Dark Mode
UK Road Safety Incident Records
Public Safety & Security
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
Data contains accident, vehicle, and casualty records for road traffic incidents across the United Kingdom from 2016 through 2020. This resource, sourced from the UK's Department for Transport, is essential for analysts studying road safety, insurance risk modelling, and accident pattern recognition. It allows users to join three key tables—Accident, Vehicle, and Casualty—using a shared index to perform detailed statistical analysis of road safety factors.
Columns
The primary accident file contains 36 columns, with detailed descriptions available in a separate provided lookup table. Key fields include:
- accident_index: A unique key used to link the Accident, Casualty, and Vehicle records together.
- accident_year: The year the incident occurred, covering the period from 2016 to 2020.
- accident_reference: A unique reference identifier for the police report.
- location_easting_osgr & location_northing_osgr: Ordnance Survey Grid Reference coordinates specifying the location of the accident. Easting values span from 64.1k to 655k, and Northing values range from 10.2k to 1.21m.
- longitude & latitude: Geographic coordinates, where the mean longitude is -1.26 and the mean latitude is 52.4.
- police_force: Identifies the police jurisdiction that recorded the incident, with codes ranging from 1 to 99.
- accident_severity: Classification of the accident impact, using values from 1 (most serious) to 3 (least serious). The majority of records (481,745) are classified as severity 3.
- number_of_vehicles: The count of vehicles involved in the incident, which can range from 1 to 24, with a mean count of 1.84.
Distribution
The material consists of three primary CSV files:
dft-road-casualty-statistics-accident-last-5-years.csv (which is 98.21 MB), dft-road-casualty-statistics-casualty-last-5-years.csv, and dft-road-casualty-statistics-vehicle-last-5-years.csv. The main accident file includes 598,000 valid records. A supplementary Excel lookup table is also included to interpret the coded variables, such as weather conditions, urban status, and specific pedestrian details. All records are 100% valid for the time period covered.Usage
This resource is ideal for conducting statistical analysis focused on road safety trends and predictive modelling. It is well-suited for performing geospatial analysis using the coordinate fields to pinpoint high-risk areas. Analysts can build clustering scripts and develop insurance risk models by studying factors like weather, location type, and the relationship between vehicle count and accident severity.
Coverage
The geographic scope includes road traffic accident data recorded across the entire United Kingdom. The time range captured is five years, from the beginning of 2016 through to the end of 2020. The scope includes details on the accident event itself, the vehicles involved, and casualty information. The data is static and is not scheduled to receive updates.
License
CC0: Public Domain
Who Can Use It
The dataset is intended for beginner and intermediate data scientists, researchers, academics, and professionals working in the insurance sector. It is highly suitable for anyone looking to apply statistical and geospatial techniques using standard tools like PostgreSQL and R (Tidyverse).
Dataset Name Suggestions
- UK Road Traffic Accident Data 2016–2020
- UK Road Safety Incident Records
- Department for Transport Casualty Statistics
Attributes
Original Data Source: UK Road Safety Incident Records
QUALITYLoading...