Linking national crash data to detailed vehicle characteristics

The RAC Foundation has been awarded a grant from the Road Safety Trust to explore the links between deprivation and vehicle technologies in road safety. Specifically, how differences in access to certain vehicle technologies based on income and other aspects of societal deprivation may have a bearing on road safety outcomes. 

To answer this question, the researchers were given two years of GB police crash data (STATS19) including vehicle registration mark (VRM, sometimes called numberplate) and vehicle identification number (VIN, the manufacturer-issued permanent identifier). This was then linked to government and commercial sources of VRM- and VIN-based data to build a database where cars in crashes had additional data fields on roadworthiness testing, usage, physical characteristics and available safety technologies for each vehicle. 

This presentation will explain the development of a repeatable methodology to (a) quality assure vehicle ID in crash data and (b) join to other datasets in a way that is both reliable and offers significant reductions on commercial data costs. 


Bruce Walton, Technical Director, Agilysis 

Bruce Walton is a skilled and insightful data professional with substantial experience of developing and applying analytical techniques.

His expertise in road safety intelligence was instrumental in creating the multi award-winning data tool MAST Online, and he has also collaborated on many other projects and reports.

He is also an experienced and accomplished conference speaker and technical trainer.

Dr Ivo Wengraf, RAC Foundation

Dr Ivo Wengraf is Research & Data Manager at the RAC Foundation, a leading independent transport policy and research organisation. With a PhD in transport geography, his main research interests are in the use of government and other administrative datasets to better inform national road transport policy discussions.

He is involved in a number of projects which join various official data sources to produce better insights, including two projects linking STATS19 to health and vehicle characteristics data, and another linking DVLA to DVSA data to produce a single dataset for academic research purposes. He is on the RSGB Champions group, he is part of the STATS19 R package authoring group, and he was a member of the last STATS19 review panel.