The light holds traffic back in the left turn when thete is clearly plenty of time for several cars to turn left into Terramont.
I've had an SUV's left tires come well into the left turn lane as I've stood there, stopped waiting for the light to turn.
I remember getting angry because there were more than one opportunity for us to turn. But we just waited like we're were supposed to do, and I could have gotten clobbered for it.
We need to have the option back of being able to turn left when large enough gaps present themselves. Otherwise, it is broken.
Thank you for contacting the City of Roswell Department of Transportation. Your request has been assigned to one of our team members and will be reviewed in the order of priority.
Thank you for your feedback on the flashing yellow operation.
I'm sure most motorists would agree that it was more convenient to have the flashing yellow arrow operating 24/7 as it was previously. Unfortunately, follow-up safety studies conducted by our agency and the Georgia Department of Transportation revealed a significant increase in left-turn conflict crashes that were occurring during the flashing yellow arrow (FYA) phase. We were also advised of the increasing number and severity of crashes by motorists that routinely use the intersection, as well as local residents. FYA operation is permissible when certain favorable operational conditions are present, but becomes contraindicated when crash history directly related to its operation starts to occur, especially when it is increasing. Therefore, GDOT (the permitting authority for SR-140/Holcomb Bridge Road) and our agency changed the operation to protected-only left turn (green arrow) for several hours of the day corresponding to when crashes were occurring. This decision was made in the interest of public safety and was a direct result of driver behavior using the FYA. We will be revisiting the crash history in about 10 months to see if any changes are warranted, but unfortunately things are going to stay the same for now.
I find this explanation to be insufficient in light of the fact that the lights at the Calibre Creek intersection, which is the very next one heading westbound, has operational FYA at all hours of the day. I refuse to accept the fact that the CC intersection can be that much safer when one can see just as far or farther down the road at the Terramont intersection.
After all the highly regarded Georgia DOT and a statistical model agree the simple folk in Terramont (and those who have the misfortune to stray in front of us) are much safer if we just hold back and know our place.
I wonder what factors the model includes? Does it report how many human-hours are expended in this life saving measure? Hey, has anyone added a calculation to predict when and what quantity of citizens would be ticked for running the light?
Models are never corrupt or wrong. They need constant adjustments in orded to align with reality, don't they?
I happen to have worked with risk models for the last 12 years as a product development software engineer. First 4 with a prison population management software, the next 8 with a financial portfolio risk and performance attribution software, data subscription, and research company for institutional and sovereign wealth funds.
These tools should be used to as advisors, and should show several scenarios. Not just kick out one simple binary and be used as an excuse. Maybe you dumbed your answer down for the the croud to hear.
But I don't buy the answer, I don't find it compelling, I also find it insufficient (to quote Wegs).
I wanted to add a picture of the intersection we took today of one of tbe large trafgic gaps that we had to let go because of the red light. The first gap I didn't think to ask for a picture. But the second, third, and fourth we did.
Well first, I actually love the latest feedback on this because its very insightful and intelligent. I also love the fact that the core concept of "risk" and what it really means, and how it should be applied was challenged. The irony is that engineers, scientists and modelers go through myriads of analysis and scenario-playing and what-ifs, only to find out that everything they go through is usually dumbed- down and simplified so that it can be brought forth to the masses for a debate based on horribly watered down truths, or non-truths. And so you are correct, many regulatory standards are pretty simple, and there may be more accurate ways to look at things.
So regarding the main issue, we certainly don't disagree that during the hours the flashing yellow arrow is disabled at this location, there can still be many large gaps available between platoons of vehicles for which most reasonably safe and prudent drivers can navigate. I get it, I'm a driver too, and it aggravates me too when very conservative operational measures are put in place because of the few drivers that refuse to use more due care. We also do not disagree, and in fact, fully agree, that there are many other intersections that we still operate FYA 24/7 where, on the face of it, would appear to be much riskier based on a variety of physical factors such as roadway curvature, sight-distance, driveway friction, etc., but for whatever reason, they did not have an increase in crashes after being converted to FYA.
What I failed to clarify, was that our decision was not based on a model or a GDOT predictive algorithm, but basically on the fact the number of actual left-turn crashes jumped up dramatically in the 12-month period after FYA was implemented as compared to the years before it. I understand that it could be argued that the 12-month period may not be a statistically-conclusive duration of time, or there could have just been a random cluster of crashes, but the change was certainly in the range where some type of mitigation would be expected of us professionally. I would also inviter you to also consider the fact that several people were seriously injured, and it may not have been the ones that failed to use due care.
And lastly, we will certainly revisit the situation in the future. In my former agency back in Florida, we actually reversed some potentially overly-conservative signal control measures implemented by FDOT, and the crash rate actually went down slightly. We theorized the crash rate was related to people running the red arrow because of the wait time to the next one. GDOT is actually pretty liberal regarding when they will try FYA compared to FDOT, so I would not really get on their case -- the are pretty willing to try things.
11 Comments
Department of Transportation (Registered User)
Robert (Registered User)
Yes, I want to "vote" this one up.
The light holds traffic back in the left turn when thete is clearly plenty of time for several cars to turn left into Terramont.
I've had an SUV's left tires come well into the left turn lane as I've stood there, stopped waiting for the light to turn.
I remember getting angry because there were more than one opportunity for us to turn. But we just waited like we're were supposed to do, and I could have gotten clobbered for it.
We need to have the option back of being able to turn left when large enough gaps present themselves. Otherwise, it is broken.
Acknowledged Traffic Engineering (Registered User)
Traffic Engineering (Registered User)
Scott Brunner (Registered User)
Thank you for your feedback on the flashing yellow operation.
I'm sure most motorists would agree that it was more convenient to have the flashing yellow arrow operating 24/7 as it was previously. Unfortunately, follow-up safety studies conducted by our agency and the Georgia Department of Transportation revealed a significant increase in left-turn conflict crashes that were occurring during the flashing yellow arrow (FYA) phase. We were also advised of the increasing number and severity of crashes by motorists that routinely use the intersection, as well as local residents. FYA operation is permissible when certain favorable operational conditions are present, but becomes contraindicated when crash history directly related to its operation starts to occur, especially when it is increasing. Therefore, GDOT (the permitting authority for SR-140/Holcomb Bridge Road) and our agency changed the operation to protected-only left turn (green arrow) for several hours of the day corresponding to when crashes were occurring. This decision was made in the interest of public safety and was a direct result of driver behavior using the FYA. We will be revisiting the crash history in about 10 months to see if any changes are warranted, but unfortunately things are going to stay the same for now.
Thank you again for contacting us.
- Scott Brunner, City Traffic Engineer
sbrunner@roswellgov.com
Wegs (Registered User)
Robert (Registered User)
This really disappoints me.
But it certainly is the "safe" thing to do.
After all the highly regarded Georgia DOT and a statistical model agree the simple folk in Terramont (and those who have the misfortune to stray in front of us) are much safer if we just hold back and know our place.
I wonder what factors the model includes? Does it report how many human-hours are expended in this life saving measure? Hey, has anyone added a calculation to predict when and what quantity of citizens would be ticked for running the light?
Models are never corrupt or wrong. They need constant adjustments in orded to align with reality, don't they?
I happen to have worked with risk models for the last 12 years as a product development software engineer. First 4 with a prison population management software, the next 8 with a financial portfolio risk and performance attribution software, data subscription, and research company for institutional and sovereign wealth funds.
These tools should be used to as advisors, and should show several scenarios. Not just kick out one simple binary and be used as an excuse. Maybe you dumbed your answer down for the the croud to hear.
But I don't buy the answer, I don't find it compelling, I also find it insufficient (to quote Wegs).
Robert (Registered User)
I wanted to add a picture of the intersection we took today of one of tbe large trafgic gaps that we had to let go because of the red light. The first gap I didn't think to ask for a picture. But the second, third, and fourth we did.
Si I uploaded it to this issue:
https://seeclickfix.com/issues/6529447
Traffic Engineering (Registered User)
Scott Brunner (Registered User)
Well first, I actually love the latest feedback on this because its very insightful and intelligent. I also love the fact that the core concept of "risk" and what it really means, and how it should be applied was challenged. The irony is that engineers, scientists and modelers go through myriads of analysis and scenario-playing and what-ifs, only to find out that everything they go through is usually dumbed- down and simplified so that it can be brought forth to the masses for a debate based on horribly watered down truths, or non-truths. And so you are correct, many regulatory standards are pretty simple, and there may be more accurate ways to look at things.
So regarding the main issue, we certainly don't disagree that during the hours the flashing yellow arrow is disabled at this location, there can still be many large gaps available between platoons of vehicles for which most reasonably safe and prudent drivers can navigate. I get it, I'm a driver too, and it aggravates me too when very conservative operational measures are put in place because of the few drivers that refuse to use more due care. We also do not disagree, and in fact, fully agree, that there are many other intersections that we still operate FYA 24/7 where, on the face of it, would appear to be much riskier based on a variety of physical factors such as roadway curvature, sight-distance, driveway friction, etc., but for whatever reason, they did not have an increase in crashes after being converted to FYA.
What I failed to clarify, was that our decision was not based on a model or a GDOT predictive algorithm, but basically on the fact the number of actual left-turn crashes jumped up dramatically in the 12-month period after FYA was implemented as compared to the years before it. I understand that it could be argued that the 12-month period may not be a statistically-conclusive duration of time, or there could have just been a random cluster of crashes, but the change was certainly in the range where some type of mitigation would be expected of us professionally. I would also inviter you to also consider the fact that several people were seriously injured, and it may not have been the ones that failed to use due care.
And lastly, we will certainly revisit the situation in the future. In my former agency back in Florida, we actually reversed some potentially overly-conservative signal control measures implemented by FDOT, and the crash rate actually went down slightly. We theorized the crash rate was related to people running the red arrow because of the wait time to the next one. GDOT is actually pretty liberal regarding when they will try FYA compared to FDOT, so I would not really get on their case -- the are pretty willing to try things.
Anyway, have a nice weekend.
Closed Scott Brunner (Registered User)