Tag: crash data

Stats from the OECD: Comparing traffic injuries of the United States and Netherlands

For an article I’m writing for Architect’s Newspaper about the Chicago Forward CDOT Action Agenda, I wanted to know about traffic injuries and fatalities in the United States, but compared to the Netherlands and Denmark and other places with a Vision Zero campaign (to have 0 traffic deaths each year).

I already knew the OECD had a good statistics database and web application. With a few clicks, I can quickly get a table of traffic injuries (casualties) listing just the countries I want. I can easily select the years I want, too.

In one more click the web application will show a time animated bar chart. A feature I’d like to see added is dividing the figure (in this case traffic injuries) by the population. Check out the video to see what it looks like. The United States looks to be in terrible shape, but our country has several times more residents.

I had trouble downloading and opening the CSV file of the data table I created. The XLS file was damaged, also. The built-in Mac OS X Archive Utility app couldn’t open the .gz file, but I used The Unarchiver app successfully.

My calculations, based on data from OECD (national population and traffic fatalities), Illinois Department of Transportation (IDOT), and the American Community Survey:

Fatalities per 100,000 in 2009

  • United States: 11.02472
  • Denmark: 5.48969
  • Netherlands: 4.35561
  • Sweden: 3.84988
  • Chicago: 16.74891
  • United Kingdom: 3.83555

Chicago’s fatality rate per 100,000 citizens in 2009 was 16.75 (473 deaths on the roads). The fatality rate dropped in 2010: just 11.65 deaths per 100,000 residents (315 deaths on the roads; the population also decreased).

Updated September 28, 2012, to add the United Kingdom. 

Terminology debate: crash versus collision

The following is an email conversation between myself and Travis Wittwer, a cool guy in Portland, Oregon, whom I stayed with in April 2010. We’ve had similar conversations before about the language writers (mainly newspaper article authors) use when speaking about and describing situations where “people and their bicycles make contact with people and their cars” (yes, there’s an easier way to say that, read on).

Travis: Continue reading

Slicing the crash data into interesting visualizations

The Chicago Crash Browser as it looks now. This only exists on my laptop and no place else. I can’t put it online because it’s so inefficient it would kill the server. 

I presented my Chicago Crash Browser to attendees of an OpenGov Hack Night three weeks ago and gathered a lot of feedback and some interest from designers and programmers there.

We collaboratively came up with a new direction: instead of focusing on creating a huge web application that I proposed, we (anyone who wants to help) would start small with a website and a couple of crash data visualizations. The visualizations would serve two purposes:

  • attract attention to the project
  • start building a gallery of data-oriented graphics that describes the breadth and extent of the crash data

Continue reading

How high (and low) expectations can make traffic safer

I have low expectations of fellow Chicagoans who are moving their vehicles on the same roads I cycle on. I expect that every door will fling open in my path, causing me to be doored. I also expect to be cut off at any moment, and especially in certain places like at intersections (where the majority of crashes occur), bus stops, or in places with lots of parallel parking activity. Because of these expectations I feel that my journeys have been pretty safe. My low expectations cause me to ride slower, ride out of the door zone, and pay attention to everyone’s maneuvers.

This is another post inspired by Traffic: Why we drive the way we do (and what it says about us) by Tom Vanderbilt. From page 227 of “Traffic”, about expectations :

Max Hall, a physics teacher in Massachusetts who often rides his collection of classic Vespas and Lambrettas in Rome, says that he finds it safer to ride in Rome than in Boston. Not only are American drivers unfamiliar with scooters, he maintains, but they resent being passed by them: “In Rome car and truck divers ‘know’ they are expect not to make sudden moves in traffic for fear of surprising, and hurting, two-wheeler drivers. And two-wheeler drivers drive, by and large, expecting not to be cut off.”

The scooter drivers have high expectations, and it seems that they’re being met.

This all plays nicely with the “safety in numbers” theory about cycling: the more people who are riding bicycles, the more visible bicycling is, and the more aware a driver will be around people who are bicycling, and the more they will expect someone on a bicycle. Awareness means caution.

It’s difficult to gauge the safety of cycling in Chicago as we’ve no exposure rate: we don’t know how many people are cycling how many miles (nor where).

A cyclist waits for the light to change at Milwaukee Avenue and Ashland Avenue. 

Exposure rate

Exposure rate in the sense I’m using it here means the number of times someone is in a crash or injury for each mile they ride. We know how many crashes and injuries are reported each year (in the Illinois Motorist Crash reports), but we don’t know how many miles people ride (neither individually nor an estimated average).

There was a limited household survey of Cook County residents in 2008 from CMAP, called Travel Tracker, that collected trip distance information for all trips members of a household made on all trip modes – I haven’t looked into this yet.

It would be highly useful if the Chicago Department of Transportation conducted ridership counts at the 10 intersections with the highest crash rates. And if the 10 intersections changed the following year, the new intersections would just be added to the initial 10 to track the changes of the initial 10. This would be one step closer to being able to determine a “crash rate” for each intersection.

Crashes by bike or by foot at different intersections

While working on a private web application that I call Chicago Crash Browser, I added some code to show the share of pedestrian and pedalcyclist crashes. The site offers users (sorry I don’t have a web server that can make it public) a list of the “Top 10” intersections in terms of bike crash frequency (that’s bike+auto crash). You can click on the intersection and a list will populate showing all the pedestrian and pedalcyclist crashes there, sorted by date. At the bottom of the list is a simple sentence that tells what percentage pedestrian and pedalcyclists made up at that intersection.

I’m still developing ideas on how this information may be useful, and what it’s saying about the intersection or the people using it.

Let me tell you about a few:

Milwaukee Avenue and Ogden Avenue

I mentioned in my article Initial intersection crash analysis for Milwaukee Avenue that this intersection is the most bike crash-frequent.

23 crashes within 150 feet of the center, 2005-2010

82.61% bike crashes **

17.39% ped crashes.

Ashland Avenue and Division Street

28 crashes within 150 feet of the center, 2005-2010

46.43% bike crashes

53.57% ped crashes **

Milwaukee, North and Damen Avenues

46 crashes within 150 feet of the center, 2005-2010

39.13% bike crashes

60.87% ped crashes **

Halsted Street, Lincoln and Fullerton Avenues

38 crashes within 150 feet of the center, 2005-2010

42.11% bike crashes

57.89% ped crashes **

Montrose Avenue and Marine Drive (Lake Shore Drive ramps)

11 crashes within 150 feet of the center, 2005-2010

90.91% bike crashes **

9.09% ped crashes

Why do you think some intersections have more of one kind of crash than the other?

People walking at Milwaukee-North-Damen.

The Chicago Crash Browser can be made public if I have a host that offers the PostgreSQL database. Do you have one to offer?