Data Analysis on Bike Collisions in NYC

Posted on Jul 29, 2018
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.

AT A GLANCE

Project Type

Data visualization with an interactive web application

Dataset

NYPD Motor Vehicle Collisions with 1.33M rows and 29 features, size: 290MB

Source

NYC Open Data

Language/Packages

R dplyr, ggplot2, shiny, shinythemes,  googleVis, ggmap, stringr, lubridate, leaflet

Procedure

  • Cleaning dataset
  • Exploratory data analysis and descriptive stats
  • Case study to explore safety of bicycle lanes
  • Deploy interactive web application

Key findings

  • Bicycle accidents increase every year since 2013, particularly during winter months
  • Number of daily accidents peaks during rush hours after work (6-8pm)
  • Driver inattention and failure to yield right-of-way are the main reasons for accidents, regardless of vehicle type
  • Case study revealed that painted bicycle lanes might provide less safety than no bicycle lanes

Outlook

Advanced ML models such as clustering or classification could provide deeper insights into the impact of painted versus protected bike lanes on the safety of New York's cyclists


BACKGROUND.

 A resilient and sustainable New York has become a tangible goal of the OneNYC initiative, originally released in 2007 to address the city's long-term challenges including its aging infrastructure. Every year, data shows an average of 4400 bicycle accidents are reported to the NYC Police Department.

Despite a significant decrease of accidents involving pedestrians, the rate of cyclist accidents remains unchanged. To help ensure a safe environment for everyone who is out and about on NYC's busy streets, I answered OneNYC's call to speak up and analyzed the motor vehicle collision dataset that is accessible on the NYC Open Data platform. The complete analysis can be found here. For the underlying code, please refer to my Github.

What are the major causes of cyclist accidents and what can we do to prevent further accidents from happening? Let's find out...

Data RESULTS.

1. What time of the year and day do cyclist accidents happen the most?

In order to identify the main causes of bicycle collisions, it is important to understand whether seasonal or daily trends occur.  After sorting the number of accidents by season, I looked at the relative increase in accidents between 2013 and 2017.

Most accidents happened during the summer months where the total number rose by 13% from 1494 to 1701 collisions. The numbers show a similar trend for fall, increasing from 1190 to 1388 (16 %) and spring, increasing from 882 to 1069 (21%) accidents. Although bike collisions happen less during the winter months (2013: 495 and 2017: 645), the relative increase is almost three times as high as in summer (30%).

Data Analysis on Bike Collisions in NYC

Is it possible, that the lack of proper bike lights causes elevated accident rates during the darker winter months? To answer this question, I displayed the cumulative number of accidents, grouped by season, over 24h to look whether accidents happen more often during darker hours of the day. The first peak appeared during the morning rush hour between 8-10am and increased until the after-work rush hour between 6-8pm. This pattern is consistent across all 4 seasons and does not suggest a dependency on the daylight.

Data Analysis on Bike Collisions in NYC

Together, the results suggest an increasing trend of bicycle accidents throughout the year but mostly during the winter months and in particular during the rush hours after-work.

2. What type of vehicles are mostly involved in cyclist accidents and what are the main causes?

The dataset allowed me to search for the vehicle type and cause that was recorded most frequently between 2012 and 2018. Since the driving patterns and incentives for being on the street can differ between a taxi driver, a delivery car, or just a passenger vehicle, I identified the main causes for each vehicle type, shown in the form of a Sankey diagram.

Following the links between the nodes of the diagram, it becomes clear that passenger vehicles are responsible for the majority of accidents (13%) due to inattention and distraction. The second most frequent cause is the failure to yield the right-of-way. While other vehicle types such as taxis are less frequently involved, the two major causes remain the same across the categories: distraction and failure to yield right-of-way.

Data Analysis on Bike Collisions in NYC

3. Bike lanes for safety? - A case study

Increasing the attention and the focus of drivers requires, unfortunately, more than just a cup of coffee. Ideally, bike lanes help bikers to move along safer by allocating side lanes and reminding motor vehicle drivers to watch out. So maybe NYC just needs more of them?

To get an idea whether bike lanes decrease the number of accidents, I divided the dataset into locations where accidents happened more than 10 times and locations where accidents happened less than 5 times to compare "hot spots" with "safer" spots in regard to the presence of bike lanes. While I expected to see fewer bike lanes at highly frequent accidents locations and more bike lanes at less-frequent accident spots, the results suggested the opposite. Based on Google StreetView, 72% of the "hot spots" actually had bike lanes compared with only 44% of the "safer" locations.

Data Analysis on Bike Collisions in NYC

Does that mean that bike lanes are more dangerous for bikers? It depends. It's worth mentioning that the locations analyzed in this case study had painted bike lanes without providing any physical barrier between cyclists and motor vehicles. It is possible that painted bike lanes lead to a false sense of safety for the cyclist, while the lack of further protection in form of poles etc. does not prevent the worst from happening.

CONCLUSION.

Further research needs to be done to identify major contributing factors that cause the persistently high rate of bicycle accidents in NYC. Unsupervised machine learning provides the tools to cluster intersections based on multiple factors including the presence of bike lanes, which potentially will allow finding innovative solutions to make NYC's streets safer for not only cyclists but everyone. Stay tuned...

 

About Author

Melanie Uhde

I’m a Ph.D. in Immunology with hands-on experience in applying Machine Learning tools in Python and R. At Columbia University Medical Center, I led research projects on the identification of patient subsets to find new treatment options by...
View all posts by Melanie Uhde >

Leave a Comment

No comments found.

View Posts by Categories


Our Recent Popular Posts


View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI