Today’s Headlines

  • City Won’t Refund Tickets From 2.9-Second Yellows, Reps Voice Support for Cams (Tribune)
  • Chicago-Detroit High-Speed Rail Would Make 10 Trips a Day at 110 MPH (Free Press)
  • Graffiti Artists Who Vandalized Purple Train Cars Facing Felony Charges (Tribune)
  • Maldonado Transfers Land He Owned by 606 to Wife, Wants to Build Houses There (DNA)
  • Car Crash Near Brookfield Delays Metra BNSF Trains (Fox)
  • Jeff Park P-Street Proposal Moves Forward in City Council (DNA)
  • Survey on Polish Triangle: Let’s Keep the Fountain, Lose the Pigeons (DNA)
  • Hearing on Parking-Lite Logan Square Development This Thursday (DNA)
  • Bridge Closures, Drone Ban Planned for Tightrope Walk Across Chicago River (Tribune)
  • Video: New Bike Cops Undergo Training (ABC)

Get national headlines at Streetsblog USA

  • CL

    I know it’s a crime and all, but the vandalized train cars look pretty cool.

    It’s ridiculous that they’re keeping the money from red light tickets where the yellow was less than three seconds (but not unexpected). It doesn’t matter that the difference is “imperceptible to the human eye” — the tickets are issued by machines, and a split second can be the difference between getting one or not.

  • BlueFairlane

    … and a split second can be the difference between getting one or not.

    Not really. At 30 mph, 0.1 seconds translates to a distance of 4.4 feet, or something less than half a car length. If you get a ticket entering the intersection 2.9 seconds into the yellow, you were going to get a ticket at 3.0 seconds.

  • Kevin

    Well, you could have been 3 ft from the intersection at 2.9 seconds but in it at 3. Though I see your point, especially since most of these the driver entered well after the 3 second mark anyway.

    That said, I think they should have refunded the tickets. Even if all of them are legitimate red light running the tickets need to follow a consistent standard when issued. Quietly changing it back and forth only adds fuel to the fire that the cameras are revenue focused scams.

  • CL

    I’m not sure what the rule is exactly — if you need to be 100% or partly in the intersection at the light change? — but if 0.1 seconds is enough time to travel 4.4 feet, it seems like the exact moment of the cutoff is going to make some difference in the number of tickets issued in the aggregate — even if the behavior of people who just missed a ticket and just barely got one is essentially the same. Whatever the cutoff is, some people are going to just barely miss the ticket by whatever standard they’re using for being in the intersection (part of your car over the line? all of your car?)

  • BlueFairlane

    I *think* the camera only triggers as cars enter the intersection after the red, but I could be wrong on that. Either way, you’re going to have a wiggle zone of 4.4 feet (at 30 mph) that people are going to dispute.

    My thinking with this, though, is that this is an incredibly narrow window that very few cars will fall into. There’s no way of knowing, of course, but the percentage of people who got tickets on a 2.9-second yellow who also fell into this tiny window is miniscule. Those that do fall into the window would have been at least 127.6 feet from the light when the yellow started, leaving them more than enough time to come to a stop and save themselves some trouble. But they made the decision to try beating the light.

  • ohsweetnothing

    Google image searching the tickets (most results used a pic from the Expired Meter, hah.) it looks like the camera takes a pic before you enter the intersection and the light is red, and after you have entered the intersection when the light is red. The ticket also shows how long the light was red when you entered the intersection as well, which seems to be an incredibly overlooked fact to me.

    So if you got a ticket that had an amber time of 2.94 seconds and a red phase time of 0.09 seconds…you may have a case. If you got a ticket with an amber time of 2.94 seconds and a red phase time of 1.4 seconds…get a grip. I guess my question is under which scenario were more tickets getting thrown out?

  • Social_werkk

    I’d love a Chicago to New York High Speed train or Chicago-Toronto.
    We just need a freskin’ national high speed rail network.

  • skyrefuge

    I semi-randomly looked at 29 different citations from 4 different intersections (using the Trib’s citation database: )Those citations were issued before the yellow change interval threshold was lowered to 2.9 seconds, so all citations show a yellow change interval of 3.00 seconds or more.

    The maximum variation at a light was 0.04 seconds (3.01 to 3.05 seconds), with a variation around 0.02-0.03 seconds being more common. So this somewhat supports CDOT’s assertion that the variation does not actually approach the full 0.10 seconds allowed under their 2.9-second standard. It’s very unlikely that many people were cited in cases where the yellow change interval was only 2.90 seconds.

    I then looked at the amount of time the light had been red before the driver entered the intersection. If that time was lower than 0.10 seconds, then such a driver could be cited “unfairly” if the actual yellow change interval was 2.90 seconds. However, in 100% of the cases I looked at, the driver entered the intersection more than 0.10 seconds after the yellow change (the lowest was 0.11 seconds, average 0.66 seconds, and max 2.95 seconds). In other words, no one cited entered the intersection less than 3.00 seconds, or even 3.10 seconds, after the light turned yellow (3.11 seconds was the closest).

    Unknown is whether CDOT had a grace period, where it would allow an entry-on-red of up to 0.10 seconds without issuing a citation (it sort of appears like they did), or whether such a grace period continued during the period of the 2.9-second yellow change interval standard.

    So I don’t think it’s ridiculous that CDOT is keeping the money. I think there’s a very good chance that, even if the yellows on those citations had been a full 3.0 seconds, the drivers all still would have been in violation. What’s ridiculous is that CDOT isn’t explaining that more clearly.

  • skyrefuge

    As I noted in my post above, from my rough analysis, it seems unlikely that anyone would have been cited for an yellow time of 2.94 seconds and a red phase time of 0.09 seconds. First, in the 29 cases I looked at, none had the yellow time off from 3.00 seconds by more than 0.05 seconds, and none showed a red phase time of less than 0.11 seconds.

    But it would be nice if CDOT would just confirm that. Unfortunately the Trib’s downloadable database doesn’t contain the measured phase times, so the best I can do is extrapolate from a small sample (and a sample from before the 2.9-second standard was implemented).

    My understanding from the Trib reporting was that the judges were simply throwing out tickets based on the measured yellow time, regardless of what the red time was (essentially “letting them off on a technicality”). If the Trib uncovered an actual example where someone was cited with a yellow time of 2.94 and a red of 0.09, I think they would have mentioned it, and they didn’t.

  • Did the Tribune ever elaborate on this or just wrote about it and let readers determine?

  • Did the tickets report that many significant digits?

  • “What’s ridiculous is that CDOT isn’t explaining that more clearly.”

    They’re trying, and so is the Inspector General. I was at the Committee on Pedestrian & Traffic Safety meeting on Tuesday for two hours listening to some smart and dumb questions from aldermen about CDOT’s policy on the yellow lights. The meeting lasted longer than I did – I had somewhere else to go.

    What is really a waste of everyone’s time is 1) asking questions not about red light cameras, 2) asking the same question someone else asked, and 3) and not reading the IG’s report, which is written at an upper classperson high school level. The IG’s report is very clear and was better than even CDOT’s own materials and speech at the meeting.

    CDOT, at the meeting, did give a little more as to how the “2.9” seconds thing works. Xerox was actually recording yellow light durations to the third decimal point (1/1000ths) and then truncating the second and third digits left of the decimal point. Thus, 2.999 became 2.9 (there was no rounding). Is 2.999 the same as 3.0?

  • C Monroe

    well hopefully the Detroit station would take VIA trains from Windsor to Toronto trains and vice versa for Amtrack trains. There are two train tunnels connecting Detroit and Windsor.

  • skyrefuge

    Yeah, I agree that the IG’s report was really good. CDOT’s written response was good too, though not as complete and detailed as I would have liked; there might have been more in the appendixes, but those were unfortunately not posted on the IG’s website. I figured 100% of the questions at the council hearing would have been stupid, so it’s nice to hear that there were at least a few smart ones, but I think prepared written communication is just a much better way to get facts like this across.

    The stuff about truncating the 2.999 to to 2.9 was already in the IG’s report (or CDOT’s response). However, CDOT’s implication that the yellow interval variance is within *thousandths* of a second seems to be understating the variance. At least using the old citations, variance on the order of hundredths of a second seemed common.

    Maybe the way the timers are designed/calibrated, they err on the side of “too long”, but again, that’s something CDOT should just detail for us. Since I saw yellow intervals as long as 3.05 seconds, does that imply that light also can vary 0.05 in the other direction, down to 2.95 (which is a lot lower than 2.999)? Or is it biased towards a longer duration, so variance is really 3.03 +/- 0.02?

  • skyrefuge

    Yes, under the old system, all durations were reported to the hundredth place.

    Which seems overly precise, because I assume the times are calculated by counting video frames. If the video is 60 frames per second, that means each frame is 0.017 seconds long, so they can’t even measure to the level implied by that precision. A yellow whose duration was actually 3.0200 seconds could be reported as either 3.02 or 3.03 seconds depending on when the on/off fell within the frame.

    But maybe they have some different method to calculate the time? I hope so, because Xerox would have to be using some kind of super-high frame-rate camera to allow them to measure to the *thousandth*. Or maybe the “truncation” to the tenths place *was* their admission/realization that they were using numbers more precise than their measuring tools.

    Again, CDOT, just lay out your methods. Or point us to somewhere where they’re already laid out.

  • And yet, you are not permitted to take a train across the border. Every time anyone I know has tried in the past twenty years, they take you off the train, search your bags, do customs and immigration, and then bus you.

  • C Monroe

    why not have customs come on the train? That is how it was done in Europe before the EU?

  • Because you don’t actually stay on the same train at all. They have you get off, do customs, get on a bus, be bussed across the bridge, and get on a train in the other country.

    I have no idea why they do it that way, but it’s been like that since before 9/11, so it’s not a “recent security theater” thing.

    In fact, it is sometimes much cheaper (if you are coming from southern Ontario to Chicago) to book a Toronto-Windsor leg, WALK across the bridge, have lunch in Detroit, and get on a separately booked Detroit-Chicago leg.

    The customs are often faster too.