Newsletter of the
ITS Cooperative Deployment Network
(Please read the Disclaimer)

ICDN: What is the primary objective of the IDAS tool?

Alexiadis: The primary objective is to help transportation planning agencies estimate benefits and costs of different ITS deployments. IDAS can do a cost/benefit analysis for over 60 different types of ITS technologies in isolation or in combination. The primary reason for developing IDAS was that, traditionally, DOTs and MPOs had special ITS divisions to handle deployments. Those divisions used their own analytical procedures, which were different from the non-ITS analyses. IDAS was developed to help "mainstream" ITS. It uses the same background assumptions that are used in the analysis of non-ITS deployments, such as carpool lanes and transit improvements.

ICDN: What do you mean by "background assumptions"?

Alexiadis: Traditionally, the long-range transportation planning and forecasting function uses so-called "travel demand" models. Metropolitan Planning Organizations (MPOs), State Departments of Transportation, counties and others use those models, which have been around for the past 30 or 40 years. It’s a 4-step process: they estimate how trips are generated, distributed on the network, assigned to the network, how mode shifts occur, and so on. IDAS relies on outputs from these travel models to estimate the effectiveness of ITS deployments. After inputting this information from the travel model, you can then deploy one or more of over 60 types of ITS wherever appropriate.

ICDN: I presume you wouldn’t want to try to create the network manually.

Alexiadis: No, IDAS doesn’t allow you to create a network manually. We are not in the business of competing against those traditional planning tools. We are going to use what is available and take it to the next step.

ICDN: Is the user interface on IDAS similar to the user interface on those tools?

Alexiadis: It is actually much easier than in those tools. Our steering committee consisting of a mix of MPOs and actual technical people guided us very well in terms of user friendliness.

ICDN: Is it a drag and drop kind of thing?

Alexiadis: Yes, it’s drag and drop. You just drop a traffic-actuated meter onto the network, and then IDAS asks you a set of questions: When will construction start or when is the mid-point? When are you going to open it? Select the ramps where you are going to deploy it. Select the freeway links that are affected by it. (See the IDAS Screen Shot for this operation.)

ICDN: Is this the same interface that was in place with IDAS Build 1 (check out the ICDN's earlier interview), or has it changed for Build 2?

Alexiadis: It’s more robust, but the general drag-and-drop idea is the same. In the previous build we had only about 20 types of ITS available, and now we have 60.

ICDN: What are some of the major types of ITS that you’ve added?

Alexiadis: We’ve added traveler information systems, commercial vehicle operations, advanced vehicle controls systems, vision enhancement.

ICDN: How do you show the benefits of "vision enhancement," either in dollars or in throughput improvement? How can you actually show any benefits of that technology?

Alexiadis: Vision enhancement primarily has a safety benefit; it reduces the accident rate. This type of microscopic accident analysis is done primarily by looking at traffic volume over the year, where a certain facility type has a certain accident rate. Some of these technologies have a direct impact on reducing the accident rates.

ICDN: Is there competent data showing that benefit?

Alexiadis: Yes. We have reviewed hundreds of before/after evaluation studies, and we base all of our results on those studies. We did not include technologies that did not have before/after studies associated with them, like automated highway systems.

ICDN: IDAS contains references to those studies, right?

Alexiadis: Yes. For example, for vision enhancement the default benefit in IDAS is a 3% accident reduction as a result of deploying the system. This data is from the California Department of Transportation Advanced Transportation Systems Program Plan Update. You can go to the source and look at it.

ICDN: In cases where studies are contradictory, do you pick the study or studies most supportive of ITS deployment?

Alexiadis: No, we usually do an average. We also put in a warning that says that the range of observed impacts is from X to Y, and encourage the user to go to the library and look at those studies, because this is a national average and people drive differently in Kansas City than in Boston.

ICDN: Again, if I am a user can I link to the actual reference?

Alexiadis: No, you can’t link to it.

ICDN: Will IDAS tell me how I can acquire it?

Alexiadis: If you go to any library system and provide this title you will get it. Eventually there will be a web site to support users, and that site may link to other databases and provide system updates.

ICDN: How does IDAS potentially change the way transportation agencies plan ITS improvements?

Alexiadis: IDAS is used is to compare and screen among alternatives. Traditionally, before IDAS, the selection of alternatives was often not based on a quantification of costs and benefits.

ICDN: Give me an example of 2 common alternatives people might want to compare.

Alexiadis: I'll give you a real example -- a test we did in Tucson. The Pima Association of Governments (Tucson MPO) wanted to look out to the year 2020. They wanted to analyze the differences in spending their money on deploying ramp metering -- which is mostly a highway-related investment -- versus putting AVL and an automated scheduling system on all of their buses -- which is a transit investment-- or some combination. So we looked at the expected benefits of each alternative, as well as a combination of the two.

ICDN: What type of benefits were you looking for?

Alexiadis: Travel time, throughput, safety, emissions, fuel consumption, travel-time reliability, and agency cost savings. IDAS calculates various performance measures for each potential project.

ICDN: If the user wants to find out the impact of installing new loop detectors and/or video cameras, does the system tell him where to put them on the network?

Alexiadis: No, the user places them, but then the system keeps a tally, an inventory of equipment located near links or nodes. The system also keeps track of labor required for incident detection at the management center, the hardware and software required for incident detection, and integration costs. If you have a communication line that could be used by ramp meters and by incident detection, IDAS asks you "do you want to share or not?" It can also allocate costs across different agencies.

ICDN: These costs must come from an underlying database, right?

Alexiadis: Yes. The database is from the National ITS Architecture. These costs were subsequently reviewed and updated by Mitretek.

ICDN: What types of cost information does this default database provide?

Alexiadis: It gives you a low, medium, and high range of costs by equipment. Of course, the first thing that a state DOT would do when they got IDAS would be to check these cost estimates and calibrate them to the specific local conditions.

ICDN: Is this tool primarily targeted for MPOs?

Alexiadis: Yes, MPOs, state departments of transportation, county agencies, city agencies, and consultants.

ICDN: After you've analyzed several alternatives, does IDAS let you print out a report that you can give to your boss and say, "Here’s the numbers"?

Alexiadis: Yes. You can have either a short summary or a much more substantiated, extensive performance summary, containing information such as: "Traffic (VMT) on the freeway was reduced by X%." It provides performance measures by mode and by facility type, and it also provides maps. That way you can see on the network traffic volume and speed differences from before to after the ITS deployment. So you can see impacts of adding ITS on the network. (See the IDAS Screen Shot for this operation.)

ICDN: In addition to adding 40 new ITS categories, what are the other changes in IDAS Build 2 versus Build 1?

Alexiadis: We also implemented some additional capabilities like a module to estimate travel time reliability impacts in addition to travel time impacts. This is a fine point, but it makes a big difference.

ICDN: Tell me why that's important.

Alexiadis: It addresses the "variability" of traffic. ITS has an impact on variability of traffic in addition to travel-time itself. An incident management system, for example, makes travel more reliable. That's an important factor for certain segments of the transportation market -- like truckers -- that are involved in just-in-time delivery. In today's business environment trucks are actually the warehouses, providing inventory. The delivery time is planned right into the manufacturing process. The trucking company's customers need to know the range of time for delivery.

ICDN: So, what do you specify, the standard deviation?

Alexiadis: Yes, we do use the standard deviation of travel time, which is traditionally ignored in transportation planning, which looks at a "typical day." When transportation planners get data to calibrate their models for this typical day they try to avoid incidents. However, there's usually at least one incident every day in most major metropolitan areas. So we looked at the HPMS [Highway Performance Monitoring System] database, a database of freeway traffic data collected by the Federal Highway Administration. We looked at the number and duration of incidents relative to the V/C (volume/capacity) ratio for two, three, and four-lane freeways. So when IDAS redoes its traffic assignment as a result of a new ITS deployment, it then goes to this travel time reliability module and tabulates changes in and impact of incident delay. We use 3 times the dollar cost for incident delay when contrasted to "normal" delay due to recurring congestion.

ICDN: Why did you do that?

Alexiadis: Because unexpected delay, which is incident delay, is more costly. You miss a meeting. If you expect the delay you plan for it and leave earlier. Incident delay is more costly to society than day-to-day recurrent delay.

ICDN: Did anyone use IDAS Version 1 for real world ITS deployment planning?

Alexiadis: I don’t know if it was actually used for any real deployments.

ICDN: In your opinion, is Version 2 now a usable tool?

Alexiadis: Yes. It’s already been applied in eight metropolitan areas with reasonable results. Of course, like any tool it can be misused. This is the warning we put in the users manual: check the assumptions before you use the tool. Calibrate IDAS to your local situation.

ICDN: Let me ask you a related question. What if an agency says "We want to be really conservative about ITS deployments, so we want to apply a standard 50% reduction to IDAS' benefit estimates because we think they might be inflated. Can they do that across the board?

Alexiadis: Yes they can do this across the board.

ICDN: When will IDAS Build 2 be formally released?

Alexiadis: We expect that McTrans and PCTrans will start selling it in March.

ICDN: You distributed Build 1 for free, as I recall. Why are you now "selling" Build 2 of IDAS?

Alexiadis: We asked $40 for shipping costs for Build 1, which was just a prototype, had bugs, and was not fully operational. The primary purpose of distributing IDAS Build 1 was to get early feedback to improve the usability of Build 2, the fully supported version. However, IDAS Build 2 has several additional costs associated with it. For example, it uses a proprietary SyBase database engine that costs about $400.

ICDN: Wasn’t that database engine in Version 1?

Alexiadis: It was in there, but because Build 1 was a test version SyBase let us send it out for free. For Build 2, McTrans and PCTrans also charge distribution and installation support fees. I expect that the price for Version 2 is going to be between $500 and $1000, which is not much for transportation software. The TSIS software is going for $600 right now. Any travel demand model system that you buy sells for $10,000 to $20,000.

ICDN: What about training? This is not the type of tool that someone would probably use right out of the box.

Alexiadis: It’s not. We are currently working with the FHWA to put together a set of training workshops. Actually there is an RFP out on the street from the National Highway Institute (NHI), and we are working with some other team members to respond to that. The intent is to provide four 2-day training sessions at each one of the 4 Federal Highway Resource Centers in Baltimore, Atlanta, Chicago, and San Francisco.

ICDN: Where do you go from here on this? Is there a plan for IDAS Version 3?

Alexiadis: The Federal Highway Administration is trying to secure some funds to do updates and to continue development. However, the non-ITS community has also looked at IDAS and likes its benefit/cost aspect and performance-based planning capability. So IDAS is probably going to be given a different name and be used to do performance-based planning for non-ITS transportation deployments -- carpool lanes, transit lines, mixed-flow lanes, and other transportation investments. That way, there would be a direct basis for which to compare ITS with non-ITS improvements.

ICDN: Can you say what changes you'd like to make in a next version of IDAS?

Alexiadis: Well, it’s primarily user-friendliness things, such as adding more textual descriptions of ITS elements. Analytically, the IDAS Version 2 is fine.

ICDN: How about benefits of ATIS technologies. For instance, does IDAS calculate the benefits of a route guidance system?

Alexiadis: Yes it does. IDAS provides for the analysis of in-vehicle information systems or hand-held personal devices with route guidance and without.

ICDN: IDAS includes an ITS service called "centralized route guidance." What does that phrase mean?

Alexiadis: That the control center provides information in real-time to the vehicle.

ICDN: Is there concrete data about the potential benefit of that technology in the U.S.?

Alexiadis: We use a lot of Japanese studies and studies by universities. We also use Mitretek's work that compiled a lot of studies. From all that we came up with a methodology for centralized route guidance.

ICDN: Show me how IDAS computes the benefits of that technology.

Alexiadis: IDAS takes into account several factors. We require several types of input from the user: How many links provide actual traveler information -- 10%, 20%, 80%? That parameter affects the range of impact. How many devices are there? How many vehicles are available? What is the market penetration? We have some defaults for market penetration: ½% for the year 2000, 5% for the year 2005, 10% for 2010, and so on. You can change those numbers if you think they are too aggressive. We also need to know the percentage of travelers that have their systems activated during the trip -- not everybody will have their system activated. Of those that are activated, what is the maximum amount of time that one would save if the system had information on all of the route segments that they use? IDAS calculates that one would have 25% savings in travel time with up to 10% market penetration. The fewer the number of people that have devices, the more the benefit.

ICDN: That may seem counterintuitive to some people. Why would the benefit of route guidance be greatest with a small number of devices?

Alexiadis: The more the number of ATIS devices, the more likelihood that the drivers that have devices will all be guided to the same route, in turn causing congestion on that route. This phenomenon is sometimes called the "hunting effect." With a 40% market penetration, IDAS calculates that the delay reduction (when compared to the 10% penetration figure) would be cut in half. At 60% market penetration, the benefit goes to zero. OK

ICDN: That would only be true if individual vehicles have the same algorithms, which tell all the drivers to divert on the same routing, isn't that so?

Alexiadis: Actually, the benefit numbers would be true even if the system told them to divert without telling them where to go -- say if it alerts them that there is an accident downstream. Again, these are average numbers.

ICDN: Are there any ATIS technologies that have been promoted in recent years that you found failed to achieve a positive cost/benefit ratio?

Alexiadis: Other than kiosks, no.

ICDN: Kiosks are the only one?

Alexiadis: Yes. Again, it is a cost/benefit issue. A kiosk may be useful to provide some help to tourists that don't know their way around, but that is not a quantifiable answer for IDAS.

ICDN: Can IDAS assess the relative benefit of ITS deployment compared against building new lanes? Can it compare the impact of adding ramp meters and lane control signals and other freeway-based ITS technologies to simply adding another lane? In some ways that's the classical argument for ITS. Can IDAS calculate that comparison?

Alexiadis: We're thinking about explicitly supporting this type of analysis in Build 3. However, there is a way to "trick" Version 2 into providing this type of comparative data. You need the MPO to develop an alternative that has the new lane and an alternative that doesn’t -- a "no build" and a "build" alternative. They have to do this for their long-range plan anyway. Then you pass these two alternatives through IDAS without deploying any ITS, and you export the results to an Excel spreadsheet that lets you do a comparison. It’s a roundabout technique and doesn’t have any cost information in it, so you have to provide cost information.

ICDN: Does it have benefit information?

Alexiadis: It will give you benefits in terms of travel time savings, safety, emissions, travel time reliability -- all of the normal IDAS reports.

ICDN: IDAS can compute all those factors on bare networks without ITS?

Alexiadis: It will reroute traffic and balance the traffic across the whole network due to the addition of a new lane. You can then assess the relative effectiveness of ITS solutions by comparing the "build" alternative (that is, an extra lane but no ITS) to an ITS-enhanced version of the "no build" alternative.

ICDN: It seems to me that IDAS' success rests entirely on its credibility. However, it is being funded by the U.S. DOT, which encourages the use of ITS. Do you think credibility is going to be an issue? Some people might look at this and say "well, everything in the database is skewed toward showing the benefits of ITS, because the people that put it out are trying to promote ITS."

Alexiadis: In some instances the case studies we use showed negative benefits. We were very careful about basing the benefits on substantiated studies, not on benefit assumptions or people's wishes.

ICDN: I’ve heard people argue that there aren’t many studies that show concrete benefits of ITS.

Alexiadis: There are a few, and our literature review encompassed hundreds of studies, including operational tests and international studies. IDAS incorporates results from all over the world. For example, we have Japanese studies for ATIS -- they have deployed more ATIS technologies than we have in the U.S. We also have European studies.

ICDN: Most of these studies are written by people that have a vested interest in demonstrating or showing the success of their projects, so you might argue that their objectivity might be somewhat in question.

Alexiadis: Perhaps, but with IDAS you can always look at which studies we’ve used to come up with 38% accident reduction, for example, and change that number if you feel it's overly optimistic.

ICDN: But what if you don’t know what the real number is, because you weren't involved in the study?

Alexiadis: That's why we have another little tool called the "weighting factor." Say, for example, somebody comes to me and says that the methodology that we used to estimate the impact of ramp metering on carbon monoxide emissions is wrong by a factor of 5. I can put .2 in as a "weight factor." This feature is for a public hearing type of analysis. The weight factor will change the benefit/cost ratio. There are at least three or four ways that you can do sensitivity and risk analysis of your assumptions and results within IDAS. It’s up to the user. This is not a black box; it’s actually a very open box.

From what you know about it, will the IDAS tool help you or others in your organization plan for future ITS deployments? What factors might weigh into your organization's decision to use (or not use) this tool? Click on the "note" icon (or click here) to share your thoughts with others on the front lines of ITS deployment.