7.5 Webinar for Next Week

SHARON HAWKINS: [AUDIO OUT] and we have about 16,400 plus miles of state highway. We're the 12th largest state highway system in the nation. And then just a little bit on the topography of Arkansas. A lot a mountainous regions, rolling hills, and a nice flat delta region. A lot of different terrain as far as where we're building our roads, and a lot of different terrain to analyze when we're looking at our crash data. So I want to do a few statistics about our crash data before I go into the tools that we're using.

So here's a chart showing the total number of crashes in the state, and total roads versus just the state highway system for the past 10 years. And you guys can see, about 60% of all crashes in Arkansas occur on our state highway system. 70% of all crashes in the state-- total crashes-- occur in our urban areas. And then 70% of all crashes that happen in the state that have a fatality involved occur in our rural areas. And 2/3 of those are roadway departure related.

There were 550 crash fatalities in Arkansas last year. This next chart kind of gives a comparison on how our fatality rate versus the national fatality rate. And you guys can see, we're just a little bit above what the national fatality rate is. All this being said, with our crash location tool, it's very important for our new initiative towards zero deaths to make sure that we have all the spatial information that we need in order to make our routes safer for our traveling public.

So I'm going to go over kind of past to present-- the different kind of tools that we've used in the past to locate our crashes. And then I'll be introducing our new tool that we're going to use. So we've been progressively implementing new methods over the years. And so for the next few slides, I'm just going to show what we've done from the 1970s through 2011, before the implementation of our new tool.

So in the 1970s, the department was tasked with locating and maintaining all the locations for crash events on the state highway system in Arkansas, and just on the state highway system. So for this purpose, the department decided to record each crash event with that event's county, route, section, and log mile location so they could be found and studied on the highway system, since the system uses the same county, route, section, and log mile set up. And we live by the log mile here at the Arkansas Highway Department.

The department partnered with the state police and local county and city law enforcement to make sure that they could locate each crash with its needed attributes, including an accurate log mile of where the crash event occurred. So we were relying on other law enforcement-- law enforcement that weren't familiar-- very familiar, anyway-- with the log mile system to get a log mile for us.

So this slide shows one of the earliest tools that we would use, which is our county, route, and section map. It only shows the state highway system on the map. Of course, it has the beginning and ending log mile for each segment of highway, and it has a log mile for all the bridges on the map.

So we were asking them to give us an accurate log mile for a crash location based on these maps. You guys can imagine how difficult that was for them. But there was a backup. And we also-- here at the department-- we had crash locator positions. And they had to locate-- relocate-- all those crashes that happened on the state highway system.

So next, we also had our tabular road inventory information in these big books that they were able to flip through. And for each route, wherever it intersected another state highway system, of course, there was a log mile listed, and then events that happened on the route. So a number of lane changes, ADT changes. There were log miles listed for that.

As you guys can see from the picture, very difficult to understand. Our crash locators here at the department could hardly understand it, so it was very difficult for us to think that other law enforcement were going to be able to understand it also. So once again, every crash had to be relocated here at the department, even though we were getting a log mile from our law enforcement.

So we also provided static printed maps, and our workers here would write in the log mile wherever there was an intersection with another route. So we would take these maps. We would hang them up with the city so they would have a static map on their wall to look at to better locate a log mile. But you know, those maps could be out of date, you know, as early as the next day. There's always system changes on the highways. So this is not a very efficient way to get it done.

So next, we offered that we could make these maps-- we didn't have to handwrite anymore. We could use MicroStation. And we could put log miles in where routes intersected the state highway system so they could have a better feel of what log miles were available for them to choose from to report the crash location. Once again, maps could be out of date very quickly. And even though these were available digitally, most cities still relied on their printed versions that were hanging up on their wall.

So in 2004, we realized the power of our linear referencing system and how it could help our crash locators here at the department locate exact log miles of where crashes occurred digitally in our GIS software. So we set up the program with our CAD maps. And we had our linear referencing system on top of those CAD maps so they could see intersecting routes with our state highway system.

They could hover along using a tool called LRS Precision Location. As they hovered along the state routes, they could get an exact route section and log mile of where their point was. Very helpful for our crash locators here at the department. But we couldn't share this tool with law enforcement, because they didn't have the software, and not the knowledge, either, of the linear referencing system.

So still, we were in a continuous cycle of us having to relocate all crashes that happened on the state highway system to make sure we were getting an accurate log mile. And we were lucky. Our Arkansas Geographic Information Office had gotten with the counties, and they had developed a 911 center line routing system for the state.

So thank goodness we had a whole state network that was attributed with road names and such so we could utilize that with our linear referencing system, and they would be able to see the routes, their attributes, their names, to still better identify a crash location. But once again, that was a tool that we could only use here at the department, and other law enforcement just couldn't get to it.

So we also tried, during that time, GPS units. We bought 50 GPS units, sent them out with the state police. We wanted to do a trial period. Maybe they can use these GPS units, take a reading at the crash scene, get a lat/long for that crash, get it into the system. And then we would have a latitude/longitude that then we could convert to log mile for our report. So we tried this for about three months. We gathered the data.

I've heard this from many other states-- when we got the data back and mapped it, a lot of those dots landed in the state police parking lot, or they had moved to the side of the road, so-- in other parking lots where they were. It just wasn't convenient. It wasn't going to work. Not necessarily the law enforcement's fault. They didn't realize how important this was to us to make sure that we had these correct locations. And there was minimal training when we did give them those GPS units.

OK. So just a little bit about our linear referencing system, which was a major player in the tool that I'll be introducing in just a minute. So as I said before, over 16,000 miles of state highways on our highway system-- 12th largest in the nation. And also included on our linear referencing system, over 15,000 miles of other routes eligible for federal aid. So what you're looking at here on the screen is pretty much our functional class system-- so all routes eligible for federal aid. It's what we have currently on our linear referencing system.

So our linear referencing system sure works like most other DOTs. We assign the route with the county route, section, and beginning and ending log mile. The concatenation of the county, route, and section and the log mile information is its unique value, or its primary key field. And then we can map different things from around the department that also have that same primary information, such as job status, information from our road inventory, and of course, crashes.

So updating the LRS, 177 changes were made just in 2012 on the state highway system. We have removed routes, added routes, relocated routes, and often a lot of relogged routes throughout the year. We keep a live version of our linear referencing system open to everybody here, but we also keep archived copies. We started archiving our linear referencing system in the year 2000.

So we can always go backwards to see what the system looked like back then, which is helpful. We just received the 2011 statewide crash data, so we can't use current linear referencing system to locate on. We're still locating crashes that happened in 2012, so we do have to keep these archived copies to make sure that we're locating in the correct year that the events occurred.

OK. So let's look at one of our typical crash reports that we get here at the highway department. It happened on the state system that we have to relocate. Of course, all the information here. You see a little narrative about where the crash happened and a little diagram. This particular crash says it happened on US Highway 82 in Columbia County, section 3b, log mile 2.20. So we're going to relocate this crash and see how the law enforcement did out in the field.

OK. So from the narrative, we see where the red circle is in the middle of the map, that this is approximately where that crash happened. But they said it happened at log mile 2.20, but this route only goes to log mile 1.65. So couldn't have happened. So we're going to have to relocate this. OK. So after our locators got a hold of it, we relocated it to say, now it happened on Highway 79, section 1, at log mile 18.26.

Those are going to be sent back to our state police. The location is going to be re-entered into the database for the correct crash location. So once again, that's a prime example of how difficult it is for our law enforcement to give us the correct log mile. So back in February of 2011, one of our county offices-- Polk County Sheriff's Office called. We just had a big snowstorm. We were still waiting for the snow to go away.

And they said, hey, we need that map that we're supposed to use to locate crashes by log mile on the state system. Can you tell me where that is? So we directed him to the route and section map that they've used. And he said, hey, is there an easier way? And that very day, we sat down, and we came up with an easier way. And we don't blame them for asking.

In this instance, here's a route. It's about 19 miles long. There's no log miles in between for them to guess, so they have to use the little section lines to kind of guess where that log mile could be. And they have 52 different fields of information per crash to enter in. A log mile's not necessarily on their radar as the most important field. So a lot of human error. And then with our locators having to relocate all those crashes, sometimes we weren't getting those locations in in time to really do some real-time analysis with our data.

So, making it easier-- so what we did is we took our linear referencing system that included all routes eligible for federal aid around the state, and we asked our GIS platform-- we asked GeoMedia to put a point every 100 feet along the linear referencing system. So when it did that, each point was populated with the county, the route, the section, and the exact log mile of that point.

So then we say, OK. Well, let's make this a little bit better. Let's attach some of our road inventory data to those points also. So now, with county, route, and section, and the exact log mile of that point, was other information that was going to be very useful to the law enforcement to fill into their reports, also.

Well, then we took the points and we said, hey, let's shove these out to a KML file for display in Google Earth. So we did that through our program. And as you can see from the screen, in the red, we have what's on the state system. In the blue dots, we have other routes eligible for federal aid, or the rest of our functional class system. And you're able to click on those dots in Google Earth and see the exact information that was generated in our GIS platform.

Of course, these are KML files. We can email them out to anybody. You can email them out to family members. They would be able to see. So we started emailing these out to our state police for them to be able to give us a more accurate log mile of where their crash events occurred. And it's just worked out wonderfully.

We also included our city limits, as you guys can see in the top left of your screen. There is a field they had to fill out that say whether or not it's in the city or out of the city. And if it's out of the city, they have to measure how far out of the city it's out of. So we have our city limits available to them. Instead of just being able to locate on main lane, they can now see the log miles on frontage roads and on ramps.

And of course, they have Google Street View available in here to be able to help them better locate where they are in the field. And of course, this allows them-- if they have connections, they can locate right there in their vehicle at the scene of the crash. If there's no time, they can always come back to the office and, of course, use this tool also to locate the incident.

In this case, this is a pretty long bridge that we have here in the state-- the 430 bridge. The beginning log mile of the bridge with 9.87, if you guys can see in the graph on the corner of the screen. So we had a lot of accidents that happened at 9.87 on that bridge.

But as you guys can see, every 100 feet along that bridge, there's a lot of dots. It's almost two miles long. So they're able to give us a more accurate, better log mile when they're out in the field. Not only are the law enforcement outside of the Highway Department using this, but also our crash locators are taking advantage of this tool and now using it to locate those crashes.

So how can everyone use this tool? We can email out the KMLs to everybody. We have FTP sites that's open to the public. Anybody can download those KML files. We also have all these points, or dots, every 100 feet available on Arkansas' GIS clearinghouse called GeoStor. In addition to that, our Arkansas Geographic Information Office offer this data available on ArcGIS Online.

So we have all of our dots in there. There's a link that anyone can click on and go in there and see the same exact information that was offered in the KML files. And also, ArcGIS Online can be used in their smartphones, also. So if they don't have their laptop open, or a good connection, they can also use their smartphones to be able to see that information.

So the use of the LRS location tools, or our new crash location tool, the training and outreach to law enforcement agencies started in June of 2011, so just three months after we came up with the methodology. All of the state police agencies are using this, which account for 25% of all crashes statewide, or about 40% of crashes on our highway system are being located using this tool. And we constantly reach out to counties and cities. Right now, it's just a volunteer basis for local law enforcement to use it. But we have about 42 of those agencies that have started to utilize it, and we've gotten very positive feedback.

So once we have the data, our better locations that we have. Of course, numerous planning studies are done in order to figure out, hey, do we need a passing lane here? Do we need more lanes here? Maybe a turning lane here to make things safer. Maybe some rumble strips.

And then we can also run all kinds of queries, as far as types of crashes, to be able to look for areas that might need to be paid attention to for more safety initiatives to take place. As far as safety initiatives, having this data, having these locations in a more exact location for us, we've been able to prove that different things, such as cable median barriers, need to be installed in more areas around the state. Rumble strips need to be installed, especially on some of our two-lane rural highways where run off the road crashes were very prevalent.

And then we have one more tool that we started using. It's called Intergraph's Incident Analyst. It's been able to help us identify clusters and locations-- areas-- that we really need to pay attention to that we haven't been able to find as quickly without this software. So case in point, this is a study we did in Jacksonville, Arkansas. We had to make a map for our administration so they could see the clusters of crashes that happened in Jacksonville.

The map on the left, you guys can see all the clusters of crashes-- the yellow dots there. This map took about four hours for us to put together. A lot of these crashes happened at the same exact log mile, so the dots are just right on top of each other in our GIS platform. So we exported it out to our CAD. We took those dots. We moved them around so we could produce those clusters.

So once again, about four hours. But with Incident Analyst's hot spot analysis tool in our GIS platform, it took about two minutes for it to generate this hotspot map for us. You guys can see that the two maps compare very well with the clusters. On the right hand side, the hot spots in the red are showing where those clusters are that you guys can see on the map on the left in the yellow.

And then there's also a incident count tool. You can count incidents within any boundary area. So in this case, we're using county. But you can also do cities. We do a lot of House and Senate districts if our legislators want to know where most of the accidents are happening within their jurisdictions.

And there's also a repeat incident tool that we use to run-- this is on a statewide basis, but we can run it on any sized project area in the state. But it lets us know quickly where repeat incidents happen over and over again at the same log mile location so we can really focus on those areas and do some study analysis to find out if there are some safety issues within those areas.

They do have a tool in there also called Change Over Time. It kind of plays a little movie for you, what happens with these incidents. I just did January through June for this. It's in PowerPoint, but once again, imagine it's a movie. When I flip to February, keep in mind that we had a pretty big snow storm this year, so a lot of crashes happened that month.

And there's also temporal reporting. It makes nice charts for us to be able to show where the majority of the crashes happened. So at the top, you see month of the year, in the middle, day of the week, and then on the bottom, time of day-- crashes that happened. And this helps us put together nice figures, especially for public involvement.

We used this at a public involvement meeting. We did hotspots of the crashes. On this particular study segment, we wanted the public to know that we know that there are a concentration of crashes that happen in this area, and it's under study, and we have plans in place to rectify that. And also, we had our temporal reporting on top of that figure.

OK. In the future, a methodology will be in place by June 2013 to include all public roads in our linear referencing system. So from our about 30,000 miles that we have available now in our linear referencing system, approximately 110,000 miles will be included in there. You can see from the figure, in the red, this is about how much we have to add-- about 74,000 miles we're going to have to add for our linear referencing system.

But when that's done-- and we guesstimate in approximately five years-- we'll be able to locate all crashes in the state, not just on our federal aid system, which, the safety initiatives-- there's nothing like having more data to do analysis with to find out what needs to happen.

Of course, the dual carriageway system will be put in place along with that linear referencing system. So each divided highway will have two centerlines, and road inventory will include complete data for log direction and anti-log direction. Right now, we can only map on our linear referencing system based on one centerline. But you guys know, FHWA has asked us to do dual carriageway along with an all public roads linear referencing system.

And in 2014, the Highway Department, along with the Arkansas State Police, will start implementing eCrash, which was developed by the Center for Advanced Public Safety, or CAPS, at the University of Alabama for a paperless, electronic submission of crash reports. eCrash will have an actually map click tool that our law enforcement will be able to use in their vehicles. They'll be able to click on a location. It will give a latitude and longitude of that crash. But they are going to be using our points-- every 100 feet points-- within the map click tool, also, for log mile information.

OK. And this is just to illustrate we're just skimming the surface of what we can do, and look at all the potential that's ahead of us, especially when we get our all public roads linear referencing system in. OK. And contact information right here. For more information on Intergraph's GeoMedia, or Incident Analyst, James O. Brown from Intergraph's. And then more information on eCrash, there's a link on there for you to look at also. And then I'm happy to answer any questions.

Thanks, Sharon. I'm going to go ahead and open up the lines for question and answer.

[BEEPING]

PRESENTER 2: This conference is now in question and answer mode. To alert the speaker that you have a question, press 1, then 0. Each question will be asked in the order it was received.

[BEEPING]

MARK: So you should have heard some instructions. If you've got a question on the phone line, just press 1 and then 0. You can also enter in questions into our chat pod, especially if you're listening through the computer. Let me just quickly see if we've got questions on the phone. No, we don't have any questions on the phone right now. But let me see.

So Sharon, you mentioned that one of the issues that you had rolling out different-- the first couple attempts at something like this, the training seemed to be-- play a role, or trying to understand how the users, especially the state police, were going to try to enter in the information. It sounds like that what you have now, it was much easier. But are there still issues you're running into in terms of getting the--

SHARON HAWKINS: [INAUDIBLE].

MARK: --location information?

SHARON HAWKINS: Well, it's true right now, when they click on one of those dots and see the information they have to enter in their report, they do still have to enter that in by looking at what they see in the crash location tool and Google Earth, and typing it in. So there's a lot of human error in there. But aside from that, we've been constantly checking the locations that have come from this new crash location tool.

And the department's made a decision that as long as they use this tool, which there's a checkmark they can check to say they used the tool, we don't have to relocate those crashes anymore, which means, for us, that we're getting our data-- our crash data-- instead of a year and a half later, we expect to see our crash data eight months after the year ends, instead of a year and a half later. And then we hope in the future, it just comes on faster.

MARK: Mm-hmm, mm-hmm. Because I know a lot of states are-- that's one of the issues, is trying to either make it easier for the-- especially the state police. They're involved in entering a lot of this information-- making it easier for them to somehow put that in.

SHARON HAWKINS: Right. And that's why we're looking forward to eCrash with the map click tool. It's a paperless system. When they click, a lot of our road inventory information, and of course, that log mile information, will automatically be entered into the report through the eCrash system.

MARK: Mm-hmm. Looks like we have a question. A user that uses ArcMap, they want to know--

SHARON HAWKINS: Right. so-- yeah. We do have ArcMap here at the Highway Department. But GeoMedia is our GIS platform. So I hope maybe that somebody else that's on the conference would be able to answer that question. I know that ArcMap has the capability for hot spotting, but I am not quite sure exactly where that is.

MARK: How did you learn about the Incident Analyst product? How did you find out about that?

SHARON HAWKINS: Intergraph, actually, was advertising it. A lot of-- like a crime location tool where incidents happen over and over again. And then they worked with Ohio data, and they kind of started advertising it as a hotspot tool for crash analysis. So when we saw the presentation, we knew that it could be a good tool for us to be able to take advantage of for hot spotting.

MARK: Mm-hmm. Let me just quickly check on the phone lines to see if we've got any questions.

AUDIENCE: [INAUDIBLE]

AUDIENCE: [INAUDIBLE]

MARK: Hello?

AUDIENCE: Hello?

MARK: You've got questions? Yeah. What's your name?

AUDIENCE: This is David Lawler with TDOT.

MARK: OK. What's your question?

AUDIENCE: I think that we had seen that, too, that Intergraph started with, was maybe about a year-- two years ago.

SHARON HAWKINS: Right.

AUDIENCE: It looked good. They've come up with another tool that we built inside of what we call our [INAUDIBLE] system. And after that data is collected, it goes into a repository at the Department of Safety. Then we extract the data and put it into what we call our [INAUDIBLE] side over here in TDOT where we put all of our crash data. And that includes location data-- the county, route, log mile, et cetera.

And they came up with an automatic-- it's an automatic updater inside of our [INAUDIBLE]. And it looks for all the electronic reports that's inside the [INAUDIBLE] database at the Department of Safety. And it extracts those and pulls them over. Now, if there's a tolerance problem, or the officer can't get the right location out in the field, and we've got-- [INAUDIBLE] got a MAP-IT tool that was built by IBMS-- the vendor was. And it's probably about what you're using, too.

We've got a-- where the officer's sitting there, and he can take a Google Map and find the location. He can go up and click on the spot where the wreck is, and then draw it back to where his car is sitting. And we've got a pretty good accuracy rate with that. Now, all the troopers have got them, and now it's starting to go out into the other agencies.

SHARON HAWKINS: Into the local agencies?

AUDIENCE: Yes, the counties and the cities. We are getting fairly good with the cities. They're getting better all the time. We're just about-- I think about 80% electronic now. Do you remember, Ken? It's close to 80--

AUDIENCE: Yeah.

AUDIENCE: --about 80% electronic. We've still got some paper stuff that has to be handled in a different way. But we're happy with the tool. It looks good. And we also have a contract with Intergraph, too. So that door is always open there for them to come up with that, too. Right now, it's depending on the GPS stuff coming off the electronic reports out in the field. And it seems to be working pretty good. It's fast.

SHARON HAWKINS: Yeah. I'm very pleased with Google. Thank goodness it's free, or--

AUDIENCE: Amen.

SHARON HAWKINS: --the tool might not exist. But very happy with it. I see another question up here, Mark. What accuracy level are you shooting for? Plus or minus 50 feet with our crashes. Yes, that would be wonderful. In fact, our accuracy level-- we did a point every 100 feet.

Our accuracy level before this tool started-- with our law enforcement, we were sometimes finding up to a half mile it can be off, which, you know, things change on the system very frequently. So that just was not going to do for us. So yes, plus or minus 50 feet is wonderful. But we're finding that we're getting our crashes very nearly where they exactly happened.

MARK: Again, if you've got questions, on the computer, you can put them into our chat pod. I just want to make a quick plug, too. I think back in 2011, Tennessee was actually one of a couple of states that participated in a peer exchange that we did on GIS and highway safety. And we've got a summary report on our website.

The website, if you look on the lower left corner of your screen, is gis.fhwa.dot.gov. And if you go to there, and if you click on Resources, and then click on Reports, you'll scroll down, you should see a link to a report that looks at what several states have done-- Massachusetts, Washington, Ohio, Maine, Tennessee, Illinois, and Minnesota.

Sounds like it might be worthwhile to kind of do a follow up case study report or peer exchange and see where states are nowadays. And I think I'd definitely include the work that you're doing, Sharon. Let me just check to see if we've got any other questions on the phone. OK. Well, we don't have any other questions on the phone line right now. Let me-- I'm just going to click it back to your contact info, Sharon. Were there any-- I don't know, last--

SHARON HAWKINS: Well, I know that the PowerPoint is going to be available to download from the participants. So I just wanted to let everybody know I had a lot of graphics in here, but at the bottom of each slide in the PowerPoint, I wrote down very particular notes about what each slide is showing--

MARK: Mm-hmm.

SHARON HAWKINS: --if they'd like to download that.

MARK: Yeah, actually--

SHARON HAWKINS: And then, we're--

MARK: Go ahead.

SHARON HAWKINS: I was just going to say, we're available anytime to answer questions. Once again, I don't think it matters what GIS platform you have. This tool was very-- it's a very easy methodology to put together. But if anybody has any questions, please feel free to contact us.

MARK: To download the file, you should be able to see a file download window. If you click on the file name and then click on the Save to my Computer button, you should be able to download a copy of this presentation. It looks like people have started answering some of our poll questions. And again, this helps us give some ideas of what kinds of webcasts everyone would be interested in participating in.

Also, if you've got ideas for a case study report, or a peer exchange, or something that we need to highlight in a better fashion, just let me know. My email address is on the lower left corner. I want to thank Sharon Hawkins again for agreeing to do this presentation. Like she mentioned-- I guess, Sharon, you did this presentation back at GIST? Is that correct?

SHARON HAWKINS: Yes, sir. Yes.

MARK: Yeah. So there were several-- lot of good presentations that were done there. And that's one of the things that we try to do, is-- especially for people who weren't able to make it out to this year's symposium, which was in Boise, Idaho. We try to get some of those presentations and make them available as webcasts. Like we've mentioned earlier, the presentation's going to be available to download. We're also recording this, so we'll send a link out to a recording of this presentation.

Let me see. One last check for questions on the phone. OK. Well, I think we're all set, then. Sharon's contact info is on the screen. Thank you, Sharon, again. And thank you to everyone for calling in, logging in, and participating. Please feel free to fill in the questions, answer the questions as best you can. And I want to wish everyone a good day. Thank you very much.

SHARON HAWKINS: Thank you, Mark.

MARK: Thank you, Sharon.