We analyze per-second GPS driving data to create a DriverDNA for each trip. We compare the DriverDNA against our database of over 7 billion patterns to rapidly determine crash probability and climate impact.
Pattern AI works in a similar way to facial recognition. But instead of comparing data to known faces, it compares it to a database of vehicle trips, and their outcomes. Because the AI is trained with damage data, saved CO2 emissions and fuel consumption, we’re able to analyze trips in real-time and predict actual outcomes.
Read our
blog post where we explain it further.
We continually train the AI with new data. And, although the core of the model isn’t influenced much by new data, our capabilities in predicting type and severity of crash are constantly evolving. This helps our customers to predict not only who will crash but also what type of crash they are likely to be involved in; for example, a rear-end collision, or minor scrape. This is powerful knowledge that will further strengthen risk mitigation and prove transformative across the insurance industry.
We look at real world trips and their outcomes, whether they involved a crash or not, and their fuel consumption. We then analyze a driver’s movements through an ENTIRE trip – not just events – to create what we call a DriverDNA of the trip, which we compare against existing trips in our database to look for similarities. This is pattern profiling. We currently have over 7 billion driving patterns in our database, trained with data from 106 countries and 1,600 cities.
Our AI is regularly validated by customer outcomes.
Additionally, our AI has received many
awards and recognitions.
What does the driver impact mean?
We predict a driver’s crash probability and climate impact. Collectively, we call this a driver’s impact on the roads. We calculate driver impact by analyzing a driver’s movements through an entire trip and not “events” like harsh braking, harsh acceleration or speeding.
Is the technology reliable?
Our AI is based on outcomes we know have happened, and by training it extensively over the last 20 years and working with companies globally, we’ve been able to see just how accurately it can predict risk and climate impact.
Is there a difference between a small and large car?
Our AI is trained with trip data from all kinds of vehicles and has learned how to make driving comparisons no matter the size of the vehicle or type of vehicle.
Is the weather taken into account?
Not by using any sensors on the vehicle but people change their driving patterns according to weather conditions. Our AI detects these changes and makes an accurate pattern analysis no matter the weather.
Do you track my every move both during the day and at night?
Data is only tracked when vehicles are driven.
What about privacy - how do you handle the data?
All data within the Greater Than cloud provides full protection of personal data using the latest encryption technologies – this protects and ensures the highest privacy level, in line with GDPR.
Isn’t safe driving eco-friendly driving?
Safe driving and eco-friendly driving have a very strong correlation. But there are a few rare cases where they don’t match exactly. For example, having consistent speed for as long as possible on a curvy road is good for fuel consumption- but can be very bad for safety if it means the driver has to brake hard or do quick cornering!
Can you harmonize the data?
Yes, thanks to the uniqueness of our scoring, our AI harmonizes GPS data from any source and is agnostic of location or vehicle type, making it easy to benchmark performance across mixed fleets and quickly pinpoint areas for potential CO2 emissions reductions and driver safety improvements.