Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
Handhelds The Almighty Buck Transportation

Insurer Measures Driver Safety With Smartphone App To Calculate Premiums 345

Qedward writes "Motorists are being invited to help develop a new driving app that could earn them a discount of 'up to 20%' on their motor insurance. British insurer Aviva is using smartphone technology to create individual driver profiles that will be used to calculate tailored pay-how-you-drive premiums. The driver behavioral app, Aviva RateMyDrive, will monitor motorists taking part in the test for 200 miles, including acceleration, braking and cornering. This data is then turned into an individual score which helps determine the motorist's premium, with 'safer' drivers earning up to 20% off their deal."
This discussion has been archived. No new comments can be posted.

Insurer Measures Driver Safety With Smartphone App To Calculate Premiums

Comments Filter:
  • by Anonymous Coward on Friday August 17, 2012 @02:13AM (#41020875)

    Besides the fact that this is begging to be gamed, how to they tell the difference between someone driving carefully and some half-blind octogenarian that's causing traffic accidents around them by driving too slow and failing to react to near-misses that may affect the next driver?

    • Re: (Score:3, Insightful)

      Wouldn't that just be a matter of statistics? The more data that's captured, the more patterns will emerge.

    • by wierd_w ( 1375923 ) on Friday August 17, 2012 @03:31AM (#41021369)

      I'd say it is more a slippery slope:

      The insurance company incentivises people to provide very detailed information about themselves, that they would normally never provide, and may even try to prevent being obtained.

      In the process, they build a precedent that will penalize people that are unwilling to provide this data willingly.

      EG, it starts out as "If I voluntarily join this program, I could say 20% on my insurance." It then later becomes the "New standard rate metric, based on your personal driving patterns," and eventually becomes "Penalized rate for not providing data on your traffic patterns."

      While it looks good now, it wont look so good to people who value their privacy in the future. They will be lumped in with people who are clearly bad drivers but dont want to admit it, and want to hide that fact from the insurance companies.

      • They certainly won't share their methods of interpreting the driving data, so it would be an easy excuse to charge you more money. "Mr. Mahoney, according to your driving data you will be charged a 5% premium over the standard rate. Better luck next time."

        It's interesting but if they aren't correlating the data with red lights, traffic signs, and impact on other drivers, it's of limited use. Unless you're one of those people who habitually drive 50 over the speed limit.

        • by Lehk228 ( 705449 )
          they will have to, at least in NY they must disclose all methods they use in determining insurance rates
      • by N1AK ( 864906 )
        How is that different to anything else? First class transport led to a decrease in the quality of economy transport (along with a drop in price).

        Insurance is based on transferring a large personal risk to another entity at a cost you can afford. In the UK the insurance market is very competitive, there are hundreds of suppliers. If it works out that a young new driver can get insurance for £1,600 rather than £2,000 if they are willing to provide more information. If someone is willing to make
      • EG, it starts out as "If I voluntarily join this program, I could say 20% on my insurance." It then later becomes the "New standard rate metric, based on your personal driving patterns," and eventually becomes "Penalized rate for not providing data on your traffic patterns."

        Sounds all too familiar to those living in the Chicago area. While the tollways were being updated to new open-road tolling (i.e. you don't need to stop at a booth), everyone was told that those who had an iPass (the electronic toll device) would receive a discount. Of course the discount turned into: those with iPass pay the same (high) cost as before and everyone using cash pays double.

    • by Sqr(twg) ( 2126054 ) on Friday August 17, 2012 @03:57AM (#41021453)

      how to they tell the difference between someone driving carefully and some half-blind octogenarian that's causing traffic accidents around them by driving too slow ... ?

      Correlating speed to position and a database of speed limits will tell you if people are driving too fast or too slow. (In certain cases, driving slower than the speed limit is the correct action, so you'd have to look at a large dataset to differentiate between those who adapt to circumstances and those who always drive to slow.)

      In general, slow drivers aren't a problem for insurance companies. If you drive slowly and another car gets into an accident while trying to overtake, it's typically his insurace that will have to pay, because he should have waited until it was safe to pass.

      I suspect they are trying to weed out the young drivers who have never been in a near-accident and believe that they can drive 20 mph over the speed limit, because they have such a good car, and their reactions are so much better than other people's. If they can eliminate that subset of drivers, they wouldn't have to have such high premiums for young people in general.

      • by History's Coming To ( 1059484 ) on Friday August 17, 2012 @04:06AM (#41021489) Journal
        The UK insurance industry has a huge problem with bogus whiplash claims - the slightest little bump and lawyers are pushing for thousands of pounds in compensation for a medical condition which doctors admit is almost impossible to prove either way. This has lead to a fivefold increase in some insurance costs over the last fifteen years. Schemes like this, and others where rolling camera footage is stored, are an attempt to show that these low speed collisions are generating claims far beyond what is reasonable.
      • by arth1 ( 260657 )

        In general, slow drivers aren't a problem for insurance companies. If you drive slowly and another car gets into an accident while trying to overtake, it's typically his insurace that will have to pay, because he should have waited until it was safe to pass.

        The main problem I have in traffic are people who turn onto a road, and then refuse to step on it, presumably because they think acceleration is dangerous. (Then, of course, the idiots slowly creep up to 10 above the speed limit, because they don't think speed is dangerous.)

        So when someone pulls out in front of them and refuses to get up to speed, someone doing the speed limit in a heavy vehicle has to step on the brakes and do evasive maneuvers to avoid rear-ending them.

        With this system, guess who gets pe

        • I could guess, but until we see the actual algorithms they use (i.e. never) that's all it would be. Insurance companies tend to have pretty good data on what kind of behaviour gets you involved in accidents, so I wouldn't be so quick to assume they'll be idiots about it.

          Let's examine their motivation for a moment...
          They want to raise the premium for those that lodge a lot of insurance claims. I.e. those that get into a lot of accidents. Granted they SAY the opposite - that they want to lower the premiums fo

    • by SteveAyre ( 209812 ) on Friday August 17, 2012 @04:11AM (#41021509)

      Aviva developed a Pay As You Go insurance system several years ago now.
      http://www.aviva.co.uk/media-centre/story/2840/norwich-union-launches-innovative-pay-as%20you-drive/ [aviva.co.uk]

      We studied it as part of a project during my CompSci course about the time it was launched.

      Essentially you agree that they put a GPS tracker in your car. It monitors your speed/acceleration/braking/etc (just like the app). You then only pay insurance for when you are driving, and the price is affected by how well you drive. It's been around for some time now. It's fixed to your car, and if you remove it from your car so they don't see your bad driving you're illegally driving without insurance.

      All the phone app is is a free trial of that type of insurance - far cheaper to give them an app than send them a tracker. If you were to actually buy their insurance there's no way they'd let you keep using the phone app for it. Too much chance of forgetting the phone or battery dying, let alone any 'gaming'.

      • by jafiwam ( 310805 )

        I'd be interested in such an app just for the data.

        Knowing what the insurance companies have determined is statistically higher accident risk behavior would help teach better driving methods.

        I do all sorts of stuff that's a little different than what most people do to minimize accident risk. Over a lifetime of driving, it has apparently worked well for reducing accidents. (I had one, within the first year of driving when I still didn't know what I was doing. That was 18 years ago.)

        For example, I stay

        • For example, I stay way back from the wide white line at intersections, the more traffic or faster the cross traffic is the farther back I stay. There is NO REASON to get up close unless turning right. All that does is set up a situation where an accident in the intersection could push cars into mine. The extra 10 feet doesn't matter for starting up again when the light turns green.

          In some cases it also makes the sensors work. Here in the Netherlands most traffic lights detect whether there is a car waiting on the light. Most of these sensors are within 2 meters (6 feet) from the white line. If you keep 3,3 meters (10 feet) of distance then it will not notice you and thus it will not turn green.

    • They don't care. They don't have to pay for the near misses or the accidents caused indirectly. They only care about what they have to pay for.
    • by DrXym ( 126579 )
      Octogenarians don't tend to have smart phones or install apps for insurance quotes. Also, even if it could be gamed, the basic question from the insurer's point of view is do they make more money from providing the app or not? Even if we assume x% of people somehow manage to con the app, does the remainder who use it in good faith allow the insurer to more accurately calculate risk and therefore the quotes it offers? If the answer is yes then it's clear why they may do it. The easiest way to game the app of
    • by chrb ( 1083577 )

      how to they tell the difference between someone driving carefully and some half-blind octogenarian that's causing traffic accidents around them by driving too slow

      Statistically speaking, anyone who is regularly causing traffic crashes around them is highly likely to be involved in one or more of those crashes. It is the nature of being a reckless driver that, sooner or later, you will be in a crash.

      Aviva is a huge business (sixth-largest insurance company in the world) - they employ some of the best actuaries in the world to figure out their risk models. They are going to be studying the statistics of driver behaviour in detail and will surely come up with more tha

  • by Riddler Sensei ( 979333 ) on Friday August 17, 2012 @02:15AM (#41020893)

    I'm not too sure this is a universally good idea. Sometimes traffic gives you a tricky situation and you need to accelerate or do a quick lane change to avoid a potential accident. In those moments I'm not too sure it's good to introduce the thought, "Oh, but wait, that may increase my premium".

    • by kwark ( 512736 )

      "Sometimes...."

      Indeed sometimes. That sometimes you have to act will show abnormal in the data. If that sometimes becomes often and thus a pattern, either change your route (or timing) to a safer one or be come a better driver by anticipating more if you thing those anomalies in the data weren't your fault in the first place.

    • by deesine ( 722173 )

      I'm not too sure either.

      I can't help but think of the time I was in Vegas at a blackjack table. The gal next to me was asking the dealer what this "Insurance" thing was written on the table. He answered, "Well, if it's advertised on the table, by the casino, do you think it's good for you, or good for them?"

      --

      • by Cederic ( 9623 )

        False comparison. This is actually a potential benefit for both parties.

        The insurer wins as there should be a reduced number of claims, resulting in reduced staff numbers required and lower payouts for claims.

        The driver wins as they get a lower premium.

        The rest of us lose out as we'll now get stuck behind slow cocks that don't dare accelerate away from junctions, adding to overall congestion and causing accidents for others as a result. But that doesn't make it a bad choice for the insured driver.

  • by LSDelirious ( 1569065 ) on Friday August 17, 2012 @02:18AM (#41020905)
    Wonder how they would rate me if I took the bus to work for a week? Certainly wouldn't catch me speeding or accelerating/decelerating too hard, but I wonder how the frequent stops would factor in? Also if you didn't put your phone into airplane mode, would being a passenger in a 737 double your rates when they clock you doing 150+mph at takeoff before you ascend above cell reception range?
    • I wonder how different roads will affect it? Much of my driving is done on twisty country roads with speeds varying from 30mph to 60mph. I'm guessing the app won't say much about things like my road position or anticipation. If I drive on the motorway it's just a case of stick it in 6th and rumble along at 70mph and 2200rpm in a straightish line for the day, with breaks every two hours. I could be sitting there reading a book or posting on slashdot - whoops, had to change lanes for that lorry there - an

    • 150+mph, didn't crash, good driver!

  • by Gordonjcp ( 186804 ) on Friday August 17, 2012 @02:18AM (#41020909) Homepage

    Drive sensibly while you're running the app, drive like a nutter when you're not.

    On a more serious note, if this ran *all the time* then it may provide useful metrics on driver ability without the privacy concerns of GPS tracking. Yes, you could *theoretically* estimate someone's position from the accelerometer data - that is, after all, how Intertial Navigation Systems work - but it wouldn't be very accurate. You could estimate someone's position from cell handoff too, if you included that in the data, but then you'd have to be *trying* to be creepy ;-)

    One of the companies I work with installs GPS trackers in vehicles for things like lorries, heavy plant and such. Their system has an option for an accelerometer that will beep if the drivier is accelerating too quickly, and thus wasting a lot of fuel. One biggish fleet has apparently saved about 1 million Euros on diesel alone using this, never mind tyres and repairs.

    • this could be fun convincing them they have bugs in their system, since accelerometer should still record rapid vertical acceleration but that wouldn't translate to much change of latitude/longitude in gps... think "tower drop" carnival rides, bungee jumping, fast elevators, etc...
  • Drive too much? (Score:4, Insightful)

    by abelb ( 1365345 ) on Friday August 17, 2012 @02:23AM (#41020933)
    How long before the insurance company succumbs to the temptation of penalizing those who use their cars too much? The more time you spend on the road the higher the chance that you'll be involved in an incident, regardless of how well you drive. You can see how such information could be used to discriminate against people living in rural areas and those living further from their place of work.
    • Re:Drive too much? (Score:5, Informative)

      by BradleyUffner ( 103496 ) on Friday August 17, 2012 @02:27AM (#41020955) Homepage

      How long before the insurance company succumbs to the temptation of penalizing those who use their cars too much? The more time you spend on the road the higher the chance that you'll be involved in an incident, regardless of how well you drive. You can see how such information could be used to discriminate against people living in rural areas and those living further from their place of work.

      I thought insurance companies already do this. Every company I've had a policy with has always wanted current and yearly mileage when I signed up. Driving fewer miles in a year resulted in lower premiums.

    • Re:Drive too much? (Score:4, Insightful)

      by Alex ( 342 ) on Friday August 17, 2012 @02:34AM (#41021017)

      That's not discrimination, its common sense.

      If you do more miles - all things being equal, you are more likely to have an accident - so it makes sense for the insurer to charge you more. The only reason they aren't doing it yet is they've not found a good way to measure it yet, I'm sure they are working on it though.

      Presumably now you are going to complain about insurers "discriminating" against people who live on flood plains, in high risk crime areas and arsonists ?

      Alex

    • by abelb ( 1365345 )
      I agree with the comments above. I think what gets me about this is that they're selling it as a benefit to drivers when in reality it's the insurance companies that have the most to gain.
    • by sqldr ( 838964 )

      The more time you spend on the road the higher the chance that you'll be involved in an incident

      From the insurance company's point of view, that's entirely the point. What about people in urban areas who rarely use their car paying for the accidents of regular drivers?

  • Plead the 5th (Score:5, Insightful)

    by LSDelirious ( 1569065 ) on Friday August 17, 2012 @02:27AM (#41020961)
    Letting them track you is like talking to a cop who's placed you under arrest... they might convince you that you're being given a chance to prove what an upstanding law abiding citizen you are, but in reality they're only looking for the incriminating parts to hold against you. Its the marketing folks jobs to come up with hypothetical situations where you can save money so you'll switch to their brand... its the bean counters and their lawyers jobs to see that you don't ever actually qualify for said hypothetical discounts, and you are giving them the ammo...
    • by sqldr ( 838964 )

      nobody is forcing you to take the deal after they penalise you. There's other insurance companies. It's worth a punt.

  • To hell with that. (Score:4, Informative)

    by Aryden ( 1872756 ) on Friday August 17, 2012 @02:30AM (#41020977)
    Progressive is already using a feature like this in the U.S. It's just not a smart phone app. It's actually a little box you put in your car. It's called Snapshot. [progressive.com] Not my kind of thing. There is just no way for the insurance company to know what is or is not going on around you when you're driving.
    • by chrb ( 1083577 )

      There is just no way for the insurance company to know what is or is not going on around you when you're driving.

      Aviva employs some of the best actuaries in the world - don't you think that they are capable of developing statistical models that take rare incidents into account? One swerve in thousands of miles of driving is going to appear as noise, a fast turn with hard acceleration on every corner is going to appear as a substantial crash risk.

  • Only 200 miles (Score:5, Insightful)

    by AmiMoJo ( 196126 ) on Friday August 17, 2012 @02:31AM (#41020989) Homepage Journal

    So I have to drive carefully for 200 miles to get my rating up and then I can turn it off and go back to my old habits? Or just swap phones with my mum for 200 miles? Or just not take my (primary) phone when I want to have some fun?

    • by Inda ( 580031 )
      Yes, yes! Driving like a looney through residential areas is fun!

      Go to a track. Pay the price of a good night out. Speed round in something like a Formula Ford. Crash and die if that pleases you.

      I know I did. Best 80 quid I spent in a long time.

      Don't be a twat on the roads.
  • by c0mpliant ( 1516433 ) on Friday August 17, 2012 @02:39AM (#41021045)
    For years insurance companies have been doing the exact same thing of estimating how good or bad a driver you are based on your age, gender, occupation etc. Now they're proposing to allow you to determine how good a driver you are based on using an app for not too long of a time really.

    Is there a potential for it to be misused, yeah, but I'd welcome any move to judge my driving over lumping me in with a particular age group or gender.
  • Sounds like malignant nonsense. And I hope people don't accept it, else we might soon need smartphones to obtain drivers-licenses. Insurance companies, although happily masquerading as safety-options, are mafia in disguise. They have purchased their way into government and become mandatory by force, compelling people to pay for protection whether they want it or not. Of course drivers need protection from the irresponsible or even honest accidents (collisions) of others, but with such a huge amount of them
    • Slippery Slope indeed! I don't see the govt requiring smartphone monitoring for the license (unless they just make it a part of "citizenship"), but I could easily see the 200 mile "trial period" being extended indefinitely so that you always had to be monitored to get discount, which like all the 1 sided contracts coporations push can be changed at any time, meaning any action they disapprove of could invalidate the "discount". And of course this would eventually morph into always be monitored to have a pol
  • The idea to customize insurance up to individual profiles is completely opposite to the very first idea of the insurance itself which is a way to share a risk within a large pool of fellows in order to distribute the cost. If you start building precise profiles of individuals and charge them accordingly, you defeat the idea behind the insurance. At term, you will charge the whole risk to each individual and they will no longer see advantages to insure themselves. Insurance is about sharing a risk over the l
    • Between zero and full risk sharing, are you able to imagine *partial* risk sharing?

    • by Cederic ( 9623 )

      The idea to customize insurance up to individual profiles is completely opposite to the very first idea of the insurance itself which is a way to share a risk within a large pool of fellows in order to distribute the cost.

      Why do you think this? That happens to be the primary implementation in consumer markets due to the difficulty of accurately assessing individual risk, but it's also possible to go to an insurance company and get a very individualised tailored policy covering something nobody else on the planet has - e.g. a supermodel's legs.

      Insurance is about sharing a risk over the largest population possible.

      No. Insurance is about offsetting a risk. Sharing it over a large population is often a more efficient mechanism, but far from essential.

    • That's why life insurance premiums don't discriminate based on your age and whether you are a smoker, right?

  • Texting (Score:2, Funny)

    by Anonymous Coward

    Does it jack up your rates if you text or talk on the phone while driving?

  • The modern world is based on statistics and conforming to expectations, whether that's an aptitude test determining what you're "probably" good at to some crude metrics determining whether you're "probably" a safe driver. Everyone is fitted into neat little categories and self-fulfilling prophecies are created, reinforcing existing prejudices and providing little scope for social improvement.

    No more is this true than with driving: young men are essentially told that they are high risk. It's like the classic

  • MAPFRE YCar (Score:4, Informative)

    by paugq ( 443696 ) <pgquiles@elpauer . o rg> on Friday August 17, 2012 @03:00AM (#41021187) Homepage

    How is this new?

    In Spain, MAPFRE has been offering for at least 4 years the YCar line of insurance for young drivers which offers as much as a 40% discount if you install a GPS-like device which sends them information about when you drive, what speed you drive, how many kilometers, etc.

    If you speed up, drive on "dangerous" hours (e. g. weekend 2 AM - 6 AM), etc, you lose the discount for next year.

    http://www.mapfre.com/seguros/es/particulares/soluciones/seguros-coches-jovenes-ycar.shtml [mapfre.com]

    There are several policies to choose and some of them even allow to adjust the policy clauses, for instance in case you are a young driver who works the night shift.

  • The reality is that everybody thinks they are above average driver. So if they get a score that is below average, they know they will pay more to offset the drivers who pay less. So they simply walk away. So ultimately exactly half of the drivers would get a worse deal in a perfect world. The insurance company will end up with better than average drivers only ... but these drivers will pay less insurance. The insurance company pays out less but earns less per driver. That is a one to one relationship, so th
    • I think you are partially right. I’ll presume that when the insurance company manages to rid itself of the bottom half of its driver pool, its payouts decrease more than its collected premiums. In effect, the insurance company is defecting against its competitors by sticking them with the higher cost customers.

    • by Lehk228 ( 705449 )
      you are neglecting to consider that the reduction in rate would be less than the reduction in frequency and magnitude of payouts. it certainly would be, that is how insurance companies make their money
  • If anyone thinks the scumbag insurance companies are going to give anyone a 20% discount, I have some prime swamp land for you.

    • by Lumpy ( 12016 )

      And that would be the truth. In reality, good drivers get to keep their current rates, bad drivers rates go up by 20%.

  • Bring back personal liability. If you drive like an ass, you cause an accident, you are liable for all damages you caused. not your insurance company, YOU.

    you were busy texting and hit a motorcyclist? everything you own is now the property of the motorcyclists family, as well as 50% of all your income for the next 30 years.

    You dont want that liability? then buy $10,000 of liability insurance at $1500 a month, or stop driving like a moron.

Keep up the good work! But please don't ask me to help.

Working...