In this webinar we cover how you can use Urban SDK to aid in faster safety planning. Pull traffic speed data from Insights, create custom maps with the data and your own crash data to pinpoint safety hot spots.

Hey everyone, welcome to another webinar. We're excited to have you. As always, my name is Ashley Thomas.

I'm the Director of Customer Success over here at Urban SDK. We're very excited for today's webinar. We're going to be going over safety analysis and hot spot identification.

And as always, if you've missed any of our previous webinars, please let your Urban SDK representative know. We can get you that recording. Those are also hosted on our Help Center.

If you go to our Help Center, you can see our whole repository, but just reach out to one of us and we will get you those recordings. I'm going to go ahead and toss it over to Andrew to introduce himself and kick it off. Thanks so much, Ashley.

So hello again, everyone. If you've been following our webinar series, you may recognize me. I'm Andrew Larter.

I'm Senior Transportation Planner here at Urban SDK. I was formerly a customer of Urban SDK myself in my previous life as a consultant transportation engineer. Now I'm responsible for a lot of the data QA, QC, and product development here at Urban SDK.

And so on today's webinar, what we want to do is we want to talk about safety analysis, hot spot identification. And these are kind of topics that dovetail very well from last week's webinar, where we analyzed speeds in a neighborhood that had recently laid down speed bumps following a tragic incident. Now, in that webinar, we analyzed pre and post implementation, see how speeds have changed, basically trying to analyze the impact of the intervention that had been made on the road network.

But in this training, we're going to be focusing a little bit more on safety from a macroscopic scale. And we're going to be diving into a few more of the tools that are available within your Urban SDK platform. Now, if you want to know how to use Urban SDK speed data to identify these hot spots on your network, or if you're a community that's considering applying for an SS4A grant or has money to develop a safety action plan, this could be a useful demonstration for you.

It also comes in handy if you work closely with local law enforcement on enforcement of speed and safety problems. Since you have speeds on hand through your Urban SDK subscription, and because you can easily upload crash data into your account, it can become much quicker to run the safety diagnostics for a road or a corridor or segment or even from a system wide perspective. So really what we're looking to demonstrate today is how you can layer the information that you already have on top of your data from Urban SDK to get a better view of the safety concerns on your roadways.

And the intent, just like with many of our previous webinars, is to help you really hone in on areas that need the most help on your network and cut out the noise that can often come with looking at these networks from a large scale perspective. So we're going to dive right in. As always, we're beginning here on the Urban SDK splash screen, I put in my credentials.

So let's log in. I remembered my password today, it's a good thing. Immediately when we jump in, we start on the personal workspace.

So we've talked about this in previous webinars. But just a reminder, the personal workspace is a sandbox environment, you can see at the top left here, where the star is to sandbox environment where you can play around with the different features in your platform. And none of the reports or the data that you save in your personal workspace will be shared or can be shared rather, to other members of your organization or members of the public.

So if we want to move into the workspace where you can share data among members of your organization or publish it to the public, we want to click on the personal workspace tab and move down to urban SDK. Well, that's what it will show for me because I'm looking at our internal workspace, but it should have your organization's name as your workspace on the side. You click that workspace, and it'll take you where you need to go.

So what we're going to do first is we're going to take urban SDK speed data from our insights tool. And then we're going to take crash data, which in this case has been produced or provided to us by one of our customers, obviously sourced from ultimately local law enforcement who attends to crashes and collisions as they happen on the roadway. So we're going to layer the data that we had urban SDK provide with the data that you may already have on hand, as in terms of speed and crash data, and put them together into one map to identify hotspots and safety concerns on a transportation network.

So first, we're going to get the speed data, we go to the left side of our control panel, and go to the insights beta. And today, not to pick on anyone, of course, we try to be fair with our geographic areas. But today, we're going to be going back to sunny New Mexico in Santa Fe.

And I'm going to load the entire county of speed data. So that's Santa Fe County, of course, speed data. And we're going to go to June 2023.

I already have it selected. Perfect. So we're going to look at data from this past year.

So as in the previous webinars, you may notice that we have two additional supplemental months that are loaded immediately. That's the April and May 2023 data that just enables a three month comparison of speed data within insights 2.0 itself. So with all this being said, we're going to click build report.

And we'll see how quickly the data updates, which is of course dependent on my home internet, which is doing pretty well today. So it's another positive. Close the side panels so we can get a better view.

So what we're looking at here is Santa Fe County. Now, before in previous webinars, we would zoom in on our area of interest directly from this screen or directly from this area. What we want to do today is take this data that we have here and save it to our data, which is where all data sets that you upload, whether it's from our insights tool, or whether it's from data that you yourself have uploaded, they all live in the data hub, which is essentially your system of record for your transportation data sets.

So rather than zooming in on an area of interest right here, we're going to just take the speed data that we've immediately pulled, and we're going to save it so we can then layer it on top of our collision data in our Studio GIS mapping tool. So what I'm going to do is I'm going to save this data to our data. So click the three dot option menu at the top left and export to data hub.

And I'm going to call this Santa Fe Speed Data. Now, to save on time in our, I clicked the 85th percentile speed, of course, I can actually add a tag there that just says Santa Fe. Now, to save on time in our webinar today, rather than saving this data set and waiting for it to process, because when I click when I click the export data button, what we're going to see is it's going to take a moment to export and save and process to our data hub.

So to avoid that problem, what we're going to do is I've actually previously saved speed data for Santa Fe already, as well as collision data for Santa Fe in our data hub on my account. So at this point, what you would do if you wanted to do this analysis would be click export data. And then you go into Studio and load it into Studio.

I've already saved it for quicker access. So we're going to go straight over to Studio so that we can layer this data on top of some collision data. So I'm going to click that.

And on the left side panel here, we're going to click the Studio button. And obviously, that will leave our unsaved report and insights. And then immediately we're met with Studio.

Now, what Studio is, is it's kind of like a GIS mapping tool. If you're familiar with ArcGIS or with QGIS, this should look pretty familiar. On the left, we have a map layers panel, which is kind of like your table of contents.

On the top right, we have a legend panel. And what we're going to do is we're going to add some data to the Studio. So I'm going to click on these five dots or these five blue squares.

I'm going to click the Add Data button. So we're met with three different options. Import file is if we wanted to upload a geospatial data file from my personal computer here, from a local drive.

So we can accept georeferenced CSVs. We can accept Esri shape files, zipped files of all sorts, and GeoJSON files. Then you can also do the same if you have a URL link to data that's stored on a server somewhere over here.

But we've already saved our speed data. And I've previously saved our collision data into our data hub. So what I'm going to do is I'm going to click Import from Data Hub, continue.

And we have a lot of data sets here for the different projects that we at Urban SDK are completing with our customers. So I'm going to filter down just to ZFA. So I have the speeds data that I wanted to look at, as well as the collision data that I previously uploaded.

So I'm going to click both of these. And I'm going to give it a few moments for the data to load into the browser. Again, we're working on a browser-based solution here.

So it depends highly on my home internet, which, again, is doing pretty well today. So now that those two are ready, we're going to continue. We're going to continue over here.

And immediately as we import, we have the Import Data Wizard here. This is just going to give us an idea of the format of our data. You can see the attribute table essentially here.

So this is the speed data, which we were looking at. And then if I click collisions, we're going to see the same for the collision data set that we're going to be layering over top today. So I'm going to click Continue again.

And then I'm going to click Import Data. Now, that loaded the collision data pretty quickly. If we're going to give it a second to load the speed data, because the speed data is a much larger file than the collision data.

There we go. OK, so we have both data sets here and ready to use. And I think you might have seen a note that said at the top, wait or end page.

Obviously, we're just loading a fairly large data set into our browser. So sometimes our browser hangs there for a moment. Just wait, and ultimately, we have our data ready to use.

So continue to close the side menu. And then what we have here is we immediately have two different data sets here in Studio. So not really very informative with what we're doing right now.

I mean, we have collisions, which you can see if I go to the left here and I click the little I button to turn off the speed layer, we can see we have point data for collisions. And if I turn off the collisions, you can see we have our line stream data for speeds. So there's a lot of different things we can do in this platform.

Again, Studio is very much a GIS platform, just like ArcGIS or QGIS. And it has features like filtering and symbology. And we're going to make use of some of those features in a little bit more in-depth of a fashion than we did in our previous webinar to really get to the heart of the issue here.

So first things first, we're going to want to take our line stream data for speeds, and we're going to want to color them by speeds themselves, obviously. So I click the little drop down menu here, which opens all the different options for speed data, or for symbology, rather. So we can see we're working with Santa Fe speeds.

We can see that the geometry is being referenced correctly. We're going to change the stroke color of this data so we can grade it on a color ramp. So I'm going to click the three dots.

And then you can obviously just pick any stroke color if you want all the lines to be colored in the same manner. But what we want to do today is to color strokes based on speed. So what I'm going to do is I'm going to click on maybe 85th percentile speeds on a Tuesday.

In traffic engineering, typically Tuesday and Thursday are considered to be your typical days of the week. Monday and Friday have effects that sometimes can be due to long weekends, people taking days off from work. Tuesday and Thursdays tend to be your most typical days of the week.

So we're going to click based on that. Now it's automatically assigned a certain color gradient to the speed data. I want to change this to more of like a green to red to represent increasing severity higher speeds.

So I'll click here, and I'll scroll down in all the different color ramp options until I find, well, this is a red to green color ramp. But what I can do, is I can click reverse, and it reverses the direction of the color palette. And now, as you'd expect, major roads and freeways, more red, higher speeds.

Pretty simple. What I'm also going to do here is increase the stroke width just a little bit. So color's a little more vibrant now from a system-wide perspective.

What I'm then going to do is I'm going to close my Santa Fe speed data and take a look at the other layer that we have here, which is our collision data. So I'll put up the layer settings there. Now there's a lot of different things we can do with our collision data here.

Obviously, there's a lot of different aspects that go into vehicle collisions. There's levels of severity, such as property damage only, injuries, minor or major injuries, and of course, tragically, fatalities based on collisions. Another interesting and often important aspect to consider about collisions is whether or not there have been pedestrians or cyclists involved in the collisions, so to speak vulnerable road users, or if there's collisions between multiple automobiles or vehicles.

So these are all fields that are within our collision data set. And I can, if I turn off the speed data so we can see the collision data a little bit better, I can actually color the point data here that represents our collisions according to some of those values. So because this is point data rather than the stroke, we're going to be looking at the fill in the center of the point.

I can color it based on something like if I scroll down past all these different fields, I can go to this is the string, I've reviewed the data set in advance, this is the string for crash severity. So I'm going to, you can see it's kind of small here, it's kind of difficult to read. So I'm actually going to increase the radius of these dots, I'm going to make them quite a bit larger, put them to 50.

Now this is not a particularly instructive color ramp. Part of that is because there's only three values in this particular field, property damage, injury, or fatality. So I might want to choose a color ramp that has only three steps instead.

So then we're getting a little bit more diversity in our colors here. If I highlight, I can immediately see different data sets that are available here, excuse me, the different points of data that are available. But you may notice when I highlight the data here, we're not actually seeing that crash severity field.

So what I might want to do is I want to go to the interactions tab here, which highlights or gives me the fields that when I hover over a piece of data, these are the fields that show up when I hover over the piece of data. This is called the tool tip, the little table that shows when I hover over a piece of data. So if I want severity, and I don't care so much about the other pieces of information, I can actually remove the other pieces of data here.

In fact, I can just click clear all, and I can select crash severity field. So now when I hover over a piece of data, we may just see injury crash. So this light red is our injury crash, dark red is our fatality, and the light green is our property damage.

Makes sense, color ramp that goes in increasing order severity. But what we're really looking to do here in our safety analysis and hotspot identification is layer these two different pieces of data on top of each other and understand what does this mean for our transportation network. So what I'm actually going to do is I'm going to use the filter feature within Studio to really filter down to an area of interest or an area where there's a problem on the transportation network.

So I'm going to turn the speeds back on, and immediately looks like we're really met with a lot of data. So don't worry, we're going to filter some of this out in just a second. First things first, though, we're actually going to change the order of drawing of the layers.

You can see speeds is above collisions. So if I click, hover over these dots here, and I just drag it the other direction, now speeds are drawn underneath the collision point data layer. This makes sense if we want to really look at how points overlay over top of a line, we want the points to be on top.

And actually see you can actually see there's some transparency in the point data. So I'm actually going to change the opacity of this data, which I can do by going back to the fill tab from 80% to 100%. Now they're completely opaque.

So I'm going to zoom out here. And actually, one more thing I'm going to do is I'm actually going to go back and change. So change the fill color so that we're no longer based on the fill color on a field.

And all the data points for collisions have the same color. And I'll show you why in just a second. Instead of this brown color that automatically assigned, I'm going to click.

Now, what we're going to do here is we're going to take a look at the transportation network of Santa Fe. And we're going to filter down to areas where there are problems, basically. We're going to see what kind of conclusions we can draw about that from the data we have here.

So if I go up to the Filter tab, which looks sort of like a sieve here, I can actually add a filter for either of the two data sets or for both data sets that we have available. So the first thing that I'm going to do is I'm going to filter our collisions down to only fatal and injury collisions, not collisions where there's property damage, where the highest level of severity exceeds property damage. I'm going to select the field.

We'll go back to severity. And the values I want, I want only fatal and injury. So as you saw, when I delete the injury, we only have a smaller number of fatal collisions, which makes sense.

But the very first thing that we notice, there's a very high concentration of these fatal injuries in this stretch of road sort of to the center of the city here. If I add injuries back in, we see that that same stretch of road is pretty dense with these white dots. So this is telling me something here.

There could be a pattern that there's a greater number of collisions, and specifically a greater number of collisions with severe injuries or fatalities on this one stretch of road. If I want to focus instead of on severity, if I want to focus on maybe vulnerable road users, we can change from looking at severity, we can change instead to a different filter, such as the pedestrian involved filter, where I can choose involved or not involved to determine if there's a pedestrian involved in the collision. So when I change to pedestrian involved only, obviously, we have a bit of a higher density in the downtown area where there's a lot of different pedestrian oriented activities.

But once again, this sort of diagonal road here seems to have a pretty high density. What if we change to cyclists instead? Well, in this particular data set, that's the pedal cycle field. We change to involved again.

Once again, we're seeing this diagonal road as a particularly high density. So we're noticing there might be an issue here. One other aspect that we're going to look at instead of just looking at collision data is we're also going to filter down our speed data.

Firstly, I'm going to increase, I'm going to go back, I'm going to increase the size of our, the width rather, of our speed data once more, maybe from one to two, so we can really sort of see the pop. Or you know what, maybe even all the way up to 23, that's a three. So this is kind of all turning into a bit of a color, colorful mess here.

So we're going to filter down our speed data. So I'll add a filter. But this time, instead of looking at collisions, we're going to look at speeds.

And we were using 85th percentile speeds on Tuesday. So since this is a numerical field instead of string field, rather than that discrete list of options, we now have this kind of continuous distribution of speeds. And I can either play with the sliders, and you can see dynamically, the speeds are sort of updating as I filter out lower speed roads.

Or I can specify a certain limit basically to what I want to filter. So we're actually going to filter out higher speed roads like freeways, basically. So let's filter out anything above 55 miles per hour.

Or how about even 50 miles per hour. And then we're also going to filter out our more local roads at the bottom to 25. So we still have some local roads here.

We could maybe even bring it to 30. And then I'm going to go back and quickly toggle off the collision layer so we can just see the speeds data. So this road that we were looking at, this diagonal east west road, this is Cerrios Road in Santa Fe, New Mexico.

We can see it's actually not a very, not to say it's not a high speed road compared to some of the other arterial roads in the area. But we can see that it's pretty firmly distributed across the maybe 30 to 50 miles per hour range. And this doesn't sound like a lot in 85th percentile speed, but the likelihood of a pedestrian experiencing a fatal injury when they collide with a car traveling between 30 miles per hour and 50 miles per hour, the likelihood of fatality jumps pretty drastically.

It's a non-linear correlation between likelihood of fatality and vehicle speed. And when we go from 30 miles per hour to 50 miles per hour, that likelihood drop goes from about 50% chance that a pedestrian will experience a fatal injury at 30 miles per hour to almost a certainty, 100% chance at 50 miles per hour. But we layer our collisions back on.

The conclusion that I really like to draw from this data here and from our hotspot identification is that there's a safety hotspot along Cerrios Road, particularly in this diagonal section kind of in the center of our screen. Now from local knowledge, this is actually a project that we at Urban SDK have been working with our customers in Santa Fe on before. Cerrios Road, if you were to visit Santa Fe, which I've never had the pleasure of doing unfortunately, it's a six to seven lane wide arterial road.

It does have sidewalks and painted cycling lanes, but one of the key features is that there's a lot of businesses, service businesses, stores, restaurants, gas stations, that kind of thing on either side of the road, which means there's very, very frequent driveways to access all these businesses that are accessing this very wide road that has a very significant traffic volume and also has frequent traffic signals. So what we're seeing is a lot of turning conflicts and a lot of points of interaction between traffic and pedestrians, but that's traffic that's coming off of a pretty moderately high speed road, particularly a very vehicle oriented road where there's a lot of traffic and drivers may not be immediately thinking, I need to look for pedestrians. So my conclusion as a practitioner here is that we've been able to use the speed data from Urban SDK and collision data provided by the Santa Fe MPO to ultimately determine there is a safety or hotspot issue here on Surreo's RIP.

Now at this point, you'd be able to hone in on this is an issue. You would probably at this point then send out field surveyors to see how traffic operations work in action, whether there's a lot of near misses happening on the street. But at this point, we've been able to use the speed data and the collision data that we have here to really hone in from a network wide perspective on an area where there could be a hotspot or safety concerns.

So I'm going to, before we turn it over to Q&A, I'm just going to really quickly run through how we get back to get speed data into Studio. Delete our map here. And just to recall, we go to the left control panel here, go to insights, choose our geographic area.

I'll do your census tract if you know a specific area, but we're looking at a macro perspective. So we're going to choose Santa Fe County, traffic speed analysis, take our time period of in question, and build our report. And then once the speed report builds, we click options and export to Data Hub.

Once we've exported to Data Hub, we then go to Studio, add data, and import from Data Hub. So that's, long and short, how we can use the different tools within your Urban SDK subscription to conduct a bit of a high level safety analysis and identify safety hotspot. So I hope that you've been interested in, or you've been enjoying, rather, this webinar today, and I'd be happy to answer any questions.

Thanks, Andrew. We do have a couple of questions, and if you guys have any other additional questions, please go ahead and pop those in the chat. But for right now, we do have just a few that have come in.

The first being, where is the speed data coming from? The speed data is coming from, that's our Urban SDK data. It's ultimately sourced from here connected vehicles. So it obtains a sample of speeds on the roads, and we are applying a gap fill methodology as well to fill in where there's not a sufficient sample size to really get a good understanding.

But that's the speed data. The collision data, maybe that's another question, I'm not sure, but the collision data is most typically provided by customers. It's often something that is collected by the local law enforcement, or even by state level law enforcement, and that's something that can then be loaded into the Data Hub from outside sources.

Fantastic, you read my mind. That was a second question there. There you go, two in one.

And then last but not least, our third question that we have is, are we able to share these maps that you just created with people who do not have an Urban SDK license? Yeah, absolutely can. So I unfortunately, I went and I exited out of the beautiful studio map we were creating before. But even if we just want to create a blank studio map here, just for demonstration purposes, we can click the save button in the bottom left here.

And we'll just say, test webinar studio. Just for demonstration purposes, when we save a map, we give it a category. Let's just go down and say speed.

When we save the map, it'll save to our workspace, our organizational workspace, which then means that anybody within our workspace is able to access it. But if we want to share it to the public, once we save the map, this little arrow is going to appear next to the save button, and click publish report, and then turn the toggle on. Now the report is public, the map rather is public.

And if you just copy this link here, you can then send this link to anyone that whether they have an Urban SDK subscription or not, and they'll be able to view the map, they won't be able to edit the data. But they'll be able to view the map. So this can be useful for presenting to decision makers, to members of the public, council meetings, all that kind of thing.

Fantastic. We did have another question come in. The question is, is the speed data just the 85th percentile, or does it compare to the posted speed? And the example that was given was, if the posted speed limit is 50 miles per hour, and we're driving 50 miles per hour, we're not speeding.

But Andrew, can you explain how that would work in the platform for us? Of course, yeah. Now we have within the speed data itself, we have what we call Urban SDK speed category, which is not a great ground truth source for speed limit data nationwide. So it's a modeled set of speed limit data.

Now with certain customers, I mean, perhaps some of the customers that are listening in today, we've worked with you to obtain like ground truth speed limit data within your community and upload it into the platform instead. So instead of our modeled speed category, which we're pretty confident in the results based on things like functional class and road width, instead of that, we've been able to actually have ground truth speed limit data for certain customers. If you have that data available and you'd like to update instead of speed category to speed limits, you can do so by contacting your Urban SDK representative as well.

But a speed category or speed limit can be filtered as well as the actual average or 85th percentile speed. So those are all basically being able to do that kind of filtering is all possible within the filter tab. Fantastic.

Those are all the questions that we have for today. If you guys have additional questions, please feel free to reach out to myself, Andrew, or whoever your Urban SDK representative is. We're happy to help.

And then we are kind of back on a normal schedule after this webinar. So you guys got a bonus one this month. We will obviously be skipping next week for Thanksgiving.

We hope you all have a very safe and lovely holiday. And we'll be back the following week. I believe that is November 29th for our next webinar session.

So if you haven't signed up for it and you need the link, again, please reach out to your Urban SDK representative. But we'll also be sending out some emails coming up to give you that link as well. Thank you guys so much for joining us today.

And we look forward to talking to you in a couple weeks. Looking forward to it. Take care, everyone.