I'm thrilled to present the outcome of my data analytics project, centered around a Kaggle dataset accessible through this link. The dataset encompasses a comprehensive collection of 1.5 million accident records that occurred on UK roads spanning the years 2005 to 2014. The dataset contains accident details such as Road types, Speed limit on the Road, Road surface, Light condition and Weather condition.
Here are the pivotal takeaways gleaned from the analytics:
Contrary to my initial expectations, it came to light that a higher incidence of accidents happens on roads with a speed limit of 30mph compared to other limits, constituting 64% of the total recorded accidents. Strikingly, the most severe accidents were observed on roads with a 15mph limit.
Furthermore, the analysis revealed that Fridays witness a greater frequency of accidents and casualties in comparison to other days, while Wednesdays marked the occurrence of the most severe accidents. Sundays emerged as the day with the least reported accidents, casualties, and severity.
The dataset shows a consistent escalation in the number of accidents across quarters, with both the first and last quarters having the highest severity. While accidents and casualties exhibited a gradual decline from 2005 to 2011, there was a conspicuous surge in 2012. Intriguingly, even though 2012 had the fewest severe accidents, its overall accident count rose significantly.
Noteworthy is the fact that an overwhelming 73% of all accidents took place during daylight hours when street lights were on, whereas 20% occurred during nighttime when street lights were illuminated.
When dissecting the road types, a striking 75% of accidents happened on single carriage roads, whereas dual carriage roads and roundabouts constituted 15% and 7%, respectively.
These insights and more are vividly illustrated through the dashboard's visuals below:
Notes: The dataset does not contain information for 2008.
Cheers.
Akinloluwa.
Comments