it seem to be the case, but those last cars through often seem to be accelerating, not holding a steady speed or slowing down.

There was a time (at least in my childhood, if not more recently) when the north/south light turned green at the same time that the east/west light turned red (and vice versa). Somewhere along the way, the traffic engineers added a momentary delay between the red light activating in one direction and the green light activating in the other. Somewhere else along the way, the onset of the green light was further delayed because the pedestrian walk light was activated.

Now we often have something like this sequence of events when a traffic light changes to red:

  1. north/south light turns red
  2. one mississippi
  3. east/west pedestrian walk light changes to "walk"
  4. two mississippi
  5. east/west light turns green

I think what's happening is that the last guy heading north or south as the light is changing is taking advantage of the fact that the east/west traffic won't really be able to go for a heartbeat or two. So, lots of red lights are run.

I don't know why the original green light pause was inserted, too many fender benders when red light bandits got there late and the stop light racers left a touch early? I suspect the pedestrian pause was inserted to give them a fighting chance to cross the street before cars turning right could cut them off. Then there's the added danger of what seems to be frequent red light bandit acceleration. Not to mention which, all those one mississippis must add up to a bit of wasted gasoline when considered in aggregate. Maybe it's time to rethink this whole mess.

Are there any traffic engineers here who can comment?

Views: 1435

Reply to This

Replies to This Discussion

Determining traffic signaling requires decisions trading safety for flow; finding the appropriate balancing point is intrinsically subjective. There are lots of factors that go into these decisions, including:
• the posted speed limit,
• actual traffic speed,
• vehicle types,
• vehicle mix,
• road surface conditions,
• sight distances,
• geometric considerations,
• driver age,
• turning movement

You’ll notice that pedestrians and cyclists aren’t mentioned; for the purpose of traffic signaling, they count as particularly difficult-to-handle vehicles that skew the vehicle mix and also alter the geometry of intersections.

You discussed something called ‘all-red phasing’, where all the lights will show red between cycles. It improves safety (just as you supposed) at the expense of reducing flow.

One big issue is what constitutes running a red light. Most states (including Illinois) have what is known as a ‘permissive yellow’ light, where vehicles may not enter an intersection on a red light but are not in violation if they don’t clear the intersection before the light turns red. This leads to the stacking of vehicles in the intersection, especially when traffic is congested and during left turns, and increases the tendency to accelerate into/through a yellow light. Alternatively, some states have ‘restrictive yellow’ rules, where being in the intersection when the light turns red constitutes running the red. Permissive yellow increases traffic flow (mostly) at the cost of more accidents.

Protected and passive left turn signals also have a significant impact on intersection safety and flow, especially whether or not a protected left turn signal leads or lags the cycle. Combine a restrictive yellow light with a lagging protected left turn (pretty much the opposite of what Illinois does) and you improve intersection safety but at the expense of slowing traffic.

Protected pedestrian crossing improves safety but significantly impedes the flow of other traffic. When protected pedestrian crossing is used, it’s almost imperative that right turn on red be prohibited, that the pedestrian crossing signal leads the cycle and that any protected left turn signal lags the cycle.

Finally, what might be the biggest factor in preventing red light running is the length of the yellow light. Simple things like a slope approaching an intersection make a huge difference in the amount of time required to stop, likewise an asphalt road requires much more time for a stop than concrete and accounting for all of these factors is important. Some communities, in the wake of implementing red light camera enforcement, have experimented with lengthening the yellow light phase and have found that it can reduce the violation rate so much that it makes the camera enforcement unnecessary and even non-viable.

But it all comes back to this: safety and flow are incompatible objectives and finding the appropriate balance between them requires decisions based as much upon values as data.

I've lived in Chicago all 35 years of my life and had a driver's license since 17. I don't think I perceive many more drivers running red lights on the change from yellow ... at least not since I started driving. If there is an uptick I'd put more of that on cars having more horsepower and acceleration than before making it easier to accelerate through a red light. Some drivers can't think more than 3 seconds ahead of themselves so I'm not sure their brains can handle computing the time delay between the cross traffic's light turning green.

Not sure if there are any statistics out there that track how often red lights are run to see if there truly is an uptick. Red light tickets might not be useful since it's dependent on enforcement but better than nothing I suppose. 

As a biker I like the delay. The mouths of most intersections are tight between one lane going each direction, left-turn lane in the middle, and parallel parked cars on both sides. The bike lane doesn't usually start until there's enough width for cars and bikes (10-20 feet in). If it's clear enough for me to see both directions then I'll proceed when all lights are red. This allows me to take the lane entering the mouth of the intersection without having to deal with any drivers trying to cut me off or squeeze through with me. By the time the cars catch up I'm usually entering the bike lane and it's safer and easier for everyone. 

What a response, Reboot Oxnard!  You bring up a good point with "what constitutes running a red light?"  I think most drivers I know believe that if the paint on your front bumper passes the painted stripe on yellow, you made it, unless there is a red light camera at that intersection.  

That would be consistent with my experiences on the road.

Shawn A Conley said:

What a response, Reboot Oxnard!  You bring up a good point with "what constitutes running a red light?"  I think most drivers I know believe that if the paint on your front bumper passes the painted stripe on yellow, you made it, unless there is a red light camera at that intersection.  

That or you could be like Evanston which makes **** up as they go.

Reboot Oxnard said:

Determining traffic signaling requires decisions trading safety for flow; finding the appropriate balancing point is intrinsically subjective. There are lots of factors that go into these decisions, including:
• the posted speed limit,
• actual traffic speed,
• vehicle types,
• vehicle mix,
• road surface conditions,
• sight distances,
• geometric considerations,
• driver age,
• turning movement

You’ll notice that pedestrians and cyclists aren’t mentioned; for the purpose of traffic signaling, they count as particularly difficult-to-handle vehicles that skew the vehicle mix and also alter the geometry of intersections.

You discussed something called ‘all-red phasing’, where all the lights will show red between cycles. It improves safety (just as you supposed) at the expense of reducing flow.

One big issue is what constitutes running a red light. Most states (including Illinois) have what is known as a ‘permissive yellow’ light, where vehicles may not enter an intersection on a red light but are not in violation if they don’t clear the intersection before the light turns red. This leads to the stacking of vehicles in the intersection, especially when traffic is congested and during left turns, and increases the tendency to accelerate into/through a yellow light. Alternatively, some states have ‘restrictive yellow’ rules, where being in the intersection when the light turns red constitutes running the red. Permissive yellow increases traffic flow (mostly) at the cost of more accidents.

Protected and passive left turn signals also have a significant impact on intersection safety and flow, especially whether or not a protected left turn signal leads or lags the cycle. Combine a restrictive yellow light with a lagging protected left turn (pretty much the opposite of what Illinois does) and you improve intersection safety but at the expense of slowing traffic.

Protected pedestrian crossing improves safety but significantly impedes the flow of other traffic. When protected pedestrian crossing is used, it’s almost imperative that right turn on red be prohibited, that the pedestrian crossing signal leads the cycle and that any protected left turn signal lags the cycle.

Finally, what might be the biggest factor in preventing red light running is the length of the yellow light. Simple things like a slope approaching an intersection make a huge difference in the amount of time required to stop, likewise an asphalt road requires much more time for a stop than concrete and accounting for all of these factors is important. Some communities, in the wake of implementing red light camera enforcement, have experimented with lengthening the yellow light phase and have found that it can reduce the violation rate so much that it makes the camera enforcement unnecessary and even non-viable.

But it all comes back to this: safety and flow are incompatible objectives and finding the appropriate balance between them requires decisions based as much upon values as data.

That's what a 'permissive yellow' is - Skip's "red light bandits" are doing exactly what the traffic planners expect them to do.

The 'all-red phase' Skip asked about is a response to the practical limits of lengthening the yellow light. At about 95% compliance, adding more time to the yellow light doesn't actually increase compliance, it starts to reduce it. If a driver stops in response to a yellow light and it doesn't turn red, he tends to learn and the next time won't be so quick to stop. Thus, the all-red phase is an attempt to 'reward' a driver for stopping by presenting him with the anticipated red light while allowing traffic that didn't stop a safe opportunity to exit the intersection.

The light turning green does not give you the right of way. That belongs to anyone who is already in the intersection - you're supposed to wait until they clear.

Shawn A Conley said:

What a response, Reboot Oxnard!  You bring up a good point with "what constitutes running a red light?"  I think most drivers I know believe that if the paint on your front bumper passes the painted stripe on yellow, you made it, unless there is a red light camera at that intersection.  

There are lots and lots of statistics but aggregating and comparing them isn't easy to do and much of the time isn't very useful. With the caveat that there are humongous variations from intersection to intersection (and even at an individual intersection depending upon time of day and weather) that make the statistic almost irrelevant, the national 'average' seems to be about 3 to 5 red light runners per hour. Red light cameras decrease violations and the most dangerous T-bone accidents but tend to increase the less damaging rear-end collisions, setting up an odd juxtaposition: fewer RLR violations but more accidents. It's also clear that a 50% reduction in RLR can be achieved in urban settings simply by increasing the yellow light duration by .5 to 1.5 seconds to a maximum of 5.5 seconds. Any yellow light of less than 3 seconds was set that way to maximize traffic flow (the red light bandits) and not out of concern for RLR violations or safety.

There are two approaches to the problem of red light running: engineering and enforcement. Traffic planners are less concerned with how often a light is run (remember, the trade-off is efficiency for safety: no harm = no foul) and more concerned with how often and what type of accidents occur. The general consensus for considering engineering changes at any given intersection is more than 4 accidents in a month or 6 accidents in two months. When people start focusing on the RLR violation rate they are generally less concerned with engineering solutions to traffic flow and safety issues than they are with enforcement opportunities and the revenue potential that represents.

Not light reading but informative: Engineering Countermeasures to Reduce Red Light Running


Rich S said:


Not sure if there are any statistics out there that track how often red lights are run to see if there truly is an uptick. Red light tickets might not be useful since it's dependent on enforcement but better than nothing I suppose.

While the driver in the intersection has the right-of-way, I suspect s/he will often have attained that position by running the red light... After all, you said the all-red phase thing was to give drivers the opportunity to clear the intersection, not jump into it.

A related, but mostly annoying (not as dangerous), phenomenon happens at slow, congested intersections. (Think North and Wells at rush hour.) People jump into the intersection while the light is yellow, likely knowing full well that they can't clear it before the light turns red, since traffic ahead of them has obviously ground to a complete halt.


Jeff Schneider said:

If only drivers remembered this!

Reboot Oxnard said:

The light turning green does not give you the right of way. That belongs to anyone who is already in the intersection - you're supposed to wait until they clear.

Blocking the box! Illegal, no?

I once saw a bike cop force a motorist to continue straight rather than turn left, as she was attempting to do, when it was obvious that she had pulled into the intersection knowing full well that the backed-up-traffic was not going to allow her to turn left without blocking oncoming traffic for a full cycle. It was awesome.

Skip Montanaro 12mi said:

A related, but mostly annoying (not as dangerous), phenomenon happens at slow, congested intersections. (Think North and Wells at rush hour.) People jump into the intersection while the light is yellow, likely knowing full well that they can't clear it before the light turns red, since traffic ahead of them has obviously ground to a complete halt.

You're right, of course, but as far as I've seen, the ratio of green-lighters to yellow-lighters is pretty close to 1:1. On some spots along Wacker during rush hour, it's probably more like 1:2, excluding taxis (who will soar through a solid red to make a left turn).

It's probably some pedicabber's fault.

Cameron 7.5 mi said:

There are many cases where a vehicle remaining in the intersection when the light turns red could have entered when the light was green. The most common is vehicles waiting in the center of an intersection to make a left turn.



Skip Montanaro 12mi said:

While the driver in the intersection has the right-of-way, I suspect s/he will often have attained that position by running the red light... After all, you said the all-red phase thing was to give drivers the opportunity to clear the intersection, not jump into it.

I have a brother-in-law who used to have a lot of crashes, but he always had the right-of-way . . .

There is a rule in boating that every boater has the obligation to avoid an accident regardless of right-of-way. It is a good rule.

Good rule, indeed.

Lisa Curcio 4.1 mi said:

I have a brother-in-law who used to have a lot of crashes, but he always had the right-of-way . . .

There is a rule in boating that every boater has the obligation to avoid an accident regardless of right-of-way. It is a good rule.

RSS

© 2008-2016   The Chainlink Community, L.L.C.   Powered by

Disclaimer  |  Report an Issue  |  Terms of Service