Team Collision

From Visual Analytics for Business Intelligence
Revision as of 01:43, 9 October 2016 by Tingzhi.lim.2013 (talk | contribs)
Jump to navigation Jump to search

Team Collision Logo.png

PROPOSAL

POSTER

APPLICATION

RESEARCH PAPER


Problem and Motivation

In United Kingdom (UK), road traffic accidents have resulted in 1,732 deaths in the year 2015, which is a 2% dip as compared to the year 2014. Despite the drop in the number of deaths, the casualties across all severities remained at an alarming figure of 186,209. As there is an increasing demand for the use of public roads, there is a strong need for us to prevent road traffic accidents and make the public roads as safe as possible. In order to prevent such accidents, it is therefore crucial to understand what are the different factors that contribute to road traffic accidents, and these understandings may then be used to prevent road traffic accidents from occurring.

Objective

The objective of the project is to:

  • Understand the demographics of drivers
    1. Distribution of age of drivers.
  • Understand the demographics of casualties
    1. Distribution of age of casualties.
    2. Distribution of severity of casualties.
    3. Distribution of type of casualties.
  • Examine the underlying factors which contributes to accidents. The following are some factors, but not limited to:
    1. Temporal patterns: Accident records based on time.
    2. Weather conditions: Which type of weather conditions would cause more accidents?
    3. Road conditions: Which type of road conditions would cause more accidents?
    4. Location: Which city has the most accidents?
  • Develop appropriate interactive visualisation to allow discovery of insights from multiple dimensions from the dataset.

Data

In this project, our team will be focusing on 2015 road safety data. The data is obtained from data.gov.uk (https://data.gov.uk/dataset/road-accidents-safety-data). It contains only personal injury accidents on public roads that are reported to the police and recorded using the UK STATS19 accident reporting form. It consists a total of 3 datasets that provide information about the accidents, the types of vehicles involved and the demographics of the casualties. Most of the data attributes are coded and re-coding would be done with the lookup tables provided by data.gov.uk.

The following data attributes are used in this project:

  • Accident Dataset [Accidents_2015.csv]
    1. Accident_Index [Accident No.]
    2. Longitude
    3. Latitude
    4. Day_of_Week
    5. Time
    6. Local_Authority_(District) [City Name]
    7. Weather_Conditions
    8. Road_Surface_Conditions
  • Vehicles Dataset [Vehicles_2015.csv]
    1. Accident_Index [Accident No.]
    2. Vehicle_Reference [Vehicle No.]
    3. Age_of_Driver
  • Casualties Dataset [Casualties_2015.csv]
    1. Accident_Index [Accident No.]
    2. Vehicle_Reference [Vehicle No.]
    3. Casualty_Reference [Casualty No.]
    4. Casualty_Class [Driver/Rider, Passenger or Pedestrian]
    5. Age_of_Casualty
    6. Casualty_Severity

Research Visualisation

Visualisations Comments
TeamCollision ResearchViz 1.JPG
Interactive Visualisation to Track Fatal Accidents
(http://news.bbc.co.uk/2/hi/in_depth/uk/2009/crash/8414354.stm)
  • The pie-chart on the left allows you to select the desired category of data to display.
  • The radial bar chart will then allow you to look at the distinctive pattern of each age group over a time-period.
  • The radial bar chart will be beneficial for high number of bins, where we will be able to look at all the bars or columns from one view without scrolling back and forth.
TeamCollision ResearchViz 2.JPG
Interactive Visualization to Rush Hour Danger
(http://www.bbc.co.uk/news/uk-15975564)
  • Visualization of the pattern of pedestrian casualties across the week
  • Map which shows the deaths involving pedestrians and buses on London’s world-famous Oxford Street between 1999 and 2010
  • Although this visualization is interesting with the data points of casualties over the years, it lacks the interactivity that our team envision our storyboard to be. We will use this visualization as a reference when we implement an interactive map in our storyboard.

Visualisation Strategy

Team Collision Visualization Strategy.png
 We intend to use a top-down approach, where our visualisation is being segmented into 3 major portions.

From the top, the first visualisation (Top-Left) is a map of UK, where it allows users to focus on different cities of UK.

The second visualisation (Top-Right) serves as an intermediate navigation-step, where users can then focus on more in-depth details such as the different casualty types.

Lastly, at the bottom section shows the underlying factors. Through the first 2 visualisations, users will be able to study how the casualties are associated with the underlying factors.

Tools

The following tools are used in this project:

  • JMP Pro
  • Tableau 10
  • D3.js
  • Text Editor

Technical Challenges

The technical challenges that the team will potentially face is the usage of the following tools:

  • D3.js: As we do not have prior experience, and thus we might need get used to D3.js within the shortest time possible.

Milestone

References

Comments

Retrieved from ‘https://wiki.smu.edu.sg/1617t1IS428g1/index.php?title=Team_Collision&oldid=4096’