Analyzing Crash Data to Understand Transportation Safety
This study examines Chicago crash data from 2023, US Census Bureau demographics, and Advanced Transportation Feature data from Ecopia AI.
Between 2018 and 2022, fatal car accidents in the US increased by more than 16% . While 2023 numbers are still being calculated at the time of this publication, and initial findings seem to indicate a decrease, there is no denying that traffic safety has been declining nationwide in recent years.
According to the US Department of Transportation (DOT), over 350,000 people lost their lives in motor vehicle accidents between 2011 and 2020. In 2022 alone, there were 42,795 fatalities - the highest number recorded since 2005. These deaths account for 95% of all fatalities related to US transportation infrastructure , and disproportionately impact disadvantaged communities.
Traffic fatalities have been steadily increasing since 2011; source: US DOT .
Fatalities do not only include motorists; too often, pedestrians and cyclists are also killed or injured in traffic accidents. In 2021, 7,300 pedestrians and 966 cyclists suffered fatal injuries from motor vehicle crashes.
Motor vehicle crashes have overall been increasing across the country. US DOT data shows that in 2021 there were over 6.1 million police-reported traffic crashes , representing an 8.7% increase since 2012. Nearly 30% of these crashes resulted in injuries.
Given these alarming trends in traffic safety, it's no surprise that state DOTs and metropolitan planning organizations (MPOs) are devoting more and more resources to understanding - and preventing - crashes. Initiatives like Vision Zero are growing in popularity, while data and analytics are providing an unprecedented level of insight into traffic safety conditions nationwide.
Geospatial data plays a particularly important role in helping DOTs and MPOs not only understand traffic safety, but also design and implement strategic mitigation efforts to reduce traffic incidents. Layering multiple geospatial datasets together to derive these insights is critical, as many community characteristics contribute to transportation safety.
In this study, we layer 2023 crash data from the City of Chicago with demographic data from the US Census Bureau and Advanced Transportation Feature infrastructure data from Ecopia AI to understand where crashes occurred, what conditions contributed to those crashes, and which communities have the most traffic risk.
Where did crashes occur in 2023?
To start our analysis, let's look at where the 109,597 police-reported motor vehicle crashes occurred throughout the City of Chicago in 2023. Zoom in and click on a point for more information about that specific crash.
If we look at the breakdown of total crashes by census block group (CBG), we can better understand the geographic distribution of these crashes, and where they are most likely to occur.
As can be expected, more crashes occur in the downtown Chicago area, although there are also hot spots near O'Hare airport and along major roadways.
Click into a CBG to learn more about the types of crashes that occurred there in 2023, as well as who lives there and what demographic characteristics may contribute to their crash risk.
Now that we have a baseline understanding of where crashes occurred, we can start to see what types of crashes are more common in certain geographic areas.
140 fatal crashes happened in the City of Chicago in 2023. Plotting these crashes on the map, we can see the exact locations where these traffic incidents took place.
Visualized by CBG, we can see where fatal crashes are more likely to occur. The vast majority of CBGs in our study area experienced no fatal crashes in 2023, and at most, CBGs only experienced 2.
A few CBGs in downtown Chicago stand out as hotspots for fatal crashes, but so do a few other areas of the city, including the Pilsen and South Deering sections.
Plotting the 16,620 crashes that involved injuries, we can see that they are much more evenly dispersed throughout the city, much like the map of overall crashes. While these only make up about 15% of all crashes, this visualization suggests there is no clear geographic pattern.
Similarly, when visualizing by CBG, the map generally shows the same hotspots for crashes with injuries as overall crashes.
Throughout the year, 2,722 crashes occurred involving pedestrians. When mapping the crashes, it makes sense that most appear to be clustered in downtown Chicago.
A clear pattern can be seen when mapping pedestrian-involved crashes by CBG. These crashes are most likely to take place in the urban downtown section of the city.
If we overlay these pedestrian-involved crashes onto daytime population density, we can clearly see that these crashes are more likely to occur where there are higher concentrations of people during the day.
Mapping the 1,931 crashes involving cyclists, there is definitely a trend towards these crashes happening in downtown Chicago, but also a higher likelihood of these incidents occurring in the northern parts of the city as compared to the southern.
This pattern is confirmed when mapping by CBG, but this visualization method makes it clear that more some cyclist-involved crashes do occur in the northern part of the city, the vast majority take place in the downtown core.
How do demographic characteristics relate to crashes?
Layering census data with historical crashes provides deeper context about who lives and works in the areas experiencing crashes. This insight can be used to inform decision-making about where to focus resources aimed at enhancing traffic safety.
Understanding commuting beavior
We can also use census data to understand how people living in each CBG commute to work, and how that relates to crash incidents. Explore the map below to see how commuting behavior varies by location.
Predominant means of commuting to work by CBG.
The data shows that CBGs with "walking" cited as the predominant commute type have 2.5x more crashes on average. These CBGs experienced 1.5x more fatal crashes, 2x more crashes with injuries, 3.7x more pedestrian-involved crashes, and 3.6x more bike-involved crashes.
What role does race play in traffic safety?
When layering demographic data into our analysis, we can examine how race relates to the likelihood of traffic incidents. We first mapped CBGs by predominant race and quantified the crashes that occurred in each in 2023.
65% of crashes in 2023 took place in CBGs with a predominantly non-white population. On average, these CBGs each experienced 8.5 more crashes than predominantly white CBGs.
Crashes by CBG predominant race.
In fact, CBGs with a majority non-white population experienced more crashes involving fatalities and injuries.
These CBGs also had more crashes citing 'no traffic controls' as a contributing cause. In majority non-white CBGs, 39,512 crashes happened in locations with no traffic controls, while in majority white CBGs only 20,328 of these crashes occurred.
Likewise, crashes in majority non-white CBGs areas cited 'worn reflective material' on traffic signage as a contributing factor 2x more often.
Crashes caused by obstructed crosswalks also occurred 6x more often in majority non-white CBGs.
What role does income play in traffic safety?
We can do a similar analysis using median household income. First, we classify each CBG into income quintiles and then quantify the crashes that occurred in each in 2023.
The differences in crash occurrence between CBGs in the top and bottom 20% of median household income are less stark, but still significant. Over 31% of all crashes in 2023 took place in CBGs in the bottom 20% of income, compared to 19% in the top 20%. That amounts to a difference of more than 13,000 crashes per year.
Crashes by median household income quintile.
Diving deeper, lower income CBGs have 2x more fatal crashes on average than higher income CBGs.
However, crashes involving cyclists are more than 3x as common in higher income CBGs.
In lower income CBGs, 19,315 crashes happened in locations with no traffic controls, while in higher income CBGs only 10,472 of these crashes occurred.
Lower income CBGs note worn road surfaces as a contributing cause in 2.8 more crashes than higher income CBGs.
Similarly, lower income CBGs had 4x more crashes citing "worn reflective material" on traffic signage as a contributing factor.
Analyzing transportation infrastructure with historical crash data
These statistics show how traffic and pedestrian infrastructure play a key role in transportation safety. While some crashes occur due to driver behavior and weather, many are caused by inadequate or nonexistent safety infrastructure.
To better understand how crashes relate to existing infrastructure, more and more DOTs and MPOs are digitizing transportation features into vector data. This data can then be layered with historical crash and census data to dive deeper into where and why crashes are most likely to occur.
Leveraging our AI-based mapping systems, Ecopia recently digitized Advanced Transportation Features across the entire City of Chicago. Pan around the map below to explore how transportation infrastructure varies throughout the city.
Ecopia AI Advanced Transportation Feature data; inconsistencies may exist in this visualization as basemap imagery was not the imagery used for feature extraction.
We can then layer in 2023 crash data to better visualize and understand what pedestrian and traffic infrastructure exists near crash locations. This can help determine where to make improvements that will reduce the likelihood of future crashes in these areas.
Ecopia AI Advanced Transportation Feature data overlaid with 2023 historical crash data; inconsistencies may exist in this visualization as basemap imagery was not the imagery used for feature extraction.
Layering demographic, infrastructure, and historical crash data for decision-making
With these vector layers, we can also understand how infrastructure relates to the population.
Explore the map on the right to discover how sidewalk availability varies by CBG.
We can then overlay crashes involving pedestrians to see how these relate to the amount of sidewalks per CBG.
While pedestrian crashes are dispersed throughout the city, there seems to be a relationship between the number of these crashes and the amount of sidewalks in northeastern CBGs.
If we visualize the total amount of sidewalks per CBG by the number of crashes in that CBG involving pedestrians, we can see where a lack of sidewalks may be leading to more pedestrian-involved crashes (dark purple).
This confirms our observation that sidewalk coverage is related to pedestrian crashes in the northeastern part of the city.
Our earlier census data analysis highlighted that CBGs where walking to work is the most popular commuting method are more likely to have crashes, so we can also layer in historical crash data to find hot spots (dark purple) where pedestrians are most at risk for crashes due to lack of sidewalk access.
Again, the northeastern part of the city seems to be a good candidate for sidewalk infrastructure improvements to reduce the likelihood of pedestrian crashes.
We can do the same analysis for crosswalks. First, we visualize the amount of crosswalks in each CBG to understand where pedestrians lack safe infrastructure for crossing streets.
Next, we overlay crashes involving pedestrians.
We can again see that particularly in the northeastern part of the city, many pedestrian-involved crashes occurred in CBGs with fewer crosswalks.
When we visualize the crosswalk data by crashes involving pedestrians, we can see the same area highlighted (dark blue).
Adding in commuting data, we can begin to understand where the lack of crosswalks puts walking commuters at risk.
Layering demographic data with historical crashes and infrastructure vectors provides helpful insight into where improvements can be made to enhance transportation safety.
Digitize transportation infrastructure with AI for deeper analysis and insights
The above examples are just a few of the many safety analytics use cases unlocked with high-precision transportation vector layers. At Ecopia AI, we work with MPOs and DOTs across the US, as well as leading transportation authorities around the world , to create comprehensive, accurate, and up-to-date maps for strategic decision-making. This data opens up new opportunities for analysis and improvements in infrastructure not possible without a digital source of truth for current transportation networks. What's more, our AI-based mapping systems can produce GIS-professional quality maps across an entire city or state in just a matter of weeks without any resources spent on manual digitization.
Whether you're working on active transportation planning, multimodal network analysis, ADA compliance, green infrastructure creation, or another use case requiring high-precision geospatial data, let Ecopia AI scale your data creation so you can focus on what really matters: analyzing information and implementing positive changes.
