QUESTIONS AND ANSWERS ABOUT ACTUATED SIGNAL CONTROLLER OPERATIONS
Mike Boydstun, assistant traffic operations engineer of Ada County Highway District,
answered questions asked by graduate students of NIATT at the University of Idaho unless otherwise noted.
Question 1: how to set up gap reduction using cars waiting before reduction (number of cars waiting before starting gap reduction)?
Answer:
Here's what it says in the Naztec Controller Manual about this feature:
Cars-Before-Reduction (0-255 vehicles) is an alternate method to delay gap reduction after a serviceable conflicting call. This feature applies the total number of detector actuations received during the yellow and all-red intervals to calculate the delay. Gap reduction begins when the total detector actuations exceeds the Cars-B4 value or after the Time-B4 timer expires (whichever comes first). After the Cars-B4 or Time-B4 delay, passage time is reduced to the Min Gap in a linear fashion during the Time-to-Reduce (TTR) period.
Cars-Before-Reduction does not replace Time-Before-Reduction and both are active at the same time. Therefore, set Time-Before-Reduction greater than Max-1 to force the controller to use Cars-Before-Reduction.
The detector option, Added.Initial must also be enabled for calling detector to enable Cars-Before-Reduction.
Note from Lei: for an Econolite ASC/2S-2100 controller (main program version 1.69), in order to activate Cars-Before-Reduction, the option of Lock Memory for a phase’s upstream detector should be set as on in order to log calls during the red time. On the contrary, this option for the stop bar detector should be set as off. Otherwise, calls accumulated on the stop bar will be counted in the parameter of Cars-Before-Reduction. If none of the detectors for a phase has Lock Memory on, then calls would not be counted toward Cars-Before-Reduction unless a constant call exists.
So, I would say that it supplements the TBR function, by saying that if "x" number of vehicle detections are received during Yellow / Red of a phase, go ahead and begin the Gap Reduction immediately, otherwise, wait for the TBR timer to elapse.
This would require a separate detector input for each lane that this feature is enabled on. Otherwise, a single vehicle may place 3 or 4 counts, then rest on the stop bar detector, and not allow the controller to recognize additional vehicles arriving behind the first vehicle. Most likely, this separate detector would need to be upstream as far as possible (within the detector layout) to prevent the queue from stacking onto this detector.
So, to answer your question, the controller obtains the value from a detector in the lane of a phase that this feature is enabled for. The controller counts the number of detector actuations, and compares it to the value programmed as Cars Before Reduction, and responds accordingly.
Question 2: Are vehicle extension time and detector extend time the same? If not, what is the difference between them?
Answer:
There is a difference between Detector Extend time and Vehicle Extension
(Passage) time.
The Detector Extend time actually causes the controller to think that a vehicle is present on a loop longer than it actually is. When the vehicle leaves the loop, the "simulated presence" of a vehicle is continued for the amount of time programmed into this value.
For example, let's say that you program Detector Input # 2 with an Extend time of 3 seconds. When a vehicle drives over the loop(s) that is (are) tied to this input, under normal conditions, the controller recognizes that the vehicle is there only as long as vehicle is on the loop. Once the vehicle leaves the loop, the Passage timer begins timing down. With the 3 second Extend value programmed, the vehicle is still recognized for the entire time that it is on the detector. Once the vehicle leaves the loop, the controller continues to think that a vehicle is still on the loop for an additional 3 seconds, and then allows the Passage timer to begin its count down.
So, the bottom line is that there is a difference between the functions of the Detector Extend and Vehicle Passage functions. If the Detector Extend time is set to zero, the detectors will operate "normally".
Question 3: Typical detector configurations
Answer:
Normally, there is only one lane per channel of detection (sometimes there are multiple channels per lane - I think I'll get to that in answering one of your other questions). This allows for "problems" in one lane to not cause problems for an entire approach. Additionally, features such as detector delay, extension, etc. can be enable easily when there is only one lane per channel. For example, let say that you want to put a 15 second delay on the detector channel for a Right Turn Only lane. If this is the only lane on the channel, you simply program the delay for that channel, and it is done. If ALL lanes of an approach are on a single channel, then the delay applies to ALL detectors on that channel, and it may mean that you end up NOT putting the delay in, and the intersection changes to green for a vehicle that has already left the intersection, as it made the right turn on red.
You can place the stop bar detectors and the upstream (or advanced) detectors on seperate channels. There are limitations as to how many you can "split out". The limitations are: cost, number of controller inputs, and number of detector rack slots. This would allow you to enable the extend only feature that you want to use. Most agencies (actually all of them that I am aware of) in Idaho DO NOT seperate these advanced detectors out, unless they are doing so to use them as counting detectors. Even in this case, most agencies still just use them as both presence and extension detectors, as this is the default setting in most controllers. The same information typically applies for High Speed approaches.
By not utilizing the extend only feature, the controller operation can either be considered more efficient or less efficient. It's all perspective. If a vehicle is approaching the intersection on a side street, is it more ( or less ) efficient to have the advanced detector(s) begin the process of calling that phase to green? If you are on the Main St., it appears to be inefficient; if you are on the side street, it appears to be efficient. That is also where the delay feature comes into play - during normal controller configuration and operation, a 1 or 2 second delay can allow the detectors to work as "default", but prevent the phase from turning green until the vehicle has arrived at the stop bar.
Most agencies leave this "default" programming in place, so that it is easier for maintenance technicians to troubleshoot detector problems. When the techician arrives at an intersection which has been identified as having a detection problem, one of the first things that they will look for is "is the detector amplifier working correctly, and is the call making it to the controller?" In using the extend only feature, the detector input does not show up on the primary controller status display until the phase turns green. The technician would need to go to a status screen strictly for the detector inputs to identify whether the input is actually making it to the controller. If the extend only feature is not implemented at all actuated signals with agency boundaries, technicians can be come confused, a correct troubleshooting becomes much more difficult. The technician may decide to change out the controller, when in reality, the controller is actually working correctly; the technician is just not aware of the way that the controller has been set up.
I hope this information helps, and that it isn't too much information. Let me know if you have any other questions. I will be out of the office until next Tuesday, at the ITE Meeting in Jackson Hole, WY. I normally try to check my e-mail in the evenings, when I can, especially if the motel has a Wireless network that I can connect my laptop to.
Question 4: Causes for phase failure
Answer:
It can be caused either by timing or by driver behavior. Sometimes drivers can allow large enough gaps between themselves and the vehicles in front of them to cause the phase to gap out. This may be due to something on the roadside (signs, crashed vehicles, emergency vehicles, etc.) or it may be a distraction inside the vehicle.
If it is caused by the signal timing, there are a few things that can be done. One is to evaluate the Maximum Green time, and see if the value is correct. It may need to be increased. Another option, if the phase is gapping out, is to increase the Passage value. This can help a particular phase overcome some of the driver distractions and use the entire amount of green time needed to clear the phase. However, you must be careful in doing this, so that you don’t loose too much of the efficiency of the fully actuated signal.
There may be other causes of Phase Failure. Maybe there aren’t enough lanes. Maybe there is a lane utilization problem (i.e. – Let’s say that there is an arterial with two through lanes in each direction and an center turn lane that turns into a left turn bay at the signalized intersections. Everyone wants to be in the left through lane at Intersection A, because the majority of the traffic is turning left at Intersection B). With these types of Phase Failures, signal timing adjustments may not be able to overcome the failure. If the adjustments are able to overcome the failure, what impacts did they have on the other movements at the intersection? Are other movement now failing that weren’t before?
Question 5: serviceable conflicting calls vs. non-serviceable conflicting calls
Answer:
Yes, there are non-serviceable conflicting calls. If a phase is omitted (such as a leading left turn) when another phase is active (such as the through movement opposite of the left turn), it would be considered a non-serviceable conflicting call. As an example, if Phase 7 is omitted when Phase 8 is green, and a call becomes active for Phase 7, this call would be considered a non-serviceable conflicting call. Also, under coordination, if the permissive period has past for a particular phase, and a call is present, that would also be considered a non-serviceable conflicting call.
Question 6: simultaneous gap out
Answer:
The Simultaneous Gap Out feature, usually implemented on a Phase by Phase basis, is used to allow a phase that has “gapped out” to continue to time if additional demand arrives on that phase, as long as one of its associated phases is still timing. For example, let’s say that Phases 4 and 8 are timing. If Phase 4 gaps out, but Phase 8 still has demand, Phase 4 will remain green. Then, if additional demand arrives on Phase 4, while Phase 8 is still running its green, Phase 4’s Passage (or Gap) timer will become active again. The same idea applies to Phase 4 if Phase 8 gaps out, and Phase 7 becomes active; Phase 4’s Passage timer will be able to be “reactivated” if additional demand arrives on Phase 4. Basically it means that ALL Phases running on one side of the barrier must either gap out or reach the maximum green value before the controller will service vehicles on the other side of the barrier.
As for saying that it “always provide(s) dilemma zone protection”, there are other variables that would override this, such as Max Out or Force Off, which will disregard the Passage timer and terminate the phase. Always is a very dangerous word to use. However, I would say that in many cases, Simultaneous Gap Out does help provide dilemma zone protection, and should be enabled on the controller, especially on the Main St. I typically use it for ALL through movements at an intersection, either isolated or in actuated coordinated operation. In the actuated coordinated operation, it can allow the side streets to utilize more of their green time, even if there are larger gaps between vehicles.
Question 7: controller types
Answer:
Controller types:
NEMA:
TS-1 – older controllers with the 4 circular MS connectors (like the LMD series controllers in the lab), and possibly an RS-232 port and another comm. port that utilizes the manufacturer’s proprietary communications scheme. NO SDLC port.
TS-2, Type 1 – All communications within the cabinet is done via the SDLC port / bus in the cabinet. Only has 1 circular connector for power.
TS-2, Type 2 – has 4 circular MS connectors, similar to TS-1, but also has the SDLC port to connect with the MMU (malfunction monitoring unit) (like the Econolite ASC-3 controller? – the one used on Day 5 of the TSSW). Allows the flexibility / features of the TS-2 controllers, but maintains the familiarity of the older TS-1 equipment.
Each manufacturer, with their proprietary communications scheme, utilizes there own different PC software applications to communicate with their controllers. For example, Peek uses the LMSystem software or CLMATS, depending on the generation of controller; Traconex uses the TNET software; Econolite uses the Airies software; Naztec uses the Streetwise software.
To gain some compatibility between the manufacturers, NTCIP was developed / is being developed. This is supposed to allow different manufacturers equipment to be used over the same communications link. However, so far, this has had marginal success in the controller markets (mostly because each manufacturer still wants to be unique, so that users will still need to use their products).
Other:
170
2070
ATC
These controllers utilize firmware to provide flexibility and features. Manufacturers provide basic hardware, at a fairly inexpensive cost. Then, to make them work, users must purchase firmware that will provide the functions and features desired. The firmware is usually purchased separately for each controller, so the cost can get expensive in a hurry, depending on the complexity of the firmware. I have not worked with any of these controllers, so I am not terribly familiar with exactly how all of this works.