Telematics technology may be a game changer for the automobile insurance industry but it comes with some caveats as Colin Sowman discovers. James Bielak, (P&C) program manager at the US office of ACORD (the Association for Cooperative Operations Research and Development), has an unenviable job: to devise a standard form of communicating vehicle data between telematics providers and insurance companies. To that end he has gathered together a group composed of insurers, telematics providers and other intere
Telematics Data Standards Ecosystem
Telematics technology may be a game changer for the automobile insurance industry but it comes with some caveats as Colin Sowman discovers
James Bielak, (P&C) program manager at the US office of ACORD (the Association for Cooperative Operations Research and Development), has an unenviable job: to devise a standard form of communicating vehicle data between6224 Telematics providers and insurance companies. To that end he has gathered together a group composed of insurers, Telematics providers and other interested parties – but it hasn’t been easy.
“Insurance companies are extremely reluctant to divulge what data they collect and how they use it because to do so could reveal their strategy for coping with the new data-rich landscape Telematics provide,” says Bielak. This situation is further complicated as one insurance company has patented certain processes that supports its Telematics systems and has filed law suits against others for infringement.
Despite what may appear to be something of a land grab, a move to Telematics-based insurance is not a panacea for insurance companies or drivers.
A paper published last year by1979 Deloitte Consulting¹ highlighted the fine balancing act associated with a shift to Telematics-based insurance: the necessary infrastructure is not cheap while those opting to use Telematics-based policies are doing so to reduce their premiums. Some of the safest drivers can qualify for discounts of between 30% and 50%.
Almost half of American drivers surveyed would require a discount of more than 20% before they would consider having a Telematics system fitted to their car.
If the trend for safer drivers to realise lower premiums through the use of Telematics systems is not balanced by an increase in the premiums of drivers not using Telematics, the result will be that insurance companies suffer reductions in income. But if, as Deloitte makes clear, an insurer decides not to offer its customers a Telematics-based option, it risks losing its safest (and most profitable) drivers to its competitors.
It is therefore easy to see why American insurance companies are keen to reduce the implementation and running costs of Telematics-based insurance and that’s where Bielak’s efforts come in. The implementation of a standard form of communication should lower data collection costs considerably as he explains: “Currently the Telematics Provider may send reports to insurer X in one format and insurer Y in another and this pushes up administrative costs. Equally, if the insurer receives data from one Provider in one format and a second Provider in a different format, this again increases administrative costs.
“If all reports used a standardised format, it will reduce costs for both the data Provider and the insurance company.”
Basically he is trying to get suppliers and insurers to agree to a standardised XML format for what is considered ‘core’ data. This means column A will always be, say, the vehicle id, column B is the device id, C is the time and date, D the latitude and longitude coordinates, E the odometer, F is speed and so on.
Technically this should be straightforward enough but because insurance companies are so reluctant to say what data they collect compiling the column headings for the proposed ‘spreadsheet’ has taken more than a year. Currently this ‘spreadsheet’ runs to more than 30 columns of what the insurers consider to be ‘core’ data and more can be added when (rather than if) needed.
“Not all insurance companies will use all the data available, but if all the core parameters are reported, each insurer can take what it needs for its own calculations,” says Bielak.
It is likely that insurers will be interested in the time of day a person is driving, how far they are driving and how fast, as well as events like hard braking and cornering. All this data and more is available from the Telematics unit installed in the car and is transmitted to the system supplier along with certain information from the car’s electronic system.
Information such as if the indicators or Brake lights are on at any particular instant and whether the horn has been sounded would be particularly pertinent immediately before an accident.
However, with each instrumented vehicle typically creating a set of figures every five seconds, the amount of information the insurers will receive will be enormous. Insurance companies have put systems in place to cope with this volume of data and extract the salient information they require, but the existing systems may have to be changed, modified or updated to accept information in the proposed standardised format.
Bielak stresses that this does not mean the end result will be identical because each insurer can and will interpret the data they receive in a different way. Even on the simplest level each insurer will have its own interpretation of what constitutes hard braking so one company may set a more lenient threshold than another.
While accepting that differences will remain, by processing each individual’s Telematics-derived driving data the insurance companies may ‘rate’ every driver and calculate their premiums for the level of cover required. Other insurers may simply use a pay by the mile formula.
As insurers move from an age/experience, vehicle, annual mileage and claims/penalties based assessment to bespoke appraisals of individual drivers’ routines and behaviours, there will be winners and losers. Drivers that are not currently differentiated under the traditional system could find themselves paying very different premiums if they adopt the telematics approach. Deloitte highlights the scenario of two older women, neighbours, who share similar credit and driving records and educational backgrounds: One aggressively drives 25 miles every day during rush hour to her job while the other is a slow and cautious retired driver who only uses her car on weekends.
Given that on average both drivers cover a similar number of miles per year and are accident free, using traditional pricing models they would pay similar premiums. But under a telematics-based policy, the premium of the weekend (and potentially safer) driver could be reduced.
According to Bielak, in some areas of the world insurance companies are already using this type of driver scoring (rather than their individual data) to quote insurance premiums. However, insurance companies may not be allowed to restrict a driver’s choice of insurer and up to 40% of drivers may change their insurer on any given year. Therefore there needs to be a degree of transparency (if not coordination) about how drivers’ scores are calculated.
Bielak says: “Because of the different ways insurers will process the data it is likely that drivers will get slightly different scores from one company to another. But, as with credit rating scores, the insurance industry will get to know that company A rates higher than company B and make the necessary allowances.”
He believes the data transmission standard on which the ACORD working group volunteers have been working will be agreed this year and is confident insurers and telematics providers will start to adopt the protocol by the end of 2014. As soon as one of the big insurance companies adopts the system all the telematics providers will be able to supply their data in the standardised format making it less onerous for other insurers to make the switch.
Once the protocol is in use by one or more insurance companies, Bielak’s working group will assess how the system is working and implement any changes that are needed. “Once the protocol has been proven, it would then be possible to make it available worldwide for anyone wishing to adopt the standard,” he says.
Assuming the solution does reduce the cost of implementing telematics-based insurance, it might be attractive for insurance companies to use telematics to rate all drivers. However, any form of compulsion would raise privacy issues that current telematics users are willing to set aside in return for potentially lower premiums. Equally, concerns could also be raised if insurance becomes unaffordable for certain groups of individuals such as young or disadvantaged drivers.
For Bielak, his work is set to continue as the group will move on to consider what additional data needs to be added to the XML file for reporting telematics data from fleet and commercial vehicles.
Some tasks, it seems, never end.
It is therefore easy to see why American insurance companies are keen to reduce the implementation and running costs of telematics-based insurance and that’s where Bielak’s efforts come in. The implementation of a standard form of communication should lower data collection costs considerably as he explains: “Currently the telematics provider may send reports to insurer X in one format and insurer Y in another and this pushes up administrative costs. Equally, if the insurer receives data from one provider in one format and a second provider in a different format, this again increases administrative costs.
“If all reports used a standardised format, it will reduce costs for both the data provider and the insurance company.”
Basically he is trying to get suppliers and insurers to agree to a standardised XML format for what is considered ‘core’ data. This means column A will always be, say, the vehicle id, column B is the device id, C is the time and date, D the latitude and longitude coordinates, E the odometer, F is speed and so on.
Technically this should be straightforward enough but because insurance companies are so reluctant to say what data they collect compiling the column headings for the proposed ‘spreadsheet’ has taken more than a year. Currently this ‘spreadsheet’ runs to more than 30 columns of what the insurers consider to be ‘core’ data and more can be added when (rather than if) needed.
“Not all insurance companies will use all the data available, but if all the core parameters are reported, each insurer can take what it needs for its own calculations,” says Bielak.
It is likely that insurers will be interested in the time of day a person is driving, how far they are driving and how fast, as well as events like hard braking and cornering. All this data and more is available from the telematics unit installed in the car and is transmitted to the system supplier along with certain information from the car’s electronic system. Information such as if the indicators or brake lights are on at any particular instant and whether the horn has been sounded would be particularly pertinent immediately before an accident.
However, with each instrumented vehicle typically creating a set of figures every five seconds, the amount of information the insurers will receive will be enormous. Insurance companies have put systems in place to cope with this volume of data and extract the salient information they require, but the existing systems may have to be changed, modified or updated to accept information in the proposed standardised format.
Bielak stresses that this does not mean the end result will be identical because each insurer can and will interpret the data they receive in a different way. Even on the simplest level each insurer will have its own interpretation of what constitutes hard braking so one company may set a more lenient threshold than another.
While accepting that differences will remain, by processing each individual’s telematics-derived driving data the insurance companies may ‘rate’ every driver and calculate their premiums for the level of cover required. Other insurers may simply use a pay by the mile formula.
As insurers move from an age/experience, vehicle, annual mileage and claims/penalties based assessment to bespoke appraisals of individual drivers’ routines and behaviours, there will be winners and losers. Drivers that are not currently differentiated under the traditional system could find themselves paying very different premiums if they adopt the telematics approach. Deloitte highlights the scenario of two older women, neighbours, who share similar credit and driving records and educational backgrounds: One aggressively drives 25 miles every day during rush hour to her job while the other is a slow and cautious retired driver who only uses her car on weekends.
Given that on average both drivers cover a similar number of miles per year and are accident free, using traditional pricing models they would pay similar premiums. But under a telematics-based policy, the premium of the weekend (and potentially safer) driver could be reduced.
According to Bielak, in some areas of the world insurance companies are already using this type of driver scoring (rather than their individual data) to quote insurance premiums. However, insurance companies may not be allowed to restrict a driver’s choice of insurer and up to 40% of drivers may change their insurer on any given year. Therefore there needs to be a degree of transparency (if not coordination) about how drivers’ scores are calculated.
Bielak says: “Because of the different ways insurers will process the data it is likely that drivers will get slightly different scores from one company to another. But, as with credit rating scores, the insurance industry will get to know that company A rates higher than company B and make the necessary allowances.”
He believes the data transmission standard on which the ACORD working group volunteers have been working will be agreed this year and is confident insurers and telematics providers will start to adopt the protocol by the end of 2014. As soon as one of the big insurance companies adopts the system all the telematics providers will be able to supply their data in the standardised format making it less onerous for other insurers to make the switch.
Once the protocol is in use by one or more insurance companies, Bielak’s working group will assess how the system is working and implement any changes that are needed. “Once the protocol has been proven, it would then be possible to make it available worldwide for anyone wishing to adopt the standard,” he says.
Assuming the solution does reduce the cost of implementing telematics-based insurance, it might be attractive for insurance companies to use telematics to rate all drivers. However, any form of compulsion would raise privacy issues that current telematics users are willing to set aside in return for potentially lower premiums. Equally, concerns could also be raised if insurance becomes unaffordable for certain groups of individuals such as young or disadvantaged drivers.
For Bielak, his work is set to continue as the group will move on to consider what additional data needs to be added to the XML file for reporting telematics data from fleet and commercial vehicles.
Some tasks, it seems, never end.
James Bielak, (P&C) program manager at the US office of ACORD (the Association for Cooperative Operations Research and Development), has an unenviable job: to devise a standard form of communicating vehicle data between
“Insurance companies are extremely reluctant to divulge what data they collect and how they use it because to do so could reveal their strategy for coping with the new data-rich landscape Telematics provide,” says Bielak. This situation is further complicated as one insurance company has patented certain processes that supports its Telematics systems and has filed law suits against others for infringement.
Despite what may appear to be something of a land grab, a move to Telematics-based insurance is not a panacea for insurance companies or drivers.
A paper published last year by
Almost half of American drivers surveyed would require a discount of more than 20% before they would consider having a Telematics system fitted to their car.
If the trend for safer drivers to realise lower premiums through the use of Telematics systems is not balanced by an increase in the premiums of drivers not using Telematics, the result will be that insurance companies suffer reductions in income. But if, as Deloitte makes clear, an insurer decides not to offer its customers a Telematics-based option, it risks losing its safest (and most profitable) drivers to its competitors.
It is therefore easy to see why American insurance companies are keen to reduce the implementation and running costs of Telematics-based insurance and that’s where Bielak’s efforts come in. The implementation of a standard form of communication should lower data collection costs considerably as he explains: “Currently the Telematics Provider may send reports to insurer X in one format and insurer Y in another and this pushes up administrative costs. Equally, if the insurer receives data from one Provider in one format and a second Provider in a different format, this again increases administrative costs.
“If all reports used a standardised format, it will reduce costs for both the data Provider and the insurance company.”
Basically he is trying to get suppliers and insurers to agree to a standardised XML format for what is considered ‘core’ data. This means column A will always be, say, the vehicle id, column B is the device id, C is the time and date, D the latitude and longitude coordinates, E the odometer, F is speed and so on.
Technically this should be straightforward enough but because insurance companies are so reluctant to say what data they collect compiling the column headings for the proposed ‘spreadsheet’ has taken more than a year. Currently this ‘spreadsheet’ runs to more than 30 columns of what the insurers consider to be ‘core’ data and more can be added when (rather than if) needed.
“Not all insurance companies will use all the data available, but if all the core parameters are reported, each insurer can take what it needs for its own calculations,” says Bielak.
It is likely that insurers will be interested in the time of day a person is driving, how far they are driving and how fast, as well as events like hard braking and cornering. All this data and more is available from the Telematics unit installed in the car and is transmitted to the system supplier along with certain information from the car’s electronic system.
Information such as if the indicators or Brake lights are on at any particular instant and whether the horn has been sounded would be particularly pertinent immediately before an accident.
However, with each instrumented vehicle typically creating a set of figures every five seconds, the amount of information the insurers will receive will be enormous. Insurance companies have put systems in place to cope with this volume of data and extract the salient information they require, but the existing systems may have to be changed, modified or updated to accept information in the proposed standardised format.
Bielak stresses that this does not mean the end result will be identical because each insurer can and will interpret the data they receive in a different way. Even on the simplest level each insurer will have its own interpretation of what constitutes hard braking so one company may set a more lenient threshold than another.
While accepting that differences will remain, by processing each individual’s Telematics-derived driving data the insurance companies may ‘rate’ every driver and calculate their premiums for the level of cover required. Other insurers may simply use a pay by the mile formula.
As insurers move from an age/experience, vehicle, annual mileage and claims/penalties based assessment to bespoke appraisals of individual drivers’ routines and behaviours, there will be winners and losers. Drivers that are not currently differentiated under the traditional system could find themselves paying very different premiums if they adopt the telematics approach. Deloitte highlights the scenario of two older women, neighbours, who share similar credit and driving records and educational backgrounds: One aggressively drives 25 miles every day during rush hour to her job while the other is a slow and cautious retired driver who only uses her car on weekends.
Given that on average both drivers cover a similar number of miles per year and are accident free, using traditional pricing models they would pay similar premiums. But under a telematics-based policy, the premium of the weekend (and potentially safer) driver could be reduced.
According to Bielak, in some areas of the world insurance companies are already using this type of driver scoring (rather than their individual data) to quote insurance premiums. However, insurance companies may not be allowed to restrict a driver’s choice of insurer and up to 40% of drivers may change their insurer on any given year. Therefore there needs to be a degree of transparency (if not coordination) about how drivers’ scores are calculated.
Bielak says: “Because of the different ways insurers will process the data it is likely that drivers will get slightly different scores from one company to another. But, as with credit rating scores, the insurance industry will get to know that company A rates higher than company B and make the necessary allowances.”
He believes the data transmission standard on which the ACORD working group volunteers have been working will be agreed this year and is confident insurers and telematics providers will start to adopt the protocol by the end of 2014. As soon as one of the big insurance companies adopts the system all the telematics providers will be able to supply their data in the standardised format making it less onerous for other insurers to make the switch.
Once the protocol is in use by one or more insurance companies, Bielak’s working group will assess how the system is working and implement any changes that are needed. “Once the protocol has been proven, it would then be possible to make it available worldwide for anyone wishing to adopt the standard,” he says.
Assuming the solution does reduce the cost of implementing telematics-based insurance, it might be attractive for insurance companies to use telematics to rate all drivers. However, any form of compulsion would raise privacy issues that current telematics users are willing to set aside in return for potentially lower premiums. Equally, concerns could also be raised if insurance becomes unaffordable for certain groups of individuals such as young or disadvantaged drivers.
For Bielak, his work is set to continue as the group will move on to consider what additional data needs to be added to the XML file for reporting telematics data from fleet and commercial vehicles.
Some tasks, it seems, never end.
It is therefore easy to see why American insurance companies are keen to reduce the implementation and running costs of telematics-based insurance and that’s where Bielak’s efforts come in. The implementation of a standard form of communication should lower data collection costs considerably as he explains: “Currently the telematics provider may send reports to insurer X in one format and insurer Y in another and this pushes up administrative costs. Equally, if the insurer receives data from one provider in one format and a second provider in a different format, this again increases administrative costs.
“If all reports used a standardised format, it will reduce costs for both the data provider and the insurance company.”
Basically he is trying to get suppliers and insurers to agree to a standardised XML format for what is considered ‘core’ data. This means column A will always be, say, the vehicle id, column B is the device id, C is the time and date, D the latitude and longitude coordinates, E the odometer, F is speed and so on.
Technically this should be straightforward enough but because insurance companies are so reluctant to say what data they collect compiling the column headings for the proposed ‘spreadsheet’ has taken more than a year. Currently this ‘spreadsheet’ runs to more than 30 columns of what the insurers consider to be ‘core’ data and more can be added when (rather than if) needed.
“Not all insurance companies will use all the data available, but if all the core parameters are reported, each insurer can take what it needs for its own calculations,” says Bielak.
It is likely that insurers will be interested in the time of day a person is driving, how far they are driving and how fast, as well as events like hard braking and cornering. All this data and more is available from the telematics unit installed in the car and is transmitted to the system supplier along with certain information from the car’s electronic system. Information such as if the indicators or brake lights are on at any particular instant and whether the horn has been sounded would be particularly pertinent immediately before an accident.
However, with each instrumented vehicle typically creating a set of figures every five seconds, the amount of information the insurers will receive will be enormous. Insurance companies have put systems in place to cope with this volume of data and extract the salient information they require, but the existing systems may have to be changed, modified or updated to accept information in the proposed standardised format.
Bielak stresses that this does not mean the end result will be identical because each insurer can and will interpret the data they receive in a different way. Even on the simplest level each insurer will have its own interpretation of what constitutes hard braking so one company may set a more lenient threshold than another.
While accepting that differences will remain, by processing each individual’s telematics-derived driving data the insurance companies may ‘rate’ every driver and calculate their premiums for the level of cover required. Other insurers may simply use a pay by the mile formula.
As insurers move from an age/experience, vehicle, annual mileage and claims/penalties based assessment to bespoke appraisals of individual drivers’ routines and behaviours, there will be winners and losers. Drivers that are not currently differentiated under the traditional system could find themselves paying very different premiums if they adopt the telematics approach. Deloitte highlights the scenario of two older women, neighbours, who share similar credit and driving records and educational backgrounds: One aggressively drives 25 miles every day during rush hour to her job while the other is a slow and cautious retired driver who only uses her car on weekends.
Given that on average both drivers cover a similar number of miles per year and are accident free, using traditional pricing models they would pay similar premiums. But under a telematics-based policy, the premium of the weekend (and potentially safer) driver could be reduced.
According to Bielak, in some areas of the world insurance companies are already using this type of driver scoring (rather than their individual data) to quote insurance premiums. However, insurance companies may not be allowed to restrict a driver’s choice of insurer and up to 40% of drivers may change their insurer on any given year. Therefore there needs to be a degree of transparency (if not coordination) about how drivers’ scores are calculated.
Bielak says: “Because of the different ways insurers will process the data it is likely that drivers will get slightly different scores from one company to another. But, as with credit rating scores, the insurance industry will get to know that company A rates higher than company B and make the necessary allowances.”
He believes the data transmission standard on which the ACORD working group volunteers have been working will be agreed this year and is confident insurers and telematics providers will start to adopt the protocol by the end of 2014. As soon as one of the big insurance companies adopts the system all the telematics providers will be able to supply their data in the standardised format making it less onerous for other insurers to make the switch.
Once the protocol is in use by one or more insurance companies, Bielak’s working group will assess how the system is working and implement any changes that are needed. “Once the protocol has been proven, it would then be possible to make it available worldwide for anyone wishing to adopt the standard,” he says.
Assuming the solution does reduce the cost of implementing telematics-based insurance, it might be attractive for insurance companies to use telematics to rate all drivers. However, any form of compulsion would raise privacy issues that current telematics users are willing to set aside in return for potentially lower premiums. Equally, concerns could also be raised if insurance becomes unaffordable for certain groups of individuals such as young or disadvantaged drivers.
For Bielak, his work is set to continue as the group will move on to consider what additional data needs to be added to the XML file for reporting telematics data from fleet and commercial vehicles.
Some tasks, it seems, never end.
