T3 Webinar Question and Answer Transcript

Virtual Intersections: Using Simulated Traffic Signals in Mobile Signal Timing (MOST) Training (April 15, 2009)

Back to Webinar Files


Q. What was the dollar amount of funding for this project?
A. Michael Kyte: For the phase that we just completed, it was about 600 thousand dollars. As Tom mentioned, this was a fairly long and intensive effort to get the details right, and so that total cost is about 600 thousand.
Q.  Does the software have capability to routinely save data for a certain period of time? I am asking this as such a function may help in troubleshooting some failures by replaying controller operation when the problem occurred.
A. Michael Kyte:  Let me address the overall simulation package and then maybe ask Kiel to pop in if I've missed anything. The MOST version, the software version that's available is based on VISSIM 4.3. It's not the newest version, but it has many of the features and certainly very, very good in terms of animation. It is limited to two intersections. That's the largest number we have on any of our input files. You can save files. You can make changes in the parameters, in the traffic signal controller itself, as several people had mentioned in their presentations. So, it is not a fully functional version in terms of not being able to handle more than two intersections, but for our purposes, for the MOST course, it handles everything that we need to show in terms of students learning about the topics that we talked about.
Q.  What all do we need to have in house to work with our summer interns? i.e: Do we also need to purchase VISSIM and an Econolite controller(s)?
A. Michael Kyte:  Everything is available on the website. The input files, the movie files, VISSIM. Again, it's that limited version that, through agreement with PTV and Econolite, it is available for student use for free for doing the MOST labs. So, all that is available on the MOST website.
Q.  How many intersections can be displayed/programmed?
A. Michael Kyte: We're limited to up to two for the version of the software for the MOST course. I should mention, and again, Kiel, feel free to jump in on this; you can purchase the full version of VISSIM with the Econolite controller. I'm sure Kiel would be glad to talk to you about that. Do you want to jump in on that, Kiel, at all?

Kiel Ova: Sure. Yeah. I think the key here is that all of the example sets have already been created for you. So we're already providing a lot of the example files that you can have, like Mike is saying, up to two intersections, and be able to use the MOST environment. Of course, VISSIM and Econolite, we support an unlimited number of connections, so if you wanted to model your entire city, of course, we would be able to do that, but that requires the full version of the software to be purchased from our company.
Q. Is the traveling MOST seminar a week long?
A. Michael Kyte: The full version would be about a five-day course. We are having discussions with Federal Highway Administration now as to exactly how this might be packaged. As I mentioned, we have done several pilots of the various experiments and labs, but exactly how it's going to be delivered through FHWA is still in the process of being determined.
Q. Are additional in-person (summer) courses planned. If so, where, what cost, how to apply to attend, to fund students?
A. Michael Kyte: At this point, we are not offering our traffic signal summer workshop because of the development of the MOST labs. And at some point, we do hope to get back to offering it again, but it won't has been this summer, and we're really trying to figure out what our new version of that summer workshop would look like using these MOST materials.

Paul Olson: Let me jump in here. Ultimately, we'd like to bring it to you rather than you have to go to it, and that's part of our discussions within Federal Highways is to how to make that happen.
Q. Are there plans to extend this simulation and training to different controllers?
A. Michael Kyte: I would say that that depends on the availability of a software emulator that other companies might have. I might ask Kiel or Gary to jump in and talk about that.

Kiel Ova: Certainly, this plan right now, I think, we're just thinking with one controller, but if there's a request to have specific controllers with this same type of environment that's kind of outside of the scope of what we were trying to accomplish here, which was to really focus on timing parameters and procedures that are applicable to any traffic controller.

Paul Olson: This is Paul again. I think Kiel makes an important point here, and that as potential students for this, you need to see past particular color or brand name on the unit itself. And we're trying to focus on the exact operating procedures of an intersection that would transfer to anything you'd want to put out there.
Q. Can you deliver only certain modules, say skip the very basic training?
A. Michael Kyte: I think the way that the various experiments can be packaged is still up for grabs. If someone is interested in doing only coordination, certainly, and if they have basic understanding of the controller processes, they could simply jump to labs six and seven. So, yeah, there's various ways to package this, and again, this is something that we'll be talking to Federal Highway Administration about what the various options are depending upon students' interest and needs.
Q. Have the license issues been resolved? Do we need to purchase a licensed copy of VISSIM and the Econolite ASC3/SIL software?
A. Michael Kyte: The software that's on the website is available for free. There are no license issues outstanding right-- so someone can go to the website right now, download the VISSIM ASC/3 MOST environment.
Q.  Does MOST have fail-safe measures to control the controller in real time?
A. Kiel Ova: Perhaps the question might be in terms of fail-safe that the VISSIM Simulation environment is in step with the traffic controller, so there is a direct relationship of the simulation to the controller. The time step of the controller can only proceed with the time step of the simulation. There's no way that the simulation and the controller can be out of sync in terms of time.
Q.  Paul Olson: There's several general questions and a lot of questions we've had relating to other tools, controllers or simulation models.
A.  Paul Olson:  I think the quick answer is that what we have is what we have at this point. And the goal was, again, to talk about general controller operations, features and functions, and to try to refrain from getting down into the weeds with particular vendor products, with particular special features and those sorts of things. Again, this is relatively high level to start with.
Q. Student use is free; what if we are not students?
A. Michael Kyte: It's essentially free to anybody who is interested in doing it. So, the term student is really a generic term. And I guess I might just sort of say that, you know, in the context of lifelong learning, we're all students, and so it's available to anyone, regardless of age, who's interested in learning more about signal timing. However, be careful at home with your kids using it!
Q.  You mentioned the new traffic signal manual; where can we get a copy?
A. Paul Olson:  I think if you just do a Google search on a traffic signal timing manual, you will easily find it. I recommend you do it, download it and see what you can do with it.

Michael Kyte: Paul, if I can just interrupt for a second, if they go to http://www.signaltiming.com - it has the traffic signal timing manual.

[Comment from a Participant] - The link for TSTM is http://www.ops.fhwa.dot.gov/arterial_mgmt/pubs.htm. Signaltiming.com is not maintained by FHWA. The manual is also now available from ITE.
Q.  The next question is: Is the software available as a training tool for entry level state engineers?
A. Paul Olson:  A quick answer to that is: Absolutely. Another question: Is FHWA offering grants to government officials for MOST training? I guess that's my question since I'm with Federal Highways, and we're still trying to resolve that and see where we can go. We will tell people shortly how that's going to work out and make sure that we get as many people through this training as possible.
Q.  Is FHWA offering grants to government officials for MOST training?
A. Paul Olson: We're [FHWA] still trying to resolve that and see where we can go. We will tell people shortly how that's going to work out and make sure that we get as many people through this training as possible.
Q. If we already have VISSIM License, how much would it cost to buy the Econolite software controller?
A. Paul Olson:  I think that's best talked to the Econolite people directly and see what they can do for you.
Q. Is the training extensible to interval operations? I think he's asking about an interval-based controller rather than a phase-based controller, and Econolite ASC/3 is a phase- and ring-based machine, not an interval-based machine.
A. Michael Kyte: I think the basic concepts are for phase-based operations.

Paul Olson: I think maybe the long answer to this is that interval operations might be one of our special modules to be developed in the future.
Q. What do you expect the cost to be per student (or organization) when you are ready to start the training?
A. Paul Olson:  We're not sure on that yet. We're going to try to keep the cost as low as possible. We still have some things to work through.
Q. We would be interested in a rail road preemption simulation using MOST. Is that possible with what we have right now?
A. Michael Kyte: I think you can do-- see, Kiel, do you want to jump in on this? I think you can do preemption in this. Is that right or not?
Q. We would also be interested in railroad pre-emption simulation using MOST, Saeed SDOT.
A. Paul Olson  I guess that's something we need to think about in the future.

Kiel Ova: Yes, and Gary can confirm that, too, that the preemption module is activated in this version. Is that right, Gary?
Gary Duncan: That's correct. The ASC/3 SIL in this version can do preemption, but I'm not sure if the MOST input files would allow it under the command layer.

Kiel Ova: Yeah. We would want to have, you know, that would require development of new training files where we would have the appropriate detection locations for the preempt inputs and so on. That would be a continuation of this software.
Q.  Is an actual ASC3 controller required?
A. Paul Olson:  And the quick answer is no, you've got the software so you don't need the hardware. If you think you need to have the hardware, then you're limited to real-time operation and real-time work with this, and that gets pretty slow and boring.
Q. Has a centralized training system been considered? For instance, Lambda rail connects several Universities in the Southeast. Could this set-up be used to base the software/equipment at one University so that we can access?
A. Michael Kyte: That's certainly possible, but again, it's all available in software version that each of the sites would be able to download themselves.

Paul Olson: So there's no real advantage to having one university run it all.

Michael Kyte: Right.

Paul Olson: The one part that we still have to maybe work on is the instructor part of it. And there may be some advantage to a central instructor pool.
Q. Can bikes and pedestrians be simulated with MOST?
A. Michael Kyte: Yes. In one of the labs, one of the experiments, we do have an experiment dealing with determination of pedestrian times. In terms of bikes and pedestrians, depending upon what is in the VISSIM simulation times, which we create, can be included. Let's see, Kiel, are bikes possible with the standard version of VISSIM?

Kiel Ova: Yes. Any type of mode can be simulated. So pedestrians, bicycles, motorcycles.

Michael Kyte: So the answer, then, is that there are labs that deal with pedestrians. There's nothing specifically with bikes yet, but that could be something that could be a topic for future lab development.
Q.  In your experiments did you consider human factors especially older driver behaviors?
A. Kiel Ova::  Yeah. I suppose at some level, you know, VISSIM has a reflection of certain human factors in our car following and lane changing models. There are various things in the details of the simulation to do that. I think in terms of maybe the question being specific to how you would time signals for older drivers, Mike, you do you want to address that?

Michael Kyte: Well, I might mention that in several of the labs and experiments, there are issues of different vehicle types, which, for example, larger vehicles such as trucks or buses that have different acceleration rates and responses to the change in signal status. So we captured that in a couple of experiments in which the vehicle types are different. But certainly a future lab can deal with different, you know, older drivers with slower response times and the effects that that may have, for example, on setting minimum green or yellow times.
Q. Can the course be applied to a specific situation in a specific city?
A. Michael Kyte: Well, people are free to develop additional labs or experiments. Right now, we have a generic city, in which conditions are what I would call fairly normal, from light traffic to heavy traffic. So, you know, if you're interested in tailoring something to your specific city, that's certainly possible, but I think the conditions that we cover are fairly general in the sense of applying to most experiences, you know, that people would have around the country.
Q. How long would it take to go through the course, self-guided, approximate hours or days?
A. Paul Olson:  That might go to Tom Urbanik, because he's had some of his students do the self guided.

Thomas Urbanik: Yeah. That's probably, you know, hard to say, because the students don't actually report their time. It's going to be of the same order of magnitude. I can answer the question to the extent that it's probably very easy to go through it self guided. It is not hard at all to do. The only down side of the self-guided part is if you run into a problem, then you're going to need somebody to help you out, but you could definitely do it self-guided.
Q. Can you simulate transit priority?
A. Michael Kyte: Again, it's possible. We don't deal with transit priority in the current labs or experiments that have been developed to date, but certainly in terms of future labs, it can be included.
Q. Is the detection real time or is it simulated?
A. Michael Kyte: Well, it's all a simulation, but it occurs in the manner in which it would actually occur in the field. There's a detector call that happens because a vehicle passes through a detection zone, and the call is received. In the simulated sense, it's real-time, yes.
Q. Text
A. Michael Kyte: Text
Q. Are speed changes due to turning vehicles taken into consideration?
A. Paul Olson:  I think that would probably go to Kiel.

Kiel Ova: Certainly. That's more of a modeling question, how you deal with the VISSIM simulation. Speed changes for turning vehicles are an input by VISSIM modeling, so you can control that. The sharper the radius, the greater the impact and speed reduction.
Q. Has the simulation modeled twice per cycle movements?
A. Paul Olson:  I think what we're getting into is a lot of questions about the base level environment, and I think we have to remember that what we've done is set up a training environment. Now, it's certainly possible that you could purchase the VISSIM software, you could purchase the Econolite controller module or have other controller modules built for you, and you could cobble up whatever scenario you wanted to and do this testing. You could put in your own intersection. You could put in your own controller timings, and you could model everything that you could possibly dream of that this equipment is capable of doing, but we have to remember that what we have here is a training environment, and we're trying to reach out to bring, as Tom Urbanik says, we'd like to bring the floor up a little bit so that we get more people to understand just the basics, and then go from there. But, surely, if you want to do things like double cycling, you want to do the rail road preempt, you want to do the signal priority, you could purchase the software and have that done to your own desire for your own intersection, your own situations. Then we'd suggest that that's probably something that should be done by other folks.
Q. How does MOST respond to the variations in the controller software? Saeed SDOT?
A. Gary Duncan:  I can answer that, Paul. This is Gary. The MOST package that's being distributed has a fixed version of the ASC/3 SIL software. And that's the only version available with the MOST package. However, with the standard purchase of the ASC/3 SIL as part of the VISSIM product, a user can purchase a maintenance module and get continuing updates of the ASC/3 SIL software.
Q. I know this is for training now...but in the future...I see a lot of possibilities! We have worked with a few cities that want to compare different vendor's controllers. We've done some experiments with Trafficware's CID. Looks like this could replace the Hardware interface. Only issue is you only have the Econolite. In the future...is there a future...any other vendors involved?
A. Kiel Ova:  I guess that's a question probably that seems to be outside of this training, I think. Right, Paul? Paul Olson: Yes, absolutely. And again, the sky's the limit, and if that's what you want to do. In fact, we welcome people trying to push this forward and move it into new fields.
Q. Can detection schemes be tailored to suit local standards? The examples shown are only stop bar detection.
A. Michael Kyte: Well, again, it's a matter of developing additional modules that look at other kinds of detection schemes. We did focus primarily on stop bar detection, except in lab four, where we deal with advanced detection, but again, if there are other desires to look at different detection schemes, they can be developed in future modules.

Paul Olson: Well, I guess key here is that we're just at the ground floor with this whole Software-In-The-Loop and the training, and we've got a long ways to go and a lot of things to do and probably a lot of smart people that have a lot of ideas that we've never thought about so far. So, we'd like to hear them. I'm not sure exactly right now, other than the evaluation form and the question form that we'll send you later, how we can collect those, but you have our e-mail addresses and our phone numbers, and you can send in cards and letters saying do this or do that, and we'll try to work with you and see where we can go with it.

Thomas Urbanik: I just want to follow up on the last question a little bit in that some things are easier to do and some things are harder to do relative to changes. And the reality is if you want to tailor the training to local standards, some things would be relatively easy to do but would nevertheless require someone with knowledge of VISSIM to make the changes. So, if you used a 30-foot instead of a 20-foot loop and for stop bar detection or your setback distance for the advanced detection were different, those things are probably relatively easy to do. Other things like LRT, railroad preemption and the like are all within the realm of doable; they're just a lot more complicated. So I think the answer to how much change and how difficult it is depends on the particular application. I think that's important to understand.
Questions submitted that were not addressed in the session:

Controllers
Q. Have there been thoughts or steps taken toward integrating other types/models of controllers into the MOST interface?
A. Paul Olson: This was addressed in other answers.
Q. If our agency does not use ASC/3 Econolite controllers, could we substitute other UTDF controller data for use in the MOST VISSIM model?
A. Paul Olson: No.
Q. Are other brands of controllers available?
A. Paul Olson: No.
Q. I did not understand some of the previous answers regarding using this. If we want to use this in our office for testing...can we use the software without the Econolite module?
A. Paul Olson: No.
Q.  Is it possible to use the VAP macro language to program the controller?
A. Paul Olson: We are not sure, but remember this is at this time primarily a training tool.
Interfacing with other Applications
Q. Can the system interface with other signal systems or just Econolite?
A. Paul Olson:  Not yet.
Q.  Is there any plan to extend the simulation to CORSIM?
A. Paul Olson: Not at this time.
Training
Q. Which establishment offers 'good' training on VISSIM and signal timing?
A. Paul Olson:  For this training effort you do not need VISSIM training.
Q. This is good for both students and practicing engineers! However, a 5-day course is too long for non-students.
A. Paul Olson: We recognize this is an issue and we are attempting to resolve it. You may have to break it up over several weeks a day or two a week.
Q. Is there a resource to determine when and where the training will be available? If it is not yet established, how can we get updated to when it will be available?
A. Paul Olson: Not yet.
Other Questions about MOST
Q.  In your experiments did you consider human factors especially older driver behaviors?
A. Paul Olson: That is an integral part of the VISSIM Model.
Q. Does MOST control the intersection in real time?
A. Paul Olson: It can but it can also run it faster or slower than real time.
Q. Is there any plans to use MOST for more advanced labs on subjects like Traffic Adaptive?
A. Paul Olson: That could be considered for a future module.
Q. Are the traffic volumes adjustable by phase?
A. Paul Olson: Yes.

back to top