New York City Crashes

Helping the mayors office and other stakeholders to derive insights from Motor Vehicle Collisions/Crashes

Goal

To provide an interactive means for the mayors office and other stakeholders from New York City (NYC) to easily access motor vehicle collision information in an interactive manner thereby aiding them to make informed decisions.

System Description

Stakeholders

• Mayors office
• Ordinary citizens (secondary stakeholder)

Main Systems of Interest

• A Power BI Dashboard that allows analysts in the mayors office to visualize number of crashes by time period and location.
• In addition our Power BI Dashbaord should provide some statistical information about crashes in New York City.
• Finally our Power BI Dashbaord will show the closest police station as well as closeset hospitals to crashes.

Enabling Systems

Systems that will enable us build our primary system are automated cron jobs or schedulers that will fetch, clean and store data from the sources listed below

Data Sources

• Motor Vehicle Collisions: https://data.cityofnewyork.us/Public-Safety/Motor-Vehicle-Collisions-Crashes/h9gi-nx95
• List of hospitals as of 2011 https://data.cityofnewyork.us/Health/NYC-Health-Hospitals-Facilities-2011/ymhw-9cz9
• List of police departments https://www1.nyc.gov/site/nypd/bureaus/patrol/precincts-landing.page

ETL

• Azure Data Factory

Storage

• Azure Data Lake Storage

Dashboards/Reports

• Azure Power BI
• Azure Synapse Analytics

System Boundaries

Roles and Responsibilities

• Automated Software i.e Background Processes
• Data Owners:
  • Police Department (NYPD): Motor Vehicle Collisions
  • NYC Health + Hospitals: List of hospitals
  • Department of City Planning (DCP): List of police stations
• Data Analysts in Mayors office
• Developers/System Administrators
• General Public

Functionalities of System

• Easily query data about crashes with little to no technical knowlegde
• Easily create and share informative visualizations with people
• Access to basic informative visualizations:
  • Visualize places where accidents happened for last several days
  • Visualize nearest hospitals to the places of accidents
  • Visualize nearest police departments to the places of accidents
• Provide statistics and predictions:
  • Get information about what are the districts / crossroads where accidents occured most frequently for the day before / last week / last month
  • Predict districts where an accident is more likely to occur the next day
  • Get information about how often does crashes take place on a specific street / crossroad
• Easily export data, visualizations or reports in some desired format

Software Requirements Specification

Extra Reading

https://towardsdatascience.com/new-to-data-visualization-start-with-new-york-city-107785f836ab?gi=54183ba7443f

https://cawemo.com/diagrams/0b163bee-a87b-4d0e-a0e3-72af2b8c2ec3--etl-and-other-processes?v=960,378,1

https://azure.microsoft.com/en-in/solutions/big-data/#products