Oxford University's Research of Dangerous Roads Using Webfleet Data

By Raghunath Banerjee, VP Data Solutions & Innovation at Bridgestone Mobility Solutions

In 2023, I published a LinkedIn article outlining how vehicle data is solving our road safety problem. Almost one year later to the day, here’s a check in of sorts on where we are now.

In that article, I described data as “the new oil” and “the engine powering the positive developments on our roads”. Those words are just as, if not more, relevant today – and not only because our road safety problem is still a major one.

Safe to say, the scope of data’s impact on road safety – both today and tomorrow – has grown a lot in 12 months.

‘Identifying high-risk roadways before massive crashes occur’

Last year, I talked about how we were leveraging connected vehicle data to help identify and map road damage, using AI-driven models to determine their severity. I talked about our work in the city Alkmaar, the Netherlands and the importance of cross-sector collaboration and real-time data insights to building smart cities of the future.

That work is focused on improving road safety in the here and now by understanding and addressing unsafe road infrastructure. This is, however, one component of the road safety beast.

The impact driver behaviour has on road safety is a far bigger unknown, but arguably the biggest issue we face. And it’s another area where data can illuminate the previously uncharted and power crucial research.

We’re now collaborating with the Transport Studies Unit (TSU) at the University of Oxford, working with researchers to leverage Webfleet 's telematics data and explore the connection between driving behaviours and crash risks within Oxfordshire in the UK.

As senior researcher, Dr Xiao Li , tells us “a better understanding of when and where risky driver behaviours, such as speeding and hard braking, occur can also help to reduce the prevalence and severity of road accidents”.

The goal of our collaboration is to enhance the timeliness and effectiveness of road safety assessments, and move away from relying on collision data alone – and its flaws – to understand dangerous roads.

“Traditional road safety studies are primarily conducted using historical collision data, which is reactive and only becomes valid after a certain number of crashes have been recorded. Relying solely on collision data also underestimates risks because it fails to capture near misses,” says Xiao.

“With the telematics data provided by Webfleet, researchers from TSU

examined the relationships between abnormal speed variations and crash risks in Oxfordshire. They proposed a new data-driven approach to identify dangerous roadways based on this telemetric data. This study enables the identification of high-risk roadways before massive crashes occur.”

Xiao and his team of researchers at Oxford’s TSU found that the abnormal speed variations generated from just one month of data showed statistically significant relationships with five years of historical crashes. Their work has already demonstrated huge potential for enabling proactive road safety assessments.

Whether it’s working with the public sector to map road degradation today, or powering research teams to locate the biggest road risks and identify the right changes for tomorrow, vehicle data is delivering profound insights in all corners of road safety.

When I share another update in 12 months’ time, there will surely be a new corner altogether that we’re exploring and breaking new ground in.

To learn more about data solutions from Bridgestone Mobility Solutions, please visit https://datasolutions.bridgestone-emia.com.