Can Connected Cars Predict Accidents?

We have been working on this topic for quite some time and have made some initial progress on predictive analysis. Without going too deep into the weeds we wanted to showcase data from a few cities where we noticed some interesting driving behaviors.

Driving Patterns – Harsh Acceleration & Harsh Braking

At dashroad we measure driver behavior to help give drivers feedback on how they can become better drivers. When we look at the data in aggregate, we can see that this feedback can make drivers safer. In the charts below we look at driving in Saudi Arabia and Dubai. When we started looking at driving behavior in Jeddah, Saudi Arabia we noticed a lot of aggressive driving styles. Over time we were able to bring those numbers down from an average of 250 Harsh Braking (HB) events/week to less than 20 HB events/week. For Harsh Acceleration (HA) there was a similar reduction in driving pattern.

 


 

jeddah

 

In Dubai it was similar but we noticed an uptick and decided to dig a bit deeper. What we found out was that during Ramadan there was a rash of aggressive driving late into the evening. In the last chart you see that overspeeding was happening more from 8pm to 12am during Ramadan and during the non-Ramadan hours it was happening more uniformly from 3am to 3pm.

 

dubai

 

overspeeding


 

Can Driving Behavior Help Predict Accidents?

With this type of data we are now seeing ways to help reduce accidents by giving alerts to drivers before they get on the road. For example, we know the times of the day when more aggressive driving is occurring and we can use this data to educate the driver.  While we are early in the development of accident prevention techniques, we are looking at the correlation between these driving patterns and accidents.

 

dashroad and Safety

We are working with a few government bodies to look at accident prevention and will be sharing our results here in the future. Follow us on twitter for more info.