We know that each insurer has its own unique risk assessment method in which the weight of individual factors may play a different role. Some of them use only telematics data to calculate the premium, and some combine them with demographic data. Our algorithm provides both access to specific data, support of our analysts in their processing, as well as ready-made scoring models developed and tested by us in practice.
We provide unique data and their unique combinations. In this text you will learn how we develop them. But before we get into how we do it − a brief introduction.
There are different types of telematics
‘We analyze the driving style of drivers for the purposes of motor policy. We do risk scoring for insurers’. That’s a promise made by nearly every insurance telematics provider. If you are also considering the use of telematics data in the calculation of premiums − you should be aware that, as in most complex topics, also here − the devil is in the details.
Basic telematics data
When analyzing drivers’ behavior, many insurance telematics service providers focus on parameters such as sudden braking and acceleration, distance covered, speed developed by the driver, their distraction or time spent behind the wheel. This is obviously important information that we also use in our models, but their value is relatively small if we do not compare them with the map context.
Map context in the analysis of data on drivers’ driving style
We can talk about map context when we are able to interpret the circumstances in which a specific event occurs. And it is not about the exact coordinates / GPS coordinates of a given place, but about the parameters that define this place, e.g.
– Road type / category,
– The speed limit in force on a given section,
– Information whether we are dealing with built-up areas or not,
– Dangerous places (bends or the so-called black spots).
Knowing these parameters, we don’t even need to know the exact location of the driver.
Digital maps in telematics. Why are the data so important?
The mere information about rapid acceleration and braking tells us little when it is not juxtaposed with the context of the speed limit. Imagine a situation where a driver leaves the highway / expressway at a speed of 100 km per hour and enters a built-up area very smoothly. They neither brake sharply, nor accelerate. Based only on such information, it might seem that we are dealing with an ideal driver. Given the speed limitation context, we know that this is a very dangerous driver. It is therefore important that both aspects − compliance with the rules and driving smoothness are always analyzed together.
Another example? Knowing that the driver regularly uses the phone while driving, we can assume that they are more likely to be distracted. But whether a distraction may result in a minor bump or a serious accident, we will only find out by analyzing the context of a given behavior. More details? Here you are. Writing or reading text messages requires the driver to take their eyes off the road for at least 5 seconds. One second of inattention is 13 meters at a speed of 50 km per hour. At a speed of 140 km/h, it is already nearly 40m, and with the assumed 5 seconds of distraction − 200 meters traveled! The conclusions are obvious.
The road context, and in particular the current speed limits, are crucial in assessing a driver’s profile. According to the report ‘Road accidents in Poland in 2020’ of the Road Traffic Office of the Police Headquarters, speed maladjustment was the most common cause of fatal accidents and accounted for 41% of such incidents. Ignoring these facts when profiling drivers is one of the biggest mistakes!
Comparison with other drivers
The next stage of risk analysis is to compare the driver with others, also traveling on a given road section. The analysis of a specific place becomes the key here. We have up-to-date road traffic data and we are able to check how other road users have behaved on a specific section of the road. Thanks to this, we know whether the behavior of the analyzed driver was typical for a given place (‘I was driving so fast because everyone drives there faster’) or whether we are dealing with someone who performs much worse than the average.
Analysis of the propensity to take risks
Drivers using telematics systems often complain that ‘one offense is enough to lose the discount’. But good profiling is much more than just analyzing individual events. Our system analyzes in a much more sophisticated way whether we are dealing with a risky driver. Single hard braking, whether in the city − caused by e.g. a change of traffic lights or out of town − resulting from a wildlife encounter, has less impact on the overall driving style index than repeated accelerations and braking at high speeds, which may prove a fondness for ‘bumper to bumper’ driving. Our driving style assessment model takes into account the sequence of maneuvers and the importance of individual events. Also with regard to the speed limits in force. Thanks to this, we are able to precisely ‘track’ drivers willing to take unnecessary risks.
Only after processing the telematics data, analyzing them against the map context and creating a driver profile, can we proceed to a reliable risk assessment.
The driver profile itself is only the starting point in this process. In order to estimate the risk of damage, we examine how often drivers representing particular profiles are involved in road incidents, and what conditions increase the likelihood of an accident.
Thanks to the statistical analysis of actual claims, supported by profiling of participating drivers, we are able to obtain the most reliable risk assessment, enabling the insurer to design such a premium calculation system that will simultaneously increase the competitiveness and profitability of its offer.
Unique road information database
For 17 years, Telematics Technologies has been the owner of one of the most popular navigation systems on the Polish market, the distinguishing feature of which is a unique database of information on the current traffic situation. We provide users with the most up-to-date data on:
- speed limits − obtained both from partners and with the help of our own research conducted on a continuous basis, during which our teams travel the roads, and constantly verify and update information,
- current traffic − obtained on an ongoing basis from over 600 thousand users of our applications