This data visualization project explores the changes to aviation safety over the past 25 years, particularly focusing on the steady improvements made over time.
Media and software: Adobe Illustrator, Adobe Photoshop, RAW Graphs, Tableau, Excel
Client: Shay Saharan
Creating this visualization began with the compilation of data. Because no comprehensive data set had been compiled for the statistics I wished to explore, I collected and compiled the data by hand.
I utilized multiple sources, including safety reports from the International Civil Aviation Organization and International Air Transport Association, Boeing and Airbus reports, and the Aviation Safety Network database to compile all of the data I required. I collected data on overall accident numbers, accident rates by number of flights, hull loss (total aircraft loss) rates, hull losses and accidents by model, and accident rates per category of accident, among many others.
Here is a look at my compiled Excel sheets.
I completed initial thumbnails and sketches to plan the arrangement of my data. Based on my look into style references, I knew I wanted to use engineering blueprints as a model for the visual design.
From there, I started to design a draft of the visualization. I used Tableau and RAW Graphs to begin making the graphs involved in the first and last sections, which I then edited and formatted in Illustrator.
For the final version of the visualization, I made significant improvements to the data in sections one and two (top and middle). I changed and improved the graphs in section one to make my argument—the improvement of aviation safety over time—more explicit.
Section two, focusing on the safety records of various types of aircraft, underwent the most changes. I experimented with many types of graphs to best suit the data I presented. The first strategy I tried was to graph production numbers against total fatalities, using the size of various data points to encode the number of hull losses per 500 aircraft produced. I produced this graph in Tableau and formatted it in Illustrator.
However, after working with the two graphs together, I began to feel like the first one was not as effective as I wanted it to be. I thought a timeline was much more important, given that I wanted to show changes over time, and making the timeline a secondary element was not the best choice.
I also reconsidered the needs of my audience, as I want this work to be accessible to the general public, and this visualization seemed too complex to suit the audience’s needs. Additionally, I considered the fact that the red and green color scheme was not accessible for viewers with red-green colorblindness. I ultimately decided to remove the first chart I made and instead modify the range chart for my final visualization.
I used Tableau to produce a graph with line weights that varied to encode the number of hull losses per 500 aircraft manufactured.
In Illustrator, I combined this graph with the data points from a Tableau bubble chart that encoded fatalities per aircraft produced. I added labels and plane icons to produce the final graph.
The third section of the visualization, focusing on the rates of different types of accidents, underwent only visual and design changes from the comprehensive draft.
In my final steps of completing the visualization, I decreased the overall situation to improve legibility and visual appeal. I then added handwritten post-it notes in Photoshop to provide further information and contribute to the blueprint design, finishing the visualization.