Exploring Telematics Part 2: What Exactly Is It?

The Floow - May 16th, 2014

There are many variants of telematics but in all of them, you have a device in your vehicle which records information about how, where and when you drive via a GPS receiver and in some cases, an accelerometer.

This data can now be collected from:

  • A ‘black-box’ wired to the car’s electronics and hidden somewhere in the vehicle 
  • A ‘self-installed device’ which is generally plugged into the vehicle’s on-board diagnostic port 
  • A smartphone

And there are, of course, pros and cons for each of the three device types.

A black-box is expensive to buy and has to be fitted by a professional installer but once it has been fitted, it can be used for stolen vehicle recovery as well as telematics insurance and there is little chance of fraud as it is nearly impossible for the customer to turn it off if they are in a hurry. 

Smartphones are all pervasive now so customers are used to the technology and sharing data via this medium. They are also transportable, dealing well with multi-vehicle policies but there are issues around the quality of the accelerometer for crash reconstruction and you cannot be certain whether the policyholder is the driver or whether they are even in their vehicle. 

OBD devices sit somewhere in the middle, they are easy for the customer to fit but they are also easy to unplug and so they sometimes suffer from potential fraud issues. The costs also vary wildly, around 70% of the insurance market pay under £600 for their policies so insurers are not usually keen to spend more than an additional £50 per policy to gain any telematics benefit.

Whichever device is used, the data from the GPS and accelerometer are analysed and key attributes are derived including:

  • location
  • speed
  • distance travelled
  • length of time driven for
  • time of day journey took place
  • accelerations (& decelerations) both longitudinally and laterally
  • harsh acceleration, braking or swerving events
  • crashes 

This data can then be used to inform the price of the policy as well as provide timely and accurate information when an accident does occur which can help the insurer to record quality first notification of loss information and understand liability as well as challenge potentially fraudulent claims.

There are then 3 main ways in which the policy can be constructed:

  1. Pay As You Drive

This is, in effect, a limited mileage policy. The rating factors and basic price can be pretty much the same as a normal motor policy with the exception that you can set a low mileage limit. The important thing is, of course, that now you can monitor the actual mileage driven and therefore, you can be confident on the initial discount and request additional payments per 100 miles once the limit is breached.

  1. Pay When/Where You Drive

There is a general understanding that driving in the dark is more dangerous than during the day, especially for younger drivers, but a standard motor policy cannot understand the exposure over these time periods and cannot charge for them differently. 

However, with telematics you can know exactly when the driving is taking place and charge for a different level of risk (by basing each months premium on last months driving, giving a refund at the end of the month or imposing a levy each time the policyholder drives at the wrong time). 

A similar method can be used to allow for where the policyholder drives. On a standard policy, the home postcode is the only proxy you have but with telematics you know exactly where the driving is taking place so can offer a discount for driving through good areas or driving on safe roads (e.g. motorways).

  1. Pay How You Drive

This is rapidly becoming the most common method, it uses knowledge of the driving style to build an understanding of the likelihood of the driver being involved in an accident which can be used as an additional rating factor in the pricing algorithm once some driver data has been built up.

It is this final approach that we recommend at The Floow. Having worked closely with leading academics in the UK and across the world, we have built algorithms based on the acceleration, speed, level of distraction and other factors which influence the probability of having an accident and the severity of impact if one does occur. 

We believe that each of the data collection methods has its own place in the market and we have designed our systems to accept data from fitted boxes, OBD devices or smartphones and provide a simple scoring mechanism back to both insurers and policyholders.

The challenge now for the insurer is to make it work. Success or failure seems to currently boil down to two elements:

  1. Can you process and understand the many millions of data points you receive as a telematics provider in real (or near-real) time to give scores back to the customer and the insurer when they expect them and to give swift first notification of loss when an accident happens?
  2. Can you win the customer’s heart as well as their mind and make them want to buy the policy?

This article is part of a series where we explore telematics and the history behind it, in the first part of the series we looked at the history of telematics, you can read that here. 

This content was written by Andy Goldby and originally featured in the July issue of Actuarial Post magazine published in 2013.

Share this article