Traffic Signal Progression

[tradephoric]

Traffic signal progression is the concept of linking traffic signals together along a street so that "platoons" of vehicles can pass through signalized intersections without getting stopped at a red light.  Some factors that affect traffic signal progression include signal spacing, speed of traffic, cycle length traffic signals run along the corridor, and roadway congestion.
   
While good signal progression along a one-way street should be easy to achieve it can be next to impossible along a two-way street where you often encounter inconsistently spaced traffic signals.  Below are a few examples of streets with good 2-way progression:

Road: Big Beaver Road (Troy, MI)
Start: Coolidge Road
End:  Dequindre Road
Time/Date: Wednesday, July 25, 2012; 10:00 AM
Distance: 5.0 miles
Speed Limit: 45 mph
Freeflow travel time: 6.7  minutes
Actual travel time: less than 6.7 minutes
Delay = 0%

Road: Telegraph Road (Southfield, MI)
Start: 12 Mile Road
End:  Voorheis Road
Time/Date: Friday, 5 PM drive
Distance: 10.0 miles
Speed Limit: 50 mph
Freeflow travel time: 12.0 minutes
Actual travel time: less than 12.0 minutes
Delay = 0%

Signal spacing and the design speed along an arterial are constants when a traffic engineer is trying to provide good progression along an arterial.  The major tool in the toolbox is changing cycle lengths of the signals to provide good two-way progression.  Here are the time-distance diagrams for a 70 second cycle length vs. a 140 second cycle length for two signals spaced 1/2 mile apart with a speed limit of 50 mph:




As you can see a 70 second cycle length provides near perfect dual progression while a 140 second cycle length provides the worst two-way progression possible.  The problem is a 70 second cycle isn't a practical cycle length to run for a number of reasons.

1.  During the heavy rush, even a simple 2-phased signal requires a longer cycle length to reduce delays.  The shorter the cycle length, the greater percentage of time is dedicated to running the yellow and all red to satisfy the safety minimums each cycle. 

2.  Signals with long pedestrian crossings force the signal to run longer cycle lengths to satisfy MUTCD standards.  It's not uncommon to have 130 feet long pedestrian crossings at major 6-lane boulevards with dual left turn lanes.  With this length of crossing it requires 45 seconds of pedestrian times for each thru phase.  Add to that the left turn phases and you're up to a 120 second cycle to satisfy the pedestrian times.

A good case study to look at would be Beach Blvd in Huntington Beach, California.  It's a major 6-lane boulevard with 50 mph speed limits that have signals stopping both directions of travel every 1/2 mile.  Also, many of the pedestrian crossings are over 130 feet long (some over 150 feet) that force the signals to run high cycle lengths.  Beach Blvd from the Pacific Ocean to the 405 is a great example of how not to design an arterial to achieve good two-way progression.

[tradephoric]

I attempt to record drives on days with light traffic so that any delays are solely due to the traffic signal timings.  If it takes multiple cycles to pass through a traffic signal then progression has already been lost. 

Here's an example of a road with pretty poor progression along Jefferson Ave. from downtown Detroit to Grosse Pointe  Park.  This road is a main route for anyone living in Grosse Pointe who works in downtown Detroit and is dotted with traffic signals that increase travel times.  Above each video in bold is the amount of delay experienced due to the traffic signal timings.   A 0% delay would indicate you were able to average the posted speed limit and make it through every green light.  A 40% delay would indicate a drive with a free-flowing travel time of 10 minutes would take you 14 minutes to complete, with 4 minutes of delay as a result of the signal progression.

  Road: EB Jefferson Avenue (Griswold St. to Alter   St.)
  Date: Thursday, July  14, 2011 (4:20-4:35 PM)
  Distance: 6.43 miles
  Speed Limit: 35 mph
  Freeflow travel time: 11.0 minutes
  Actual travel time: 15.4 minutes
  Delay = 40%
 

   
   
  Road: WB Jefferson Avenue (Alter St. to Griswold   St.)
  Date: Thursday, July  14, 2011 (4:35-4:50 PM)
  Distance: 6.43 miles
  Speed Limit: 35 mph
  Freeflow travel time: 11.0 minutes
  Actual travel time: 16.1 minutes
  Delay = 46%
 

[tradephoric]

An example of a road with good signal progression in only one direction of travel:
   
Road: NW Dixie Hwy (Telegraph to White Lake Rd.)
Time: Thursday @ 4:00 PM
Distance: 6.8 miles
Speed Limit: 45 mph
Freeflow travel time: 9.1  minutes
Actual travel time: 9.1 minutes
Delay = 0%

 

Road: SE Dixie Hwy (White Lake Rd. to Telegraph)
Time: Thursday @ 4:15 PM
Distance: 6.8 miles
Speed Limit: 45 mph
Freeflow travel time: 9.1  minutes
Actual travel time: 15.2 minutes
Delay = 67%

[tradephoric]

A one-way street should be the easiest type of road to provide good progression but as you can see in the below video this is not always the case:

Road: Cass Avenue (Johnson Ave. to Huron St.)
Date: Thursday @ 4:30 PM
Distance: 0.60 miles
Speed Limit: 35 mph
Freeflow travel time: 62 seconds
Actual travel time: 187 seconds
Delay = 202%

   
This one-way street might have great progression assuming the signals are running per the traffic signal timing plans.  However, a poorly maintained system can quickly turn a road with good progression into one with the very bad progression.  Here are some factors that can effect the progression along a corridor:

• Signal out of step due to emergency preempt being activated.
  
• Signal out of step due to pedestrian actuation.
  
• Clock in controller is wrong (storm rolled through, bad power in area causing clocks to drift, GPS unit not acquiring data).
  
• Signal out of step due to Dial change in controller.
  
• Broken detection causing side-street to cycle and run long.
  
• Signal timing entered into the controller incorrectly.
  
• Traffic signals along a corridor have different controller types which can lead to clocks drifting at a different rate.
  
• Signal heads turned during a wind storm and now the side-street heads are facing the main-street direction (rare, but quite unsafe when it occurs).
  

[Alps]

I was just in New York City.
* Drove 3rd Ave. northbound from 34th to 85th St. without hitting a red light. (I was at the front of the "wave.")
* More interestingly, drove Cross Bay Blvd./Woodhaven Blvd. north from Rockaway up to Queens Blvd., and hit ONE red light. In fact, I was very clearly at the front of a northbound "wave" of progression. All I can think of is that even though this is a two-way street (and 6-8 lanes at that), because it was a summer Sunday, the city set up a progression to move people off of Rockaway Beach. I was definitely not expecting that, given that most two-way streets in NY have "blocks" of streets that turn yellow/red simultaneously (see Central Park West).

[PurdueBill]

OH 18 from between I-77 in Montrose and I-71 in Medina has some notoriously bad progression now; just yesterday was a prime example.  Platoon of traffic sits at red light at Windfall Road, is released, reaches 55 mph, sees next green light staying green for a long time with cross traffic waiting and nothing on 18 either way, then platoon has to stop as light changes against it.  Repeat 4 more times.  Very frustrating that the cross traffic could be going during the time that no one was coming on 18.  There are loops on 18 approaching the intersections but evidently they don't do anything to help the inability to get a platoon through without stopping over and over.  I drive that stretch fairly frequently and it is very, very rare to get consecutive greens at any time of day.  Being ODOT signals, I would have expected better--usually it's the city signals that are bad at progression.

Akron does very well actually with progression on one-way streets like Cedar/Exchange and High/Broadway, with it being possible to get greens all the way along if you do 25-30, but not so good on two-ways.  Once Exchange becomes two-way west of downtown, it's not uncommon to have to stop at each and every single light.  I'd drive to work on the streets because it's 7 miles instead of 11 on the expressways, but using the streets takes much longer--half of the time driving and half of the time sitting at red lights.

[tradephoric]

Road: Dixie Hwy & Telegraph Road (Metro Detroit)
Start: Dixie Hwy & White Lake Road
End:  Telegraph & Northwestern S.D.
Time/Date: 12/23/12 @ 4 P.M.
Distance: 20.0 miles
Speed Limit: 45 mph
Freeflow travel time: 26.7  minutes
Actual travel time: 27.0 minutes
Delay = 1%

[Road Hog]

I've learned the traffic signal patterns on my drive to and from work (I need to count the number of those signals someday), so I know when to punch the gas to make a yellow and when to let off the gas to get to a green.

Sometimes at an actuated signal I get tripped up by a side-street yahoo, but that's pretty rare.

[tradephoric]

Road: Orange Blossom Trail (Orlando, FL)
Start: Landstreet Rd (1:25 in video)
End:  34th Street (12:25 in video)
Time/Date: Offpeak time
Distance: 5.0 miles
Speed Limit: 45 mph
Freeflow travel time: 6.7  minutes
Actual travel time: 11.0 minutes
Delay = 65%

Comparing the 5 mile drive down Orange Blossom Trail in Orlando, FL to the 5 mile stretch of Big Beaver Rd in Troy, MI (first video of the thread), the driver down Orange Blossom Trail experiences 4.3 minutes more delay (notice that the delay is due entirely to signal progression since the video was taken during light traffic).  A driver who works odd hours driving the 5 mile stretch of Orange Blossom Trail to work and back experiences 35 hours more delay each year than the driver taking Big Beaver to work (8.6 minutes delay roundtrip x 5 days/week x 50 weeks/year).   

INRIX attempts to rank cities with the highest congestion levels each year with their annual Urban Mobility Report.  The report focuses on the delays drivers experience along highways in major metros but appears to entirely ignore the delays a driver experiences on major arterials that are dotted with traffic signals.  Based on the 2012 Urban Mobility Report, Orlando drivers experienced 45 hours of yearly delay per auto commuter compared to 40 hours of delay for Detroit drivers.  Add to that the delays experienced on major arterials where stop lights can significantly increase travel times and you get a clearer picture of true delay in a city.

[tradephoric]

Another smooth drive down a boulevard in Metro Detroit. 

Road: Mound Road (Macomb, MI)
Start: 1500' N. of Utica Road (start of boulevard section)
End:  8 Mile Road
Time/Date: Saturday, Feb. 9, 2013
Distance: 12.9 miles
Speed Limit: 50 mph
Freeflow travel time: 15.4  minutes
Actual travel time: 16.9 minutes
Delay = 9.7%

[tradephoric]

Road: 4th Ave (Chula Vista, CA)
Start: Beyer Way (0:00 in video)
End:  Division Street (23:45 in video)
Time/Date: Unknown
Distance: 6.9 miles
Speed Limit: 35 mph
Freeflow travel time: 11.8  minutes
Actual travel time: 23.8 minutes
Delay = 102%

[tradephoric]

There are a lot of grade separated interchanges along this road but the traffic signals the drivers does get stopped at have long reds.

Road: Friars Rd (San Diego, CA)
Start: Sea World Drive (0:00 in video)
End:  Mission George Road(16:23 in video)
Time/Date: May 23, 2012
Distance: 7.0 miles
Speed Limit: 50 mph
Freeflow travel time: 8.5  minutes
Actual travel time: 16.4 minutes
Delay = 93%