Flight Risk

We began by reading each incident report like a short note from the past, asking what it claims went wrong and how often that same reason appears again. This postcard is relevant because it can help us see whether mechanical issues, weather, human factors, or unknown causes dominate the story.

This card reads like an evidence board, as every official cause label is counted and placed side by side so the report language becomes measurable. It is relevant because these categories are the dataset’s own explanation field, so comparing them directly shows what investigators most often record as the root reason.

Here, we wanted to know when a flight feels most vulnerable, so we compare takeoff, cruise, and landing in one view instead of treating all crashes as the same event. This is relevant to the topic because phase-of-flight context tells us where risk concentrates operationally, which is far more actionable than just total counts.

This postcard focuses on the machines themselves, grouping aircraft types to see which categories appear most frequently in the crash record. It is relevant because fleet mix and aircraft class can shape exposure, and this view helps separate whether patterns come from operating volume, design era, or reporting concentration.

This one feels more personal: it maps where survivors are found within aircraft zones and asks how position might relate to outcomes. It is relevant because it shifts the question from “how many crashes happened” to “what happened inside the cabin,” giving a human centered lens to survivability patterns.

For the final card, we looked up to the sky and asked how often storm systems, turbulence, or poor visibility are named in crash narratives. It is relevant because weather is one of the most intuitive external factors, and this view helps distinguish atmospheric pressure on flights from technical or human causes.