Good afternoon or good morning to those of you in the West. Welcome to the Talking Freight Seminar Series. My name is Laura Feast and I will moderate today's seminar. Today's topic is the National Freight Performance Measurement Data Dissemination Tool. Please be advised that today's seminar is being recorded.
Today we'll have two presentations, the first given by Crystal Jones of the Federal Highway Administration Office of Freight Management and Operations and Jeffrey Short of the American Transportation Research Institute. The final presentation will be given by Kathleen (Kitty) Hancock of Virginia Tech.
Crystal Jones joined FHWA's Office of Freight Management and Operations in October 2003. Prior to joining FHWA Crystal worked for the Department of the Army for 11 years where she held several positions in transportation and logistics including an assignment with the Office of the Deputy Chief of Staff for Logistics at the Pentagon. Crystal's primary area of expertise is with freight technology and operations. Crystal also has extensive experience in the areas of programming and budgeting and strategic and performance planning. Crystal holds a Bachelor's Degree in Industrial Technology with a concentration in Computer Science from Elizabeth City State University, and a Masters of Science in Administration from Central Michigan University. She is the program manager for FHWA's Freight Performance Measurement initiative, the topic of today's Talking Freight.
Jeff Short is a Senior Research Associate with the American Transportation Research Institute (ATRI). He is project manager for the Travel Time in Freight Significant Corridors project (also known as Freight Performance Measures), sponsored by the Federal Highway Administration Office of Freight Management and Operations, which has developed a national system for measuring the performance of truck movement along U.S. freight corridors and at international border crossings. During his time with ATRI, Jeff has also played key roles in numerous other projects, including multiple Transportation Research Board-sponsored commercial vehicle safety syntheses, analyses of transportation funding practices and outcomes, research of onboard safety technologies, and an investigation of best practices for freight operations during national and regional emergencies. His areas of expertise include transportation and freight policy analysis, program evaluation, safety, technology, economics and operations. Jeffrey holds an MS from the Georgia Institute of Technology and a BA from Emory University. He is active in both the Transportation Research Board and ITS America, and is a member of the TRB Performance Measurement Committee.
Kitty Hancock is the co-director for the Center for Geospatial Information Technology and an associate professor in the Via Department of Civil and Environmental Engineering at Virginia Polytechnic Institute and State University. She earned her Ph.D. and Masters degree in civil engineering from Vanderbilt University in 1994 and 1991. She received her bachelor's degree in civil engineering from Colorado State University in 1982.
Dr. Hancock has more than 25 years of experience as a transportation engineer and researcher in freight planning and operations, geographic information systems for transportation, highway safety, crash analysis, and roadside safety. In addition to FPMweb, Dr. Hancock has been active on several truck and freight related projects including a National Science Foundation exploratory grant on micro-simulation of freight movement integrated with logistics, information, and e-commerce. She is currently providing technical expertise to the Freight Analysis Framework effort for Federal Highway Administration and is a member of the TRB standing committee on Freight Data.
I'd now like to go over a few logistical details prior to starting the seminar. Today's seminar will last 90 minutes, with 60 minutes allocated for the speakers, and the final 30 minutes for audience Question and Answer. If during the presentations you think of a question, you can type it into the smaller text box underneath the chat area on the lower right side of your screen. Please make sure you are typing in the thin text box and not the large white area. Please also make sure you send your question to "Everyone" and indicate which presenter your question is for. Presenters will be unable to answer your questions during their presentations, but I will start off the question and answer session with the questions typed into the chat box. Once we get through all of the questions that have been typed in, the Operator will give you instructions on how to ask a question over the phone. If you think of a question after the seminar, you can send it to the presenters directly, or I encourage you to use the Freight Planning LISTSERV. The LISTSERV is an email list and is a great forum for the distribution of information and a place where you can post questions to find out what other subscribers have learned in the area of Freight Planning. If you have not already joined the LISTSERV, the web address at which you can register is provided on the slide on your screen.
Finally, I would like to remind you that this session is being recorded. A file containing the audio and the visual portion of this seminar will be posted to the Talking Freight Web site within the next week. We encourage you to direct others in your office that may have not been able to attend this seminar to access the recorded seminar.
The PowerPoint presentations used during the seminar are available for download from the file download box in the lower right corner of your screen. The presentations will also be available online within the next week. I will notify all attendees of the availability of the PowerPoint's, the recording, and a transcript of this seminar.
We're now going to go ahead and get started. Today's topic, for those of you who just joined us, is the National Freight Performance Measurement Data Dissemination Tool. Our first presentation will be given by Crystal Jones of the Federal Highway Administration Office of Freight Management and Operations and Jeff Short of the American Transportation Research Institute.
As a reminder, if you have questions during the presentation please type them into the chat box and they will be answered in the last 30 minutes of the seminar.
Good afternoon. As Laura said, I am Crystal Jones from Federal Highway's Office of Freight Management and Operations, and as you can see from the title, the main focus of today's seminar is the data dissemination tool that Virginia Tech developed on behalf of the Freight Performance Measurement program. As the program manager for the initiative I thought what I would is start by providing background information as to why Federal Highway has engaged in this project, and Jeff Short will follow with some description of the data itself, the methods that we use for processing and gathering the data, and then Kitty's presentation as Laura indicated will be the grand finale that will describe the data dissemination tool we plan to be deploying in the early summer.
As many of you may know, since our last reauthorization within the Department of Transportation, there has been many proposals and positions put forth on what the next authorization for the Department of Transportation looks like, and though the proposals have varying positions what it should look like, there have been key things that emerged in terms, Some of those things include defining the federal rules as pertains to freight and good movement, linking transportation policy and funding to environment and energy, promoting better management of existing assets, using multiple funding sources, and that last bullet which is really aligned with what we're talking about here today is incorporating performance and accountability into transportation programs.
Transportation performance measurement as the previous slide indicated has emerged as a key item that needs to be considered or will be considered most likely in our next authorization. Several groups have recommended that the next authorization consider a performance-based surface transportation program, the proposals include the previous Administration's Transportation and infrastructure reform act proposal, AASHTO proposal, the policy and revenue commission, the finance commission, and also the Government Accountability Office.
Most of the work we've been doing on the freight performance measurement initiative has been focused on highway infrastructure, so the information that we'll be presenting here today is very highway focused and the work that we have ongoing here in federal highway focuses on those key parts of the highway infrastructure that are key to freight movement. That includes the Interstate System, international border crossings and approaches, and intermodal connectors which are those parts of the national highway system that connect to other parts of the national highway system to port terminals, rail terminals, and air terminals. We won't talk in detail about the last bullet, but the FPM initiative has a component that looks at movement of freight on the interstate system and international border crossings.
A few years back we had research done within Federal Highway that took a look at if we were going to develop a freight performance measurement program what would be some of the key indicators of freight performance. Some of the indicators that is emerged in that research are point to point travel times selected freight significant highways, crossing times at international border crossings, and condition of intermodal between NHS and intermodal terminals, and farm highway currently has a program or an initiative aimed at those top three indicators, and again the presentation here today mainly focuses on the point to point travel time on freight significant corridors, and we'll talk briefly about the crossing times at international border crossing, but the FPM data dissemination tool described later is primarily focused on the first indicator, point to point travel time on selected freight significant highways.
In terms of what the FPM initiative is, essentially what we have done is established a contractual partnership with the American Transportation Research Institute who in turn has partnerships with data providers and motor carriers, and we're collecting GPS and satellite data from trucks, approximately 500,000, that are moving nationwide on our highway system. The main focus of the initiative to date has been on 25 major U.S. Interstate corridors which is what the FPM web tool will focus on and land border crossings, we have a significant amount of data for 15 U.S. Canada land border crossings.
Again, we started this program in 2003 and it was sort of a crawl, walk, run approach. We started out with selected portions of five interstate highways, I-5, I-65, I-70, I-10, and I-45, and then we gradually expanded to the entire length of all five of those corridors, and those are the initial corridors shown in red on the slide.
After we refined our approaches and methods we expanded our data analysis and collection to 25 interstate corridors which is are the ones shown in blue, so essentially from about 2006 to the present we have data for the 25 interstate corridors shown in red and blue on this slide, and I just want to highlight the FPM web tool primarily focuses on these twenty-five corridors, but through the FPM program we have access to data for practically any interstate highway in the United States as well as those parts of national highway system that have significant freight movement, so our only limitation in terms of what we have available in the FPM initiative is whether or not there is a significant amount of types of vehicles-- which Jeff will describe-- that have the equipment that we get the GPS data from moving on those highways. Again our primary focus for the initiative is 25 interstate corridors.
This is a typical visualization of what a month of processed data would be displayed like on a map, and just like a lot of our traveler information tool that state DOTs and MPOs have on their website, this visualization product shows the average operating speed for the 25 interstates that I mentioned are our primary focus, and it is color coded, anything in green is above 55 miles per hour. This is 24 hours a day, seven days a week, and anything in yellow is 50 to 55, and anything in red is below 50 miles per hour, and again what Kitty will describe is how this type of national analysis can be taken down to the state using the dissemination, but our primary focus in federal highway has been to do national type analysis.
I mentioned there is an international border crossing component to FPM. Similar to our work on the interstate corridors we started the border crossing work with five border crossings, and they were the Champlain crossing, the Ambassador Bridge Crossing, Peace Bridge, Pacific Highway and Alexandria Bay, and we have gradually extended our border crossing component and now we have data for 15 major land border crossing areas on the U.S. Canada border and this crossing handle more than 80% of the trade inbound to the United States from Canada.
In our border work we didn't want to be focus necessarily on what a customs agency is doing inside the border crossing, so the data we have for the border crossing areas include not just the immediate crossing area but also the major highways that approach the crossing area, so as an example, for the Ambassador Bridge Crossing we have data for Michigan route 3, interstate 75, 94, 96, and the other roads that connect to the border crossing on the Canada side of the border, so it is not just a look at the border crossing itself, it is also the approach roads that connect to the international and border crossing.
Mainly our national level analysis focused on developing these current measures, average operating speeds and travel time reliability, and for the border crossings we look at average total crossing time and crossing time reliability. What we heard early on when we started our research is the velocity and speed are important to shippers and carriers and freight movement, they also view reliability as being equally important, meaning that they don't necessarily want to know how slow or how fast it is, they want to know how reliable they can plan their trips and routes and deliveries based on travel time reliability, so our focus is not only average operating speed but also travel time reliability.
In addition to the basic measures or the measures on the previous slides, we also have done at least initial validation to show coupled with other data sources we can derive other measures of travel time and listed here are just some of the examples that ATRI has experimented and demonstrated the capability to develop these types of measures, and hopefully within the next six months or so the web tool that you will see will have some other measures other than the initial measure that will be present today. Again, our primary focus thus far has been average operating speeds and travel time reliability.
Though today's session is primarily focused on the FPM web tool dissemination, I wanted to take the opportunity it give a brief mention to another tool federal highway is currently finishing up development on. I mentioned, intermodal connectors are another part of the national highway system that we consider to be very key to the movement of freight and the efficiency of freight movement, so we have recently concluded development, with the help of four states, Kentucky, Kansas, Georgia, and Indiana, an intermodal connector assessment tool. The tool is meant to be is a simple tool for assessing the condition of freight intermodal connectors at the state and national level and our intent when we developed the tool was to make maximum utilization of national data sets such as HPMS and NBI and other data sets, and again we're finalizing that tool and hopefully it will be ready for dissemination later this summer.
Now I will turn it over to Jeff who will provide information on the data attributes and the methods that we use for processing and collecting the data through the FPM initiative.
Okay. As Laura mentioned earlier, I am Jeff Short with the American Transportation Research Institute. As Crystal mentioned, we have been working to produce a freight performance measures web application so they can access state performance data in local areas. What I would like to discuss are the methods by which we produce freight performance measures and the applications of those measures that can be used by FPM web users.
First let's discuss the data that goes into the FPM web tool. I will try to answer some of the questions you have before you have a chance to ask them and the first question is: how many trucks are in the database? There are approximately 325,000 large trucks that contribute data to the FPM web database. Second question, what type of vehicles? Generally speaking the database is made up of large trucks. More specifically, while there are less than truckload vehicles in this population, the database tends to be for the most part made up of over the road long haul truckload vehicles. How much data is utilized annually? This system behind the tool will process more than 1 billion truck positions. The facilities that are mentioned, Crystal went over these earlier for the initial FPM web release, dated for 25 interstates available. There are plans to increase the numbers moving forward. Historical timeframe, how far back can we look at data? The data that we will be processing for the FPM web tool will go back to mid-2006 and will be available all the way through the most recent fully completed month. If you're accessing data in April 2009, data through March 2009 will be available. Additionally, we'll have border crossing data that will come online through the FPM web at some point in the future.
Here is more information about the data. This is a visualization of the raw data that is processed into the FPM web. The data elements include a unique truck ID, time and date stamp, latitude and longitude, and interstate identifier for data management purposes.
Continuing through the data processing, walk through to process these records, which number in the billions. There are many steps that are related to data management and fusion, putting multiple data sets together, managing them, making sense of them essentially, and second sending that raw managed fused data to a data processing tool that uses several custom and off the shelf software applications and some relatively powerful hardware.
Next, this slide shows how truck positions and eventually truck speeds are mapped to interstate highway segments. First quality controls are placed on the raw data. If truck positions meet these quality measures, they are shifted to a new database and are snapped in pairs to the highway system. For each pair of truck positions a speed is determined through a time/distance calculation of data from two related truck positions. If you look at the pink line on this slide, each end of that line is where a truck position was given to calculate the distance is traveled, the time between position reads, and you have an average travel speed over that distance.
These values are given to each three-mile segment of highway that an individual truck crosses. Average speeds from multiple vehicles are eventually calculated based on criteria such as time of day, day of week, or whatever query is put in by the FPM web user.
What are the applications of this data? I will go into more detail describing specific applications, but as a high level applications include trend analysis, analysis of the effect of operational conditions, and before and after project is implemented, a highway project is implemented analysis.
As Dr. Hancock will show you in the next presentation, this is the type of data that will be returned to users after a query. The results will be based on the criterion that is given through the query tool, and these results can be analyzed in this format, but they can also be imported into GIS software. The website where we host FPM web will have a route file that will be available for download and can be imported into GIS software, and the information such as what you see here can be imported into the GIS software as well. The data can be visualized on a national level.
I got a little ahead of myself. Here is an example of data being visualized at the state level. State DOTs may want to look at it statewide, but also focus in on certain areas. Here you may want to look at Charleston, for instance. This example shows the data being used to display the effects of a wild fire in Florida several years ago. You see on May 15th on the slide on the left that traffic, truck traffic disappears essentially from the highway that is affected by this wild fire.
Next is an example of using the data to determine the hour by hour impact of a work zone on average truck speeds. This is another thing that can be done with FPM web. Some other applications I just wanted to go over really quickly. A first application of this data is border crossing data. As I said, border crossing data will be available at a later date, but as is shown here, average speeds by time of day on a segment of the Blue Water Bridge which is shown here can be calculated, for instance.
Additionally, the data can be used to assess freight bottlenecks. Here we have a chart of hour by hour average truck speeds at one of the worst bottlenecks in the United States. One final application analysis of truck parking deficiencies is possible through this data set. The map you see here depicts the level of truck demand for highway, highways and compares this information with locations in New York State of official truck parking areas.
Finally, I just had a few notes on FPM web before I turn the presentation over to Kitty Hancock. First I just wanted to touch base with everyone on data privacy. We all know that freight information is often tough to acquire because of data privacy concerns. Such concerns stem from the highway high level of competition among trucking industry members, and things like concern over regulations, concern over litigation. What FPM web offers to transportation stakeholders is data that is delivered in such a way that data privacy concerns are no longer an issue with the final statistics that are produced by FPM web. Therefore, this is a win-win for industry and the public sector.
Second, you will see a terms of service slide in the next presentation. Most of you are familiar with such terms of service agreements when you do things like online like pay bills, download software, and install software, et cetera. I just wanted to point out that such agreements are fairly standard, and there is one with this web tool as well.
Finally, I have a link posted here on the slide, and I believe it is posted somewhere else on this webinar interface, and if you go to this link, you can sign up for a notification when the FPM web tool is deployed to the public. Again, thank you for your time. I will now turn the presentation over to Kitty Hancock from Virginia Tech.
Okay. Well, while they switch presentations for us, I want to thank everyone for attending and giving us the opportunity to give you a preview of the tool that's going to be made available. I do want to emphasize that what you will be seeing is what you're seeing in this presentation is the pre-beta version. The beta version is in the process of being forwarded over to its permanent home, and we will be testing it over the next month or so to make sure that it is operating the way that we had designed it to operate.
I am Kitty Hancock with Virginia Tech. With you are a developer for the software, and recently completed Ph.D. I am with the center for Geo-Spatial Information Technology here at Virginia Tech, and we provide research and educational opportunities for students here at Tech to work on GIS projects and to get hands-on experience with both research and applications associated with GIS.
Moving into FPM web, when you log into the system or when you first click on the link to the system and the link will be available both through the American Transportation Research Institute's web page as well as Federal Highway's web page once it is up, when you click on it, you will come to the log-on page. If you don't have an account yet, you can request an account through the web page, and it will take you to a request page which then once you complete this will be sent via email to the people at ATRI for review, and once ATRI has approved it, you will be provided with a user name and password to log into the system.
As Jeff mentioned, there is an online service term of agreement that you will see, and this will come up every time you log in just to remind you this is important information that should be used appropriately. If you accept it, you move into the application. If you reject it, it takes you back to the web page that you started from. Once have you moved into the actual application, it comes up in a map interface where you can see the extent of the interstate system, and this is RTIS developed a server, and it does have most of the standard functionality of RGIS server. Across down the left-hand side, you can search, you can zoom, and you can change some of the characteristics of the map content.
Down the left-hand side are the standard RGIS navigation tools. Across the top bar right above the map you will see the query tools and comments, and then a link to go back to ATRI's main web page. There are two different types of queries. There is the first which is the customized query which will take you -- which will provide you with summary data, aggregate data in a table format which can be downloaded in a CSV file for import to Excel or some other data manipulation tool. There is also a customized query for mapping which will allows to you download information that can be imported into a GIS software package so you can generate the linearly referenced mapping that you have seen and that I will show you shortly.
This is an example of one of the RGIS queries that can be performed on the map itself. Then moving into the complied query for the data, there are two different components to the query. There is the location information where you specify either the interstate that you're interested in or the states that you're interested in or you can indicate all, and get aggregated data across that location aspect of the query.
There is also the temporal component which allows you to decide which years you're interested in, whether you want to break it down by month, whether you want to break it down by day of week, whether you want to break it down by day of month, and if you want to break it into hourly intervals. The data that comes into this software or this application, the most disaggregate form that data exists in is the average speed but per hour per three-mile segment on all of the interstates across the that are existing in the database. So that is the most disaggregate level that you can acquire the data for when do you a request.
You can aggregate it to any of the levels that you see here and then you can indicate whether you want to report it back out in the data display individually by state by hour or if you want to aggregate it and see just the total values for each of these, and we'll see a couple of examples of those.
If you're doing this on a regular basis, you can actually save your queries. You can generate them, and then save them, and come back and retrieve them at a later date and run them again either with updated data or with the data that you worked with in the past. This is specific to the user name that is logged in when you save the query and when you retrieve it.
The results of the query show up in this tabular form, and you will see the interstate depending on the level of aggregation, the interstate, the state, the year, the month, the day, the week, or day of the month if that that's what your query resulted in, the direction, if there is nothing in the field, it is both directions combined, and east-west, north-south depending on the interstate, and the average travel speed and then the standard deviation of that travel speed.
The export button allows to you export this to a CSV file which can be imported directly into Excel or into some other mathematical manipulation tool. Once you request export, it brings up the standard windows box that requests where you want to save this to locally on your computer. An example of another query where you can actually specify the mile post and this is within your individual state. The data itself is saved in a national mile posting system that was built when the original data manipulation was taking place. So from that if you look at I-10 it is 0 to 1,500 or whatever the length of I-10 is from south to north, or from east to west, but this application does adjust for individual state mile posts, so if you're in Virginia and you're working on I-95 and you want to see the average speeds along the segment of I-95 from mile post 130 to 165, you specify that, and it extracts the appropriate data and provides the results for that segment. If you were to request that information for the map query and download it, it does maintain the national mile posting which matches the route -- the download geo database download file.
When you do the customized query for mapping, it looks almost identical to the query for the tabular report of the data. The difference here is when you generate the results, it comes out in three-mile segments, and it looks something like this, so you would see the interstate, the state it is traveling through, the from mile post to mile post, the year, the average velocity or the average travel speed, and the standard deviation for that three-mile segment. This then can be imported directly into RGIS or Transcad and linearly referenced so you have the display map that is we saw earlier.
An example with Transcad, we did a query on I-40 in Oklahoma between mile post 256 and 330, and attached it to the route file within Transcad, and you can see the resulting themed speed for that segment of roadway in Oklahoma.
Similarly in RGIS we extracted the entire length of I-5 and imported it as a CSV file into RGIS and it created the themed speed map showing I-5 for the entire length. For these two, the mapping component at this time, you need to have the GIS software resident on your individual computers. The query tool itself is a web-based tool, and there is no requirement other than that you have web access.
Then finally there is a box that is available for comments for anyone who is using this. If there is any information that you would like to provide as feedback for the beta testers, this will be a place to give any problems that have been identified. Once it is made available for public use then this form to use an indication of any input that you would like to provide to either us, ATRI or Federal Highway as far as the tool itself goes.
Crystal talked a little bit at the beginning about some of the interest in freight performance measures and data from the public perspective from the federal perspective. I just want to close a little bit with some from the research perspective. I have been involved in freight for many years now, and throughout my career there has been a call for freight data. You can see just from this slide that there have been eight different conferences, studies, reports just within the past nine years that have called for improved freight data. Federal Highway has been active in trying to identify ways to help meet the needs to provide freight data to the public, to the particularly more locally to local jurisdictions, state jurisdictions, and this is one of those efforts to make it available to people for both planning and research applications.
Within this call for data typically the most common areas that have been identified are a standard data architecture which is currently being funded by the national cooperative freight research program, and they're in the process of establishing a baseline architecture or a baseline for an architecture.
Speed has typically been a very important data element that a lot of the modeling, a lot of the planning has really wanted to have that particular data element to support decisions that are made either at a state level or to support some of the models, some of the research that has gone into freight planning.
What we can do with the data Jeff showed you several different applications that they are working with this speed data to provide analyses that have not been available in the past because we have not had this level of information.
A couple of other things that have been of key interest are improvements and changes to the existing data that we have. Speed feeds into that, average travel speed feeds into that very well, data gaps within the data sets. As I mentioned, this is the first time we've really had travel speed at the level that is being provided or will be provided through this particular tool, and then to provide these attributes back into some of the planning models and into some of the research that is currently being performed.
A couple of other unique features, particularly with this data set, is the method for the data collection, the new technologies, and a new way of obtaining information that is not involved in surveys but that can be linked to survey information, and then the dissemination of this information. This is one of the several applications that is being supported by Federal Highway in getting information out to local jurisdictions, and out to states to make it available and useful for those of you who need this information for your decision making.
We saw a couple of examples of the data synthesis and analysis that ATRI has already been performing, and that we hope can be expanded by making this available to those of you who are doing this work. Finally, I want to put in a pitch being from the data side of transportation is we have continually been trying to reinforce the fact that data is infrastructure and is necessary for us to make the appropriate decisions across the transportation system. This application is another step to providing that infrastructure and to maintaining this level as we move forward. It does require money, time and commitment and Federal Highway has shown that, and with this we're hoping that we can continue to get this word out that data is important, it is an asset, and it does need to be part of any of the resource decision making that takes place.
With that, I will be happy to open it up to or turn it back over for questions.
Thank you, Kitty and Crystal and Jeff. Those were definitely interesting presentations, and it is definitely going to be a very useful tool once it is all up and running, and has a lot of potential. I would now like to start off with question and answers. I will start with the ones posted online, and then once we get those questions answered as time allows we'll open up the phone lines for questions. Let's start. I think the first question we have may have been covered by Kitty's presentation, but I want to reiterate it anyway to make sure. The question was in reference to the map showing velocity.
Question: Do you have the reference markers or other geographic markers to show the specific length of road? For Wyoming I assume the section of I-80 is red, in red is near Rock Springs and Green River or somewhere between reference marker 66 and 130. It would be helpful to have specific reference markers.
This is Jeff Short. You will be able to if you have this in the GIS software, and then you can zoom right in. The mile markers will also be listed in the output from the query, so you will be able to see down to the three-mile segment level where and what the issues are as far as what the average speed is.
Jeff, maybe you can just explain a little to the audience about the three-mile segments and the rationale and the differentiation between this and spot speed. I think you did it in the one slide that had the different description of two points and how you used time lapse and distance travel to calculate average speed. Maybe you elaborate that this is not spot speed, it really is looking at speed like an average operating speed between two points.
As Crystal indicated, we are looking at operational speeds here. It is one thing to say at this point on a highway this vehicle was going 57 miles per hour, but it is definitely another possibly more unique thing to say over 35 miles this was the average speed of this truck, and thus it starts to build a picture of what trucking operations are doing over periods of time. In one of the slides described that. Points may be 60 miles apart, and the vehicle is moving across the 60 miles in 60 minutes, and then we can all calculate the average travel rate. Of course, and that travel rate will apply that to segment during the time period it was moving.
I just wanted to reiterate on the mile posting the application allows to you pick mile posts based on your state mile posting, but when you bring it into your GIS, it is based on the entire interstate mile posting which was created when the data were being processed from the original raw data. You will be able to determine it based on the state boundaries and so on within the geo database.
Another thing I would like to add to the discussion of spot speeds, there is a much larger database of spot speeds that we may incorporate into this web tool at some point. Of course that would be disclosed, but it may be a completely separate application, or it may be worked into this operational speed method that we're using currently with the FPM web tool.
Question: What are the area polygons in yellow on the FPM example page?
What he is referring to is the yellow that's here showing Virginia. This was just a query within the AHRQ GIS server itself where we were specifying state name equals Virginia, and you run it, and it highlighted the Virginia polygon. You can do this with segments of routes or any other layer that exists within the GIS software.
On the opening page were those MPO regions, the very first convenience.
Thank you for clarifying that for us. We had a question about what is the level of confidence in the data.
This is Dan Murray. I did pose the question to the questioner about what level of confidence she is referencing. There is the legitimate question about the resolution, accuracy of the GPS or position point to where we're mapping it. There is sort of a confidential level question about the percentage of points that are within some polygon, so we need to know especially what level of accuracy to what data given there is so many different aspects to the system, what level of confidence or what degree of accuracy they're referencing.
In some ways, the most practical answer to that is how accurate is our calculated speed to the ECM engine speed. We did do some limited testing of that in a different project using some FPM data. It was an assessment of the role of speed limiters in safety, and I believe, Jeff, you might want to jump in here that our ECM engine speed data was at least in one noticed 56.2 and I think we were at 56.8 miles per hour, but in total we have never really done sort of a comparison or a probe of ECM speed data against our FPM calculated data. Anecdotally Crystal and Jeff have done some assessment of that, and I think Crystal it is safe to say you have been fairly comfortable.
And I think the point that Jeff made earlier about this not being spot speed data, so in some instances depending on how free flow and how much congestion there is, you could expect that some of the average operating speeds might be aligned with for instance what you get from an ATR.
The data that's currently going to be available in the web tool is not spot speed data so it is always going to be based on average operating speed over a given segment of highway. We have a planned effort with our and FHWA's Office of Policy that collects ATR data from all the states to do some comparison of our speed data with some of the ATR data. We do have planned enhancements for the web tool, so again as Kitty point out, just actually the pre-beta version, but we'll do some beta testing with the small select group of folks and we'll get feedback on how it can be enhanced, and we do have a research item in our next research budget that allows us to do some enhancements to the web tool. The intent is to get it at least initial version out into the hands of the folks who make the decisions and use the data and to get feedback and use that feedback to make whatever enhancements we can given the data set we have available.
Okay. Thank you. We also have had question about what about adding in crash data.
Question: How do you know where, when there is a construction or maintenance zone, and how do you anticipate the usefulness of the standard deviation? Will there be average ranges, et cetera, available?
I will answer the first part. At the national level if you remember my presentation, I said that for the most part we're focused on deriving national level measures, so at the corridor level it was sufficient for what FHWA internal needs were at the headquarters level. So the intent of putting this down to the state level is that you have the data and you begin to look at if you have a plan work zone or a work zone in place, you will be able to use the data before, after and during the work zone. The work zone example that Jeff presented is a result of a coordination effort between two entities within the office of operations, so we have a work zone management team within the federal highways office of operations, they came to us and said we have this project ongoing, can you use this data to look at the impact and effects of a work zone. We don't have any national work zone data set that we can add as a layer to this, but the intent is that as a tool that you can work with MPOs and others have an interest to look at the work zones, and how the freight movement is impacted by the work zones, if that's a performance measure you want to look at, so there is no national work zone database that we over laid data with, but if someone would come to us and say we have this planned work zone, rather it is a resurfacing, complete lane closure, whatever, we can use it to get the impact of the work zones. There is no national effort to add a layer with them, a national construction or maintenance zone. I will let Jeff answer the last part about standard deviations and average changes.
I would just add to that question in Crystal's answer that this tool shouldn't necessarily be used to identify where work zones are, but should be used to assess the impact of different types of work zones on freight movement, so you may be able to take that data and extrapolate and say in the future we're going to have a work zone here, what does that mean for freight movement? How much is that going to cost trucking for instance? So I will leave it at that. Hopefully that answers the question.
The question about the standard deviation, the way the data are aggregated to get to the three-mile segments per hour is any truck that crosses that three-mile segment, the speed of every truck that crosses that three-mile segment during that hour, so the standard deviation just provides you a measure of the variability across that three-mile segment across that hour of when the data were collected. It is really just probably a spot check if you will to make sure how much variability you see in that speed value and it is more for your analysis and not really for any reporting requirement.
Okay. Thank you. I do see there are quite a lot of questions coming in, and we will try to move through these quickly so we can cover all the questions that you're asking. The next question I saw is:
Question: How many motor carriers were involved with this research?
It is continuous, so it is not over. It is a continuous research program, and as Jeff mentioned, this particular web tool is based on approximately 325,000 vehicles, but the FPM initiative itself has somewhere near 500,000 vehicles that are both not only commercial truck but also trailer tracking devices are included also. We don't disclose who our partners are, but in terms of the total number of vehicles for the FPM effort overall, it is nearing 500,000, and for the FPM web tool, the initial version which will be rolled out, and that was described in today's presentation, it is about 325,000 nationwide, and in Canada, too, even though our current web tool doesn't include Canada and Mexico.
Question: What share of the truck traffic in these sections is measured?
Jeff, maybe you can talk about perhaps the future capability of saying how many vehicles were included in a speed, but that won't be an initial feature in the rollout, but you can talk about that, Jeff.
Since we don't collect data on the vehicle that is we don't collect data on, it is one-way to say essentially what's happening at one interstate route at one any given time is kind of unknown if it is not part of this data set. You can say there are 325,000 trucks in this and there are 3 million large trucks, therefore we've got this percentage covered, but if you look at just a three-mile segment of highway, that number is not readily available as far as what the mix is on that segment of highway.
I think we have looked at some limited instances where we have HPMS and cord and count data, and the truth is our database in terms of the representativeness of all large trucks which are typically classified as class 7 and 8, 26001 and over, usually five axle, our percentage ranges anywhere from 10% or more of all trucks out there to percentages that are probably in the scale of 25 to 40% of the traffic mix when compared to the cord and counts, so that's another way of saying if there is X number of trucks a three-mile segment, we could have as many as 40% of that percentage of trucks out there, but as Jeff points out, we would need to have the independent data on how many trucks are in a particular segment for us to estimate our representation. We can estimate that immediately and easily once we know what the total count is on the quarter and the segment or in the network because again we know what our discreet numbers are.
Okay. Thank you. Our next question:
Question: How feasible is it to add goods, value or weight to the data set? Are these data readily available for surveys or some other form? This would be of interest particularly for cross-border freight traffic.
Currently we do not have commodity information. For the most part we just analyze the data elements I showed earlier. We also have spot speeds and whatnot, but as far as commodity goes, that's something that we meant to leave out of the data set in order to protect privacy, et cetera, and also often times that information is just not linked with this data source at all.
I can't speak right now to how it could be integrated with the freight analysis framework, for instance, but certainly there are possibilities to use the two data sets in combination with one another to understand not just the speed but the value of the commodities from freight analysis framework over the interstate highway system.
The freight analysis framework right now is reporting annual values, and you can aggregate the speed data, too, in annual speed travel speed which then could be directly correlated with the data to break it down into hourly segment or hourly data isn't practical, but potentially you could do some modeling or something that could bring it down to a daily or a weekly or some aggregation in that fashion.
Question: If that data is in 60 mile increments how effectively can this application be used within metro area?
I am not sure what he's referring to with 60 miles, because it does get down to three-mile segment, so you can use it as three-mile segment.
I think Jeff may have given an example using 60 miles. It is good to know that it does break down to three miles or will be feasible for metro area to find this useful.
I don't know, you know, perhaps when we put out the guidance and user guide that comes with this we'll have up front maybe some charts that describe the attributes of the data because even though some of the data points are as infrequent as every hour, there are some points that are as frequent as every ten or 15 minutes, so all of that information will be disclosed about the attributes of the data in some format when we fully implement the web tool.
Okay. Our next question on do you have comparative data between trucking and auto trips? You see limiters installed on trucks affect the speed or calculation of reliability analysis?
I mean, some of the these truck does have speed limiters, so the characteristics of the truck could have some bias because there are speed limiters on some of the trucks, but I will let Dan maybe elaborate on that. I am not sure in terms of the speeds themselves are still going to be characteristics of the vehicles operating on a particular segment at a given point in time, but the fact that some of them have speed limiters can put some bias in the data, but I would venture to say these carriers are the more safety conscious type of driver that we to want have on our highway system so the speed limiters is not a bad thing, but I will let Dan or Jeff maybe elaborate on the impact the speed limiters on the data set.
I would say because FPM is attempting to capture what is rather than conceptually what ought to be, the fact that major carriers, for instance, Con-Way is limited at 62 miles per hour, Snyder 65, this is the ideal database for not speculating and modeling on truck speeds but calculating what the real truck speeds are and in that case it becomes even more imperative that we're able to document posted speed differentials for instance in California, 70 versus 55 along with some work that professor Steve Johnson has done with ATRI out of the University of Arkansas, that this shows there is natural differentials and natural speed inclinations between automobiles and trucks and the safety differentials are critical so we're not presuming that all trucks are driving the posted speeds in some instances. They're higher and in some instances they're lower in others, and so this is what is documented by FPM. I would say that's probably one of the strengths of the data.
One more thing to add to that is one of the major uses of this tool of course is to identify congestion in areas where trucks are not moving and speed limiters aren't going to affect these areas. If a truck speed limiter is set at 62 mph, that isn't really going to affect how bad congestion is as it shows up in the FPM data because the truck will not be moving fast enough for the limiter to engage.
The next question is a clarification maybe for the work zone and incident question from previously.
Question: how do you know if the work zones or incidents were in place when the data was being collected?
The data is continuous data collection, so there is no lapse in data collection. We get data 24 hours a day seven days a week regardless of whether there is an incident or not. The only way we were able to do those descriptions of the work zone and the wild fires in Florida is because somebody came to us after the fact and said we had this event, what was the impact on freight movement, so those examples were based on somebody telling us. For instance, in the North Carolina work zone example that there was a work zone in June 2008 and the wild fire was we went to look at the impact of a known wild fire. It is not going to be able to look at these incidents in real-time. That's probably another point that we didn't give a lot of attention to in the discussions thus far is that this data is historical data, so it is not intended to be for traveler information or for any type of application like that. It is always going to be historical data. Normally, as Jeff point out, it will have about a 30-day lag, so you won't see April 2009 data until May of 2009.
I would add we're working closely with Al Monday from TTI -- Jerry Al Monday to look at applications work zone, so there may be more coming in that area.
The application is set up so you can go back and specify to the hour, day, whatever, of an incident. So you wanted to go back and look at the impact of an ice storm you know occurred such as when our fly over ramps froze last year, you can go back and extract the truck speeds for that section of road for that time and then look at it on either end as well.
Great. We are almost through the questions.
Question: Are there any plans to eventually expand the data coverage to non-interstate, non-national highway system routes?
From the overall program, we already have access to more data than we know what to do with, so in terms of coverage, our analysis and the tool is based on those 25 interstate corridors, but in addition to the actual web tool itself, here Federal Highway you can contact me initially if you're a public entity, and we have a provision within our contract with ATRI to look at non-interstate routes. We have access to data on non-interstate routes that just won't be in this version of the web tool, and I know Jeff has already talked to me about probably some very near to medium term expansion plans to include the entire interstate system, but the non-interstate routes I am not sure when they'll be available in the web tool. If you're a state DOT or MPO and you have specific routes that you want some sort of ad hoc analysis done on for a non-interstate route, then we have the capability to get access to that data. We just have to come into Federal Highway as a request and we can consider it on a case by case basis in terms of when we might be able to provide a data product for a non-interstate NHS route. In terms of our data coverage, our only limitation is whether or not there is a significant amount of these vehicle that is are included in our data set traveling on that highway, so if you for instance look at a dredge operation where you don't have typically long haul vehicles but more of the pickups and the smaller types of vehicles operating if they aren't equipped with this GPS technology and aren't a part of the data set that we have access to, we then we won't have data but our major limitation is not necessarily by route, it is by whether or not there is a significant amount of vehicle that is are included in our data set traveling in a particular route.
Okay. We had one last question which Dan Murray was kind enough to respond to. I will summarize the question and response here. The question was whether or not the truck counts are available by section, hour, day, week, et cetera. As Kitty and Dan have mention on here, the data can in fact be analyzed by parts, days, weeks, months. I think that's the end of the typed in questions.
Maybe the person asking the question can say if they are looking for the truck counts or the truck speed return. Are you asking is the web tool going to give you the ability to see the number of trucks included or the average speed by day or week? What Dan provided an answer to is yes we can provide an average speed by day of week, part of the day, et cetera, but the question says truck counts, so just to make sure the question was fully answered are you talking counts or average speed?
My reference, as Crystal indicated, is you could look at the truck speeds by every five minutes, 15 minutes per hour, day of week, every 2 a.m. in the month of March over the three years, et cetera, but whether if you're asking specific truck count it looks like Mary Lee is asking that, then I don't believe this tool generates that data. Is the data knowable, yes, but through this tool, I do not believe that's true. Do you agree with that, Kitty? The way that we aggregate the data for this tool is it is straight average speed per segment per hour. There is only measure is standard deviation. There is no N to that that's saved in that formulation.
We did consider that, so I am not sure of the technical why or why it is not included in this version, but we will continue to investigate why the N can't be displayed. It is possible, but it is just not in this version.
There is some concern that it could be misused as a truck count which it is not.
Right, this only captures the ones that have that you're collecting data from, so if you only have 10% of the truck traffic in that area, you're missing 90% of the trucks, and you can miss use that N as being a wrong truck count.
Okay. Thank you for clarifying that response. We have a few minutes, so if you would like to ask a question over the phone Diane will provide instructions on how to do that.
Hi. I just wanted to comment, I think this looks like a really great tool, and I thought I could clarify the reason I was asking about the truck counts because we are able to take a segment of road using our APR data and give an estimate of what type of truck would be on that section such as by classification of the truck, and so I thought if we had the number and we knew it was either a 7 or an 8, then we would know what percentage would be represented on that section based on our other data, and but I mean if we have that like if I gave you that information, you could return with some of that information back to me.
Yes. I mentioned that our along with our FHWA Office of Policy, we have been talking about doing just that. Great point that again we continue and this is a first step, and we hope to make improvements, get feedback. Any feedback provided will be beneficial and helpful in making this a better tool.
Okay. I had one other question. You said it was classes 7 and 8 that were largely represented. I understood that 8 through 13 are what are accepted as a commercial vehicle size or class?
It is interesting. Industry and government use different truck classifications, so some day we'll get that ironed out, but essentially these would be 26,001 through 80,000 pound with LCDs in some instances, and so it would in theory cover up through your class 13, in some cases 14.
Thank you. We did quickly post a poll question to get your initial feedback based on what you heard today if you think the tool is going to be useful to you. You can always contact the presenters or myself if you have questions, or have any comments that you think might be useful right now. Thank you all for attending today's seminar. The recorded version of this event will be available within the next week on the Talking Freight website.
I'd like to give a brief mention about the FHWA Freight Peer to Peer Program. The Freight Peer-to-Peer Program (P2P) puts public sector freight transportation professionals in touch with experts in the field and provides technical assistance in order to enhance overall freight skills and knowledge. The program is available to public entities, including State departments of transportation (DOTs) and metropolitan planning organizations (MPOs). To learn more about the program or to arrange a peer exchange, or to discuss participating as a peer/expert please visit the Freight Peer to Peer web site. The next seminar will be held on May 20 and will be about Higher Productivity Trucks.
If you haven't done so already, I encourage you to visit the Talking Freight Web Site and sign up for this seminar. The address is up on the slide on your screen. I also encourage you to join the Freight Planning LISTSERV if you have not already done so.
Enjoy the rest of your day!