Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
GPS and Auto Insurance
9 views
Skip to first unread message
Ed M.
Feb 9, 2012, 6:21:14 PM2/9/12
to
Similar policies have been issued before. Another story:
http://www.bbc.co.uk/news/technology-16969509
"The AA is set to launch a new insurance policy which uses sat-nav
technology to track driver performance.
. . . The measures include monitoring speed, braking severity,
cornering and the types of roads used during certain times of day.
. . . Extreme speeds would be greeted with 'a stern email' to the
driver.
. . . Welsh insurance firm Motaquote has just launched a partnership
with leading sat-nav manufacturer TomTom to offer real-time feedback
on driver performance.
Other major insurers are expected to launch policies soon.
. . . Avis . . . said it might consider using the technology in its
vehicles.
Malcolm Tarling, from the Association of British Insurers, said it is
an approach which is likely to become commonplace."
http://www.theaa.com/insurance/pay-how-you-drive-car-insurance.html
https://www.theaadrivesafe.com/drivesafe_data.php
http://www.bbc.co.uk/news/technology-16969509
"The AA is set to launch a new insurance policy which uses sat-nav
technology to track driver performance.
. . . The measures include monitoring speed, braking severity,
cornering and the types of roads used during certain times of day.
. . . Extreme speeds would be greeted with 'a stern email' to the
driver.
. . . Welsh insurance firm Motaquote has just launched a partnership
with leading sat-nav manufacturer TomTom to offer real-time feedback
on driver performance.
Other major insurers are expected to launch policies soon.
. . . Avis . . . said it might consider using the technology in its
vehicles.
Malcolm Tarling, from the Association of British Insurers, said it is
an approach which is likely to become commonplace."
http://www.theaa.com/insurance/pay-how-you-drive-car-insurance.html
https://www.theaadrivesafe.com/drivesafe_data.php
Wolfgang S. Rupprecht
Feb 10, 2012, 2:46:05 AM2/10/12
to
"Ed M." <pat_...@yahoo.com> writes:
> . . . The measures include monitoring speed, braking severity,
> cornering and the types of roads used during certain times of day.
In my experience the really dangerous ones are the ones that aren't
paying attention to the road or the flow of the traffic. I don't see
how a gps and accelerometer is going to detect them.
On the other hand, the sharp braking and hard turning might be a very
good driver just goofing off on an empty road. Penalizing them (and
losing them as a customer) is going to negatively influence profits.
If they are going to be evil and spy on people they need to use the
right tools -- 4 cameras to record how well or poorly the driver is
interacting with other traffic.
-wolfgang
--
g+: https://plus.google.com/114566345864337108516/about
claudegps
Feb 10, 2012, 5:13:19 AM2/10/12
to
discounts depending on "how you drive", but it's yet clear how they
are going to evaluate this.
First there was a insurance based on "how much you drive". Many kms ->
pay a lot. Drive few kms -> pay very low. This is clear.
The worst thing is that it's seems that the government is going to
require this mandatory!
Now the insurance companies offers discounts for the user that decides
to have the "black box" installed in the car.
If the government makes this mandatory, the insurance companies does
not need to offer anything....
T.J. Higgins
Feb 10, 2012, 10:15:37 AM2/10/12
to
"Snapshot."
--
TJH
tjhiggin.at.hiwaay.dot.net
Alan Browne
Feb 10, 2012, 10:35:50 AM2/10/12
to
On 2012-02-09 18:21 , Ed M. wrote:
> Similar policies have been issued before. Another story:
>
> http://www.bbc.co.uk/news/technology-16969509
>
> "The AA is set to launch a new insurance policy which uses sat-nav
> technology to track driver performance.
I've always thought the day would come:
> Similar policies have been issued before. Another story:
>
> http://www.bbc.co.uk/news/technology-16969509
>
> "The AA is set to launch a new insurance policy which uses sat-nav
> technology to track driver performance.
-that all cars will have a black box recording many signals: GPS PVT,
speed, brake activation, gas pedal position, turn signals and so on.
Even video (forward, wide angle) of the last 60 seconds before a
collision (air bag release) and some period after.
-that roadside "radio speed limit signs" would broadcast the maximum
speed (standard or reduced due to road conditions) and that there would
be a governor on the car that would hold the car to that limit plus a
small reserve (about 10% above the limit). [ This would also allow the
increase of speed limits on most highways ]. This would also reduce the
costs of speed enforcement while reducing accidents due to excessive
speeding - and save fuel.
--
"We demand rigidly defined areas of doubt and uncertainty."
Douglas Adams - (Could have been a GPS engineer).
Hans-Georg Michna
Feb 10, 2012, 11:24:04 AM2/10/12
to
I predict that the institutions involved in this will not
publish how exactly they determine what is good or bad driving
from GPS data records.
I have severe doubts that they can actually predict the
likelihood of traffic accidents from GPS records. The
information is insufficient for the purpose.
They will probably find some correlations, but there will always
be drivers whose GPS records indicate that they are poor
drivers, and yet they have no accidents.
Hans-Georg
publish how exactly they determine what is good or bad driving
from GPS data records.
I have severe doubts that they can actually predict the
likelihood of traffic accidents from GPS records. The
information is insufficient for the purpose.
They will probably find some correlations, but there will always
be drivers whose GPS records indicate that they are poor
drivers, and yet they have no accidents.
Hans-Georg
Richard Owlett
Feb 10, 2012, 11:24:09 AM2/10/12
to
say just what you plug in. The ads also never say just how,
when, or *WHAT* the gadget communicates with/to "Progressive".
Also no mention of waiver of presumed privacy.
Alan Browne
Feb 10, 2012, 11:41:42 AM2/10/12
to
On 2012-02-10 11:24 , Hans-Georg Michna wrote:
> I predict that the institutions involved in this will not
> publish how exactly they determine what is good or bad driving
> from GPS data records.
Actually they do. They're goal is not to game the system but to pay out
> I predict that the institutions involved in this will not
> publish how exactly they determine what is good or bad driving
> from GPS data records.
less claims. So in installing such devices they benefit by telling you
exactly what they consider risky driving and what they will look for in
the record. If you don't drive "that way" they won't find it in the record.
> I have severe doubts that they can actually predict the
> likelihood of traffic accidents from GPS records. The
irresponsible behavior which increases the overall likelihood of accidents.
> information is insufficient for the purpose.
order of 5 Hz or more. (note that an intelligent recorder can record
"special events" at a higher/detailed rate and ho-hum - down the highway
at a lower rate - or not at all).
> They will probably find some correlations, but there will always
> be drivers whose GPS records indicate that they are poor
> drivers, and yet they have no accidents.
male driver pays high ins. costs due to his peers.
Collectively, it drives down the cost for "mature" drivers.
Yes, we've all met the little old lady who brags about never having had
an accident - yet her driving "skills" cause others grief and risk on
the road. But that is more exception than rule.
Demographically, the age of drivers is rising quickly in most areas. It
will be interesting to see how laws adapt (re-testing/training for older
drivers would be a good idea...).
Richard Owlett
Feb 10, 2012, 12:46:45 PM2/10/12
to
regulations for decades. In Missouri (and I presume other
states) there are provisions for knowledgeable person to
advise DMV that retesting should be done. In my case (after
a spinal cord injury) my doctor's opinion was I needed hand
controls. I thought of them as only a convenient aid. He
filled out form and less than 2 weeks later I had an
appointment for a road test. Have been driving with hand
controls for several years now :)
Alan Browne
Feb 10, 2012, 1:06:49 PM2/10/12
to
On 2012-02-10 12:46 , Richard Owlett wrote:
>> Demographically, the age of drivers is rising quickly in most areas. It
>> will be interesting to see how laws adapt (re-testing/training for older
>> drivers would be a good idea...).
>>
>
> IIRC age related retesting has been part of New York regulations for
> decades. In Missouri (and I presume other states) there are provisions
> for knowledgeable person to advise DMV that retesting should be done. In
> my case (after a spinal cord injury) my doctor's opinion was I needed
> hand controls. I thought of them as only a convenient aid. He filled out
> form and less than 2 weeks later I had an appointment for a road test.
> Have been driving with hand controls for several years now :)
There are no age re-testing requirements here other than medical and
>> Demographically, the age of drivers is rising quickly in most areas. It
>> will be interesting to see how laws adapt (re-testing/training for older
>> drivers would be a good idea...).
>>
>
> IIRC age related retesting has been part of New York regulations for
> decades. In Missouri (and I presume other states) there are provisions
> for knowledgeable person to advise DMV that retesting should be done. In
> my case (after a spinal cord injury) my doctor's opinion was I needed
> hand controls. I thought of them as only a convenient aid. He filled out
> form and less than 2 weeks later I had an appointment for a road test.
> Have been driving with hand controls for several years now :)
eyesight tests on license renewal from 75 on. (Quebec). I'm not sure
of the interval anymore - the photo renewal is every 5 years, payment is
every year (by mail) - but the license itself used to be a 2 year cycle.
Not clear what the cycle is for 75 and older. As a result of the
tests one can be limited to day driving only.
Ontario has a driver re-ed program with tests every 2 years. A road
test may be required (from age 80).
Brit-Col. has a medical test requirement every 2 years from 80.
Most other provinces have no specific rules.
Peter H. Coffin
Feb 10, 2012, 9:16:18 PM2/10/12
to
On Fri, 10 Feb 2012 10:24:09 -0600, Richard Owlett wrote:
> T.J. Higgins wrote:
>> Progressive Insurance in the USA also has this, under the name
>> "Snapshot."
>>
>
> I was wondering if that was a GPS device as the ads never
> say just what you plug in. The ads also never say just how,
> when, or *WHAT* the gadget communicates with/to "Progressive".
It goes into the OBD-II port.
> T.J. Higgins wrote:
>> Progressive Insurance in the USA also has this, under the name
>> "Snapshot."
>>
>
> I was wondering if that was a GPS device as the ads never
> say just what you plug in. The ads also never say just how,
> when, or *WHAT* the gadget communicates with/to "Progressive".
> Also no mention of waiver of presumed privacy.
driving, I'd think ti reasonable to assume you're going to have to sign
something that allows them to monitor your driving....
--
The pig is nothing but a giant dish which walks while waiting to be
served.
--Grimod de La Reynière
Peter H. Coffin
Feb 10, 2012, 9:23:07 PM2/10/12
to
On Fri, 10 Feb 2012 11:41:42 -0500, Alan Browne wrote:
> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
>> I predict that the institutions involved in this will not
>> publish how exactly they determine what is good or bad driving
>> from GPS data records.
>
> Actually they do. They're goal is not to game the system but to pay out
> less claims. So in installing such devices they benefit by telling you
> exactly what they consider risky driving and what they will look for in
> the record. If you don't drive "that way" they won't find it in the record.
And, critically, they can not pay out claims if the insured was doing
> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
>> I predict that the institutions involved in this will not
>> publish how exactly they determine what is good or bad driving
>> from GPS data records.
>
> Actually they do. They're goal is not to game the system but to pay out
> less claims. So in installing such devices they benefit by telling you
> exactly what they consider risky driving and what they will look for in
> the record. If you don't drive "that way" they won't find it in the record.
something illegal immediately prior to the claimed event. Like driving
faster than the speed limit. Or prove that the covered driver was NOT at
fault in a contested incident. May not matter a lot of the time, but ...
enough to pay for the gizmos, the reduced premium, and a pocketable
amount of difference left over on the quarterly reports....
Alan Browne
Feb 11, 2012, 11:25:59 AM2/11/12
to
damages no matter what you were doing. Drunk. Reckless. Stupid.
Breaking the law. They pay the cost to repair or replace the vehicle(s)
in the accident.
The exception to that would be if you were deliberately causing damage
to an insured property for the purpose of making a claim. (Ins. fraud).
(Where they "chisel" is on the injury side - and not over the 'cause'
but over the 'claims'.)
It's all priced into the premiums.
All monitoring will do is increase premiums on those deemed to drive in
a riskier fashion than others.
Hans-Georg Michna
Feb 12, 2012, 5:35:48 PM2/12/12
to
On Fri, 10 Feb 2012 11:41:42 -0500, Alan Browne wrote:
>On 2012-02-10 11:24 , Hans-Georg Michna wrote:
>> I predict that the institutions involved in this will not
>> publish how exactly they determine what is good or bad driving
>> from GPS data records.
>Actually they do. They're goal is not to game the system but to pay out
>less claims. So in installing such devices they benefit by telling you
>exactly what they consider risky driving and what they will look for in
>the record. If you don't drive "that way" they won't find it in the record.
Surprising for me. I will look at their definition, as soon as I
>> I predict that the institutions involved in this will not
>> publish how exactly they determine what is good or bad driving
>> from GPS data records.
>Actually they do. They're goal is not to game the system but to pay out
>less claims. So in installing such devices they benefit by telling you
>exactly what they consider risky driving and what they will look for in
>the record. If you don't drive "that way" they won't find it in the record.
get offered such an insurance.
>> I have severe doubts that they can actually predict the
>> likelihood of traffic accidents from GPS records. The
>They're not looking for likelihood of accidents, they're looking for
>irresponsible behavior which increases the overall likelihood of accidents.
accidents, then it increases the likelihood of accidents.
>> information is insufficient for the purpose.
>Depends on recording rate. For these purposes I would hope it is on the
>order of 5 Hz or more. (note that an intelligent recorder can record
>"special events" at a higher/detailed rate and ho-hum - down the highway
>at a lower rate - or not at all).
problem is that the GPS only records the movement of the car,
but the movement of the car is pretty much meaningless if you
don't know the situation it is in.
Let me give you a simple example. I drive, GPS-equipped.
Somebody in the neighboring lane swerves into my lane. I spot it
at the last moment and successfully avoid an accident by means
of a very sudden, extreme steering maneuver. My driving was
perfect and excellent, perhaps far above the average driver's
abilities (to spice the example a little).
Now guess what the insurance company reads out of my GPS record.
I could give you more examples. They are, in fact, easy to find.
Assume I avoid running over a child that jumps on the street,
but in saving the child's life I end up wrecking my car. I
suspect, had I rolled straight over the child, my GPS record
would have been impeccable, never mind that the value of a
child's life is incomparably higher than the value of my car,
and never mind that the insurance company would greatly prefer
to replace my car, rather than compensate a family for the loss
of their child.
I state that you cannot determine the likelihood of accidents
from any GPS record. I think the whole thing is almost a joke.
>> They will probably find some correlations, but there will always
>> be drivers whose GPS records indicate that they are poor
>> drivers, and yet they have no accidents.
>Individually it is often unfair: the well behaved unmarried 21 year old
>male driver pays high ins. costs due to his peers.
>
>Collectively, it drives down the cost for "mature" drivers.
Hans-Georg
Alan Browne
Feb 12, 2012, 6:05:24 PM2/12/12
to
On 2012-02-12 17:35 , Hans-Georg Michna wrote:
> On Fri, 10 Feb 2012 11:41:42 -0500, Alan Browne wrote:
>
>> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
>
>>> I predict that the institutions involved in this will not
>>> publish how exactly they determine what is good or bad driving
>>> from GPS data records.
>
>> Actually they do. They're goal is not to game the system but to pay out
>> less claims. So in installing such devices they benefit by telling you
>> exactly what they consider risky driving and what they will look for in
>> the record. If you don't drive "that way" they won't find it in the record.
>
> Surprising for me. I will look at their definition, as soon as I
> get offered such an insurance.
>
>>> I have severe doubts that they can actually predict the
>>> likelihood of traffic accidents from GPS records. The
>
>> They're not looking for likelihood of accidents, they're looking for
>> irresponsible behavior which increases the overall likelihood of accidents.
>
> That is a tautology. If anything increases the likelihood of
> accidents, then it increases the likelihood of accidents.
Ins. co's aren't looking at your past bad behaviour that didn't cause an
> On Fri, 10 Feb 2012 11:41:42 -0500, Alan Browne wrote:
>
>> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
>
>>> I predict that the institutions involved in this will not
>>> publish how exactly they determine what is good or bad driving
>>> from GPS data records.
>
>> Actually they do. They're goal is not to game the system but to pay out
>> less claims. So in installing such devices they benefit by telling you
>> exactly what they consider risky driving and what they will look for in
>> the record. If you don't drive "that way" they won't find it in the record.
>
> Surprising for me. I will look at their definition, as soon as I
> get offered such an insurance.
>
>>> I have severe doubts that they can actually predict the
>>> likelihood of traffic accidents from GPS records. The
>
>> They're not looking for likelihood of accidents, they're looking for
>> irresponsible behavior which increases the overall likelihood of accidents.
>
> That is a tautology. If anything increases the likelihood of
> accidents, then it increases the likelihood of accidents.
accident at that time. They are looking at your past bad behaviour that
makes it more likely that you will have an accident because of your bad
habits.
IOW, if you drive aggressively in corners, they believe it means you are
more likely to have an accident in any driving condition, not just
corners. It's your whole attitude towards driving that is suspect.
>>> information is insufficient for the purpose.
>
>> Depends on recording rate. For these purposes I would hope it is on the
>> order of 5 Hz or more. (note that an intelligent recorder can record
>> "special events" at a higher/detailed rate and ho-hum - down the highway
>> at a lower rate - or not at all).
>
> I don't think a high recording rate solves the problem. The
> problem is that the GPS only records the movement of the car,
> but the movement of the car is pretty much meaningless if you
> don't know the situation it is in.
recording most accurately shows the dynamics of the vehicle.
If you enter a curve and accelerate out, normally, this is revealed in
the data. If you accelerate aggressively, this is also shown.
If you jack rabbit off of a stop...
If you do not make complete stops at a stop sign...
If you drive significantly above the speed limit in a residential area...
If you consistently drive above the limit on the highway...
If you always make sharp/hard stops...
And so on.
And the more often you do this ( bad-behaviour / opportunities ) then
your "badness" goes up in the ins. co profile.
> Let me give you a simple example. I drive, GPS-equipped.
> Somebody in the neighboring lane swerves into my lane. I spot it
> at the last moment and successfully avoid an accident by means
> of a very sudden, extreme steering maneuver. My driving was
> perfect and excellent, perhaps far above the average driver's
> abilities (to spice the example a little).
>
> Now guess what the insurance company reads out of my GPS record.
valid fashion, that event would disappear into the noise.
> I could give you more examples. They are, in fact, easy to find.
> Assume I avoid running over a child that jumps on the street,
> but in saving the child's life I end up wrecking my car. I
> suspect, had I rolled straight over the child, my GPS record
> would have been impeccable, never mind that the value of a
> child's life is incomparably higher than the value of my car,
> and never mind that the insurance company would greatly prefer
> to replace my car, rather than compensate a family for the loss
> of their child.
> I state that you cannot determine the likelihood of accidents
> from any GPS record. I think the whole thing is almost a joke.
measure you have something to refer to.
Ins. companies traditionally only had your claims record. (low rate,
coarse data, a lot of subjectiveness).
Then (in most North American states/provinces) the ins. companies would
also get your moving infractions. More pertinent data (moving
infractions usually are non-safe-practice indicators).
So, with GPS recordings, they will be getting even more detailed
"behaviour" that indicates riskiness.
>>> They will probably find some correlations, but there will always
>>> be drivers whose GPS records indicate that they are poor
>>> drivers, and yet they have no accidents.
>
>> Individually it is often unfair: the well behaved unmarried 21 year old
>> male driver pays high ins. costs due to his peers.
>>
>> Collectively, it drives down the cost for "mature" drivers.
>
> I don't see how you come to this conclusion.
companies tell you specifically what drives their price. Any 20 year
old who buys a car is told by the ins. salesperson specifically all the
reasons why their rate is so high... Indeed my full coverage insurance
is 1/5 of my son's full coverage insurance for his car. (He's since
dropped to liability only - meaning damage to his car would not be
covered - that makes him a much more careful driver!).
Vincent van der Laan
Feb 13, 2012, 5:34:05 AM2/13/12
to
> Similar policies have been issued before. Another story:
>
> http://www.bbc.co.uk/news/technology-16969509
>
> "The AA is set to launch a new insurance policy which uses sat-nav
> technology to track driver performance.
Ridiculous... what about all sports car drivers?
>
> http://www.bbc.co.uk/news/technology-16969509
>
> "The AA is set to launch a new insurance policy which uses sat-nav
> technology to track driver performance.
They accelerate fast, corner fast and drive over 200km/h in Germany.
The system can't detect if I'm driving 240 km/h on an empty German Autobahn
which perfectly safe or on a severely crowded one where it would be quite
dangerous.
It can't detect if my emergency braking and swerving was for a kid that just
appeared from between two parked cars and started crossing the road without
watching out. Even at 30 km/h this could be dangerous.
A soon as my insurance company comes up with BS like this I will choose
another.
Alan Browne
Feb 13, 2012, 9:34:58 AM2/13/12
to
On 2012-02-13 05:34 , Vincent van der Laan wrote:
>> Similar policies have been issued before. Another story:
>>
>> http://www.bbc.co.uk/news/technology-16969509
>>
>> "The AA is set to launch a new insurance policy which uses sat-nav
>> technology to track driver performance.
> Ridiculous... what about all sports car drivers?
> They accelerate fast, corner fast and drive over 200km/h in Germany.
Which, by any definition, would be considered riskier driving than in
>> Similar policies have been issued before. Another story:
>>
>> http://www.bbc.co.uk/news/technology-16969509
>>
>> "The AA is set to launch a new insurance policy which uses sat-nav
>> technology to track driver performance.
> Ridiculous... what about all sports car drivers?
> They accelerate fast, corner fast and drive over 200km/h in Germany.
speed limited countries.
For example, France which has autoroute speed limits of, IIRC 130 km/hr,
has drastically limited accidents by installing a lot of radar-cameras
and fining heavily for speeding.
> The system can't detect if I'm driving 240 km/h on an empty German Autobahn
> which perfectly safe or on a severely crowded one where it would be quite
> dangerous.
account for the more liberal laws. How much do you pay per year to
insure a new €30,000 automobile even if you have a good record? (In
Canada, a "good record" would be no claims in 5 years).
In my experience it is impossible to drive over 80 - 120 km/h on the
autobahn when it is "severely crowded" so I doubt it will be an issue
except with motorcycles.
> It can't detect if my emergency braking and swerving was for a kid that just
> appeared from between two parked cars and started crossing the road without
> watching out. Even at 30 km/h this could be dangerous.
individual manoeuvre cases should disappear into the statistical noise.
> A soon as my insurance company comes up with BS like this I will choose
> another.
inspection at a higher price and others will choose the lower price at
the expense of tighter inspection.
Hans-Georg Michna
Feb 13, 2012, 11:23:31 AM2/13/12
to
On Sun, 12 Feb 2012 18:05:24 -0500, Alan Browne wrote:
>The dynamics of the car represent the commands of the driver. High rate
>recording most accurately shows the dynamics of the vehicle.
But it is still meaningless when it comes to predict the
>The dynamics of the car represent the commands of the driver. High rate
>recording most accurately shows the dynamics of the vehicle.
likelihood of accidents.
>If you enter a curve and accelerate out, normally, this is revealed in
>the data. If you accelerate aggressively, this is also shown.
accidents.
>If you jack rabbit off of a stop...
>
>If you do not make complete stops at a stop sign...
>
>If you drive significantly above the speed limit in a residential area...
>
>If you consistently drive above the limit on the highway...
>
>If you always make sharp/hard stops...
>
>And so on.
likelihood of accidents. But it seems we will be unable to
decide this question here.
It would be good to have some good data analysis from an
insurance company that uses GPS tracks. It seems to me that this
may well be a technophile toy project, because it sounds so
impressive to the layman.
Hans-Georg
Mike Coon
Feb 13, 2012, 12:23:25 PM2/13/12
to
Hans-Georg Michna wrote:
> And is meaningless again. Aggressive acceleration does not cause
> accidents.
I bet there's a correlation, though, just as there is with following too
> And is meaningless again. Aggressive acceleration does not cause
> accidents.
close behind..
Mike.
--
If reply address is Mike@@mjcoon.+.com (invalid), remove spurious "@"
and substitute "plus" for +.
Alan Browne
Feb 13, 2012, 12:41:58 PM2/13/12
to
On 2012-02-13 11:23 , Hans-Georg Michna wrote:
> On Sun, 12 Feb 2012 18:05:24 -0500, Alan Browne wrote:
>
>> The dynamics of the car represent the commands of the driver. High rate
>> recording most accurately shows the dynamics of the vehicle.
>
> But it is still meaningless when it comes to predict the
> likelihood of accidents.
>
>> If you enter a curve and accelerate out, normally, this is revealed in
>> the data. If you accelerate aggressively, this is also shown.
>
> And is meaningless again. Aggressive acceleration does not cause
> accidents.
It's not about individual events. It's about the collection of
> On Sun, 12 Feb 2012 18:05:24 -0500, Alan Browne wrote:
>
>> The dynamics of the car represent the commands of the driver. High rate
>> recording most accurately shows the dynamics of the vehicle.
>
> But it is still meaningless when it comes to predict the
> likelihood of accidents.
>
>> If you enter a curve and accelerate out, normally, this is revealed in
>> the data. If you accelerate aggressively, this is also shown.
>
> And is meaningless again. Aggressive acceleration does not cause
> accidents.
behaviours repeated often.
Again, and read carefully: insurance companies look at _behaviour_ and
_claims_. When they see an accident report, they know what it really means.
The basic fact is people can drive carelessly, recklessly and
dangerously for a long time before an accident occurs. But the more
careless, reckless and dangerous, the more grave the outcome. Insurance
companies know more about bad drivers than anyone else.
And if you believe your particular skill set is above that of most
people then you represent the great majority of people who believe they
are better than average drivers. Surely half of them are wrong.
Becoming a pilot and flight instructor had the co-benefit in me becoming
very critical about my driving skills and habits. And even as I spot
skill shortcomings and bad habits in my own driving I know I don't spot
them all. Further, when people point out a mistake I make driving I
accept it and try to improve in that respect. Even if they are wrong
about it, it draws attention to an area to check and improve. No
downside. Many drivers (most?) hate it when others note a shortcoming
in their driving. I welcome it.
There is no pride in safety. And pride seems to be what irks most
drivers where safety is concerned.
In the end, my safety and life are FAR MORE IMPORTANT than anyone's
thrills or ego on the road.
>> If you jack rabbit off of a stop...
>>
>> If you do not make complete stops at a stop sign...
>>
>> If you drive significantly above the speed limit in a residential area...
>>
>> If you consistently drive above the limit on the highway...
>>
>> If you always make sharp/hard stops...
>>
>> And so on.
>
> All of these are highly arguable when it comes to predicting the
> likelihood of accidents. But it seems we will be unable to
> decide this question here.
Collectively and repeatedly. Yes.
Aggressive driving (above) is a sign of poor driver attitude. The fact
is that the roads in most places are too crowded to allow for some to
drive aggressively and the rest defensively.
> It would be good to have some good data analysis from an
> insurance company that uses GPS tracks. It seems to me that this
> may well be a technophile toy project, because it sounds so
> impressive to the layman.
offering lower rates in exchange for such a service then they would be
best served by having a smart means of analysis. Otherwise they will
either lose customers due to rising rates on poor analysis of behaviour
or they will lose money on claims to seemingly safe drivers who are not.
Robert Peirce
Feb 13, 2012, 1:14:08 PM2/13/12
to
In article <f9SdnQgVfun71qTS...@giganews.com>,
things that I knew they were very capable of doing that would have been
considered very aggressive on the part of the average driver.
Even in normal situations, consider a three lane highway where you are
in the middle between two trucks and the left truck starts to move over
on you. Some drivers would accelerate. Others would brake. Some would
just panic.
Alan Browne <alan....@FreelunchVideotron.ca> wrote:
> Aggressive driving (above) is a sign of poor driver attitude. The fact
> is that the roads in most places are too crowded to allow for some to
> drive aggressively and the rest defensively.
Sometimes, the best defense is a good offense. I have seen people do
> Aggressive driving (above) is a sign of poor driver attitude. The fact
> is that the roads in most places are too crowded to allow for some to
> drive aggressively and the rest defensively.
things that I knew they were very capable of doing that would have been
considered very aggressive on the part of the average driver.
Even in normal situations, consider a three lane highway where you are
in the middle between two trucks and the left truck starts to move over
on you. Some drivers would accelerate. Others would brake. Some would
just panic.
Alan Browne
Feb 13, 2012, 1:23:57 PM2/13/12
to
drivers with experience: when making a change to the left lane (in
"right" lane countries) that we didn't spot a car or more likely a
motorcycle. When I hear a horn in that situation I look right and
simply back into my lane and wave "sorry". This happens about once
every couple years that I don't spot the car in the blind spot.
Truckers are usually very attuned to horns and will back off.
[2] The high end automobile industry has spent a lot of "fear of death"
advertising on the notion that acceleration is a defense. Indeed you
can cherry pick to the few cases where it is valid.
In the end damage increases as a square of the speed in an emergency.
Hitting the brakes probably gives you more dx/dt than hitting the gas
pedal in all but a very narrow range of cases.
If you prefer the automobile vendor's power fantasy then good luck to you.
Peter H. Coffin
Feb 13, 2012, 2:34:52 PM2/13/12
to
On Mon, 13 Feb 2012 11:34:05 +0100, Vincent van der Laan wrote:
>> Similar policies have been issued before. Another story:
>>
>> http://www.bbc.co.uk/news/technology-16969509
>>
>> "The AA is set to launch a new insurance policy which uses sat-nav
>> technology to track driver performance.
>
> Ridiculous... what about all sports car drivers? They accelerate fast,
> corner fast and drive over 200km/h in Germany.
>
> The system can't detect if I'm driving 240 km/h on an empty German
> Autobahn which perfectly safe or on a severely crowded one where it
> would be quite dangerous.
Wait... You're telling us, in a GPS newsgroup, that a gizmo plugged into a
>> Similar policies have been issued before. Another story:
>>
>> http://www.bbc.co.uk/news/technology-16969509
>>
>> "The AA is set to launch a new insurance policy which uses sat-nav
>> technology to track driver performance.
>
> Ridiculous... what about all sports car drivers? They accelerate fast,
> corner fast and drive over 200km/h in Germany.
>
> The system can't detect if I'm driving 240 km/h on an empty German
> Autobahn which perfectly safe or on a severely crowded one where it
> would be quite dangerous.
car couldn't possibly know what road the car's driving on?
--
"Remember that after Heracles cleaned the Augean Stables, he killed
the man who asked him to do it."
-- Robert Townsend, Up The Organization.
Vincent van der Laan
Feb 13, 2012, 4:52:26 PM2/13/12
to
On 13-02-12 20:34, "Peter H. Coffin" wrote:
> Wait... You're telling us, in a GPS newsgroup, that a gizmo plugged into a
> car couldn't possibly know what road the car's driving on?
Which road... yes obviously.
> Wait... You're telling us, in a GPS newsgroup, that a gizmo plugged into a
> car couldn't possibly know what road the car's driving on?
Local traffic density, I'm not so sure...
But it could be done I suppose:
TomTom does a good job collecting traffic jam data by tracking (anonimised)
cellphones in Holland...
Peter H. Coffin
Feb 13, 2012, 6:17:03 PM2/13/12
to
of the things on the same stretch of road. It only becomes important
after the fact, after all.
(You want to REALLY know how fast traffic is flowing on a particular
road? Ask a cell phone carrier. They know better than the transportation
authorities that have the traffic cams, how fast cell phones are
switching in and out of cell tower ranges....)
--
26. No matter how attractive certain members of the rebellion are,
there is probably someone just as attractive who is not desperate to
kill me. Therefore, I will think twice before ordering a prisoner
sent to my bedchamber. --Peter Anspach's Evil Overlord list
Alan Browne
Feb 13, 2012, 7:35:07 PM2/13/12
to
On 2012-02-13 18:17 , Peter H. Coffin wrote:
> On Mon, 13 Feb 2012 22:52:26 +0100, Vincent van der Laan wrote:
>> On 13-02-12 20:34, "Peter H. Coffin" wrote:
>>
>>> Wait... You're telling us, in a GPS newsgroup, that a gizmo plugged into a
>>> car couldn't possibly know what road the car's driving on?
>> Which road... yes obviously.
>>
>> Local traffic density, I'm not so sure...
>>
>> But it could be done I suppose:
>> TomTom does a good job collecting traffic jam data by tracking (anonimised)
>> cellphones in Holland...
>
> Traffic density is pretty easy to impute if you have more than a couple
> of the things on the same stretch of road. It only becomes important
> after the fact, after all.
>
> (You want to REALLY know how fast traffic is flowing on a particular
> road? Ask a cell phone carrier. They know better than the transportation
> authorities that have the traffic cams, how fast cell phones are
> switching in and out of cell tower ranges....)
Cell phone data is the basis of services such as Google Maps Trafic.
> On Mon, 13 Feb 2012 22:52:26 +0100, Vincent van der Laan wrote:
>> On 13-02-12 20:34, "Peter H. Coffin" wrote:
>>
>>> Wait... You're telling us, in a GPS newsgroup, that a gizmo plugged into a
>>> car couldn't possibly know what road the car's driving on?
>> Which road... yes obviously.
>>
>> Local traffic density, I'm not so sure...
>>
>> But it could be done I suppose:
>> TomTom does a good job collecting traffic jam data by tracking (anonimised)
>> cellphones in Holland...
>
> Traffic density is pretty easy to impute if you have more than a couple
> of the things on the same stretch of road. It only becomes important
> after the fact, after all.
>
> (You want to REALLY know how fast traffic is flowing on a particular
> road? Ask a cell phone carrier. They know better than the transportation
> authorities that have the traffic cams, how fast cell phones are
> switching in and out of cell tower ranges....)
Hans-Georg Michna
Feb 14, 2012, 11:12:12 AM2/14/12
to
On Mon, 13 Feb 2012 17:23:25 -0000, "Mike Coon"
<Mike@@mjcoon.+.com> wrote:
>Hans-Georg Michna wrote:
>> And is meaningless again. Aggressive acceleration does not cause
>> accidents.
>I bet there's a correlation, though, just as there is with following too
>close behind..
I wouldn't be sure. Can't remember ever having heard of an
<Mike@@mjcoon.+.com> wrote:
>Hans-Georg Michna wrote:
>> And is meaningless again. Aggressive acceleration does not cause
>> accidents.
>I bet there's a correlation, though, just as there is with following too
>close behind..
accident where a car aggressively accelerated into an obstacle.
And the "following too close" can obviously not be measured from
a single GPS track.
Hans-Georg
Hans-Georg Michna
Feb 14, 2012, 11:16:51 AM2/14/12
to
On Mon, 13 Feb 2012 12:41:58 -0500, Alan Browne wrote:
>Becoming a pilot and flight instructor had the co-benefit in me becoming
>very critical about my driving skills and habits. And even as I spot
>skill shortcomings and bad habits in my own driving I know I don't spot
>them all. Further, when people point out a mistake I make driving I
>accept it and try to improve in that respect. Even if they are wrong
>about it, it draws attention to an area to check and improve. No
>downside. Many drivers (most?) hate it when others note a shortcoming
>in their driving. I welcome it.
>
>There is no pride in safety. And pride seems to be what irks most
>drivers where safety is concerned.
>
>In the end, my safety and life are FAR MORE IMPORTANT than anyone's
>thrills or ego on the road.
Very true. I could have said most of this. (I'm a pilot too, but
>Becoming a pilot and flight instructor had the co-benefit in me becoming
>very critical about my driving skills and habits. And even as I spot
>skill shortcomings and bad habits in my own driving I know I don't spot
>them all. Further, when people point out a mistake I make driving I
>accept it and try to improve in that respect. Even if they are wrong
>about it, it draws attention to an area to check and improve. No
>downside. Many drivers (most?) hate it when others note a shortcoming
>in their driving. I welcome it.
>
>There is no pride in safety. And pride seems to be what irks most
>drivers where safety is concerned.
>
>In the end, my safety and life are FAR MORE IMPORTANT than anyone's
>thrills or ego on the road.
not a flight instructor.)
But I'm proud of being totally accident-free throughout my life
(except for one case in which another car drove into mine while
mine was standing still, something which I could not prevent).
Hans-Georg
Mike Coon
Feb 14, 2012, 11:24:54 AM2/14/12
to
drivers having the equivalent of the lorry-driver's tachygraph cuts down
accidents, I'm all for it. If the motivation is lower insurance premiums,
that's OK with me too.
Alan Browne
Feb 14, 2012, 5:02:22 PM2/14/12
to
On 2012-02-14 11:12 , Hans-Georg Michna wrote:
> On Mon, 13 Feb 2012 17:23:25 -0000, "Mike Coon"
> <Mike@@mjcoon.+.com> wrote:
>
>> Hans-Georg Michna wrote:
>
>>> And is meaningless again. Aggressive acceleration does not cause
>>> accidents.
>
>> I bet there's a correlation, though, just as there is with following too
>> close behind..
>
> I wouldn't be sure. Can't remember ever having heard of an
> accident where a car aggressively accelerated into an obstacle.
Ins. companies know one thing for sure: aggressive driving increases the
> On Mon, 13 Feb 2012 17:23:25 -0000, "Mike Coon"
> <Mike@@mjcoon.+.com> wrote:
>
>> Hans-Georg Michna wrote:
>
>>> And is meaningless again. Aggressive acceleration does not cause
>>> accidents.
>
>> I bet there's a correlation, though, just as there is with following too
>> close behind..
>
> I wouldn't be sure. Can't remember ever having heard of an
> accident where a car aggressively accelerated into an obstacle.
likelihood of accidents for that driver.
From the CAA website:
QUOTE
Are you a high risk?
Speeding, at fault accidents and DUI's will make anyone a high-risk
driver in the eyes of an insurance provider, but there are lesser-known
factors that can also contribute to this classification. These include
having numerous moving violations, unpaid tickets, driving a sports car,
being male, not having continuous coverage, and even having a poor
credit rating.
Dangerous Realities
In 2006, over 2 million Canadian displayed aggressive driving behaviors,
while 2.7 million admitted to frequently driving well over the speed
limit. Even more shocking, 2 million admitted to speeding up in order to
get through a red light, while 670,000 reported that they liked to take
risks just for the fun of it¹.
/QUOTE
claudegps
Feb 15, 2012, 5:47:07 PM2/15/12
to
Il giorno venerdì 10 febbraio 2012 17:41:42 UTC+1, Alan Browne ha scritto:
> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
> > I predict that the institutions involved in this will not
> > publish how exactly they determine what is good or bad driving
> > from GPS data records.
>
> Actually they do. They're goal is not to game the system but to pay out
> less claims. So in installing such devices they benefit by telling you
> exactly what they consider risky driving and what they will look for in
> the record. If you don't drive "that way" they won't find it in the record.
Their goal is also to avoid frauds, or overpaying for accidents.
> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
> > I predict that the institutions involved in this will not
> > publish how exactly they determine what is good or bad driving
> > from GPS data records.
>
> Actually they do. They're goal is not to game the system but to pay out
> less claims. So in installing such devices they benefit by telling you
> exactly what they consider risky driving and what they will look for in
> the record. If you don't drive "that way" they won't find it in the record.
Anyway, it's clear nor I could not find any information of what they are defining for "bad driving". I just found some figures about speed limits (that is easy), but for sharp acceleration/deceleration there is no clear definition.
(I tried to summaryze some finding about new and old insurances related GPS here: http://gnss-info.blogspot.com/2012/02/scatola-nera-assicurazioni-e-gps.html unfortunately it's for an italian audience...)
> > I have severe doubts that they can actually predict the
> > likelihood of traffic accidents from GPS records. The
>
> They're not looking for likelihood of accidents, they're looking for
> irresponsible behavior which increases the overall likelihood of accidents.
>
> > information is insufficient for the purpose.
>
> Depends on recording rate. For these purposes I would hope it is on the
> order of 5 Hz or more. (note that an intelligent recorder can record
> "special events" at a higher/detailed rate and ho-hum - down the highway
> at a lower rate - or not at all).
But the accelerometer is recorded only for few seconds before/after a crash.
The GPS is sampled at 5Hz, but usually not all the data is trasnferred: the device autonomously evaluates if a certain manouver is over the "limits" and send a notification to the server.
For tracking, a lower rate is used (1Hz or less) to avoid sending too much data.
Alan Browne
Feb 15, 2012, 6:24:55 PM2/15/12
to
On 2012-02-15 17:47 , claudegps wrote:
> Il giorno venerdì 10 febbraio 2012 17:41:42 UTC+1, Alan Browne ha scritto:
>> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
>>> I predict that the institutions involved in this will not
>>> publish how exactly they determine what is good or bad driving
>>> from GPS data records.
>>
>> Actually they do. They're goal is not to game the system but to pay out
>> less claims. So in installing such devices they benefit by telling you
>> exactly what they consider risky driving and what they will look for in
>> the record. If you don't drive "that way" they won't find it in the record.
>
> Their goal is also to avoid frauds, or overpaying for accidents.
> Anyway, it's clear nor I could not find any information of what they are defining for "bad driving". I just found some figures about speed limits (that is easy), but for sharp acceleration/deceleration there is no clear definition.
>
> (I tried to summaryze some finding about new and old insurances related GPS here: http://gnss-info.blogspot.com/2012/02/scatola-nera-assicurazioni-e-gps.html unfortunately it's for an italian audience...)
Traditional insurance terms won't go there. But as I've mentioned
> Il giorno venerdì 10 febbraio 2012 17:41:42 UTC+1, Alan Browne ha scritto:
>> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
>>> I predict that the institutions involved in this will not
>>> publish how exactly they determine what is good or bad driving
>>> from GPS data records.
>>
>> Actually they do. They're goal is not to game the system but to pay out
>> less claims. So in installing such devices they benefit by telling you
>> exactly what they consider risky driving and what they will look for in
>> the record. If you don't drive "that way" they won't find it in the record.
>
> Their goal is also to avoid frauds, or overpaying for accidents.
> Anyway, it's clear nor I could not find any information of what they are defining for "bad driving". I just found some figures about speed limits (that is easy), but for sharp acceleration/deceleration there is no clear definition.
>
> (I tried to summaryze some finding about new and old insurances related GPS here: http://gnss-info.blogspot.com/2012/02/scatola-nera-assicurazioni-e-gps.html unfortunately it's for an italian audience...)
before, this seems to be an effort at identifying behaviour that may be
correlated to riskier driving.
>
>>> I have severe doubts that they can actually predict the
>>> likelihood of traffic accidents from GPS records. The
>>
>> They're not looking for likelihood of accidents, they're looking for
>> irresponsible behavior which increases the overall likelihood of accidents.
>>
>>> information is insufficient for the purpose.
>>
>> Depends on recording rate. For these purposes I would hope it is on the
>> order of 5 Hz or more. (note that an intelligent recorder can record
>> "special events" at a higher/detailed rate and ho-hum - down the highway
>> at a lower rate - or not at all).
>
> Usually they record 5Hz GPS measurements and 100Hz acceleromenter measurements.
> But the accelerometer is recorded only for few seconds before/after a crash.
> The GPS is sampled at 5Hz, but usually not all the data is trasnferred: the device autonomously evaluates if a certain manouver is over the "limits" and send a notification to the server.
> For tracking, a lower rate is used (1Hz or less) to avoid sending too much data.
record in a way that is useful to insurance companies. Thirdly, as I've
also mentioned this can be done intelligently by sampling at high rate
but only storing what is needed to illustrate high dynamic events.
claudegps
Feb 16, 2012, 8:32:49 AM2/16/12
to
On 16 Feb, 00:24, Alan Browne <alan.bro...@FreelunchVideotron.ca>
wrote:
to riskier driving"
The insurance company does not says it... as far as I know, they don't
tell.
They don't tell also how long they can keep the records, if you can
them to review, if can ask their deletion after the insurance as
expired and so on...
Too few informations to the user (IMHO)
[cut]
used to monitor crashes and driver behaviour.
Anyway, the problem is to understand what a "high dynamic event" is.
Insurance company does not disclose how they evalute.
Nor they say that you can have the evidence of the behaviour. You just
get the result of their evaluation
wrote:
> On 2012-02-15 17:47 , claudegps wrote:
>
> > Il giorno venerdì 10 febbraio 2012 17:41:42 UTC+1, Alan Browne ha scritto:
> >> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
> >>> I predict that the institutions involved in this will not
> >>> publish how exactly they determine what is good or bad driving
> >>> from GPS data records.
>
> >> Actually they do. They're goal is not to game the system but to pay out
> >> less claims. So in installing such devices they benefit by telling you
> >> exactly what they consider risky driving and what they will look for in
> >> the record. If you don't drive "that way" they won't find it in the record.
>
> > Their goal is also to avoid frauds, or overpaying for accidents.
> > Anyway, it's clear nor I could not find any information of what they are defining for "bad driving". I just found some figures about speed limits (that is easy), but for sharp acceleration/deceleration there is no clear definition.
>
> > (I tried to summaryze some finding about new and old insurances related GPS here:http://gnss-info.blogspot.com/2012/02/scatola-nera-assicurazioni-e-gp... unfortunately it's for an italian audience...)
>
> > Il giorno venerdì 10 febbraio 2012 17:41:42 UTC+1, Alan Browne ha scritto:
> >> On 2012-02-10 11:24 , Hans-Georg Michna wrote:
> >>> I predict that the institutions involved in this will not
> >>> publish how exactly they determine what is good or bad driving
> >>> from GPS data records.
>
> >> Actually they do. They're goal is not to game the system but to pay out
> >> less claims. So in installing such devices they benefit by telling you
> >> exactly what they consider risky driving and what they will look for in
> >> the record. If you don't drive "that way" they won't find it in the record.
>
> > Their goal is also to avoid frauds, or overpaying for accidents.
> > Anyway, it's clear nor I could not find any information of what they are defining for "bad driving". I just found some figures about speed limits (that is easy), but for sharp acceleration/deceleration there is no clear definition.
>
>
> Traditional insurance terms won't go there. But as I've mentioned
> before, this seems to be an effort at identifying behaviour that may be
> correlated to riskier driving.
It is, but it's unclear which behaviours are identified as "correlated
> Traditional insurance terms won't go there. But as I've mentioned
> before, this seems to be an effort at identifying behaviour that may be
> correlated to riskier driving.
to riskier driving"
The insurance company does not says it... as far as I know, they don't
tell.
They don't tell also how long they can keep the records, if you can
them to review, if can ask their deletion after the insurance as
expired and so on...
Too few informations to the user (IMHO)
[cut]
> >> Depends on recording rate. For these purposes I would hope it is on the
> >> order of 5 Hz or more. (note that an intelligent recorder can record
> >> "special events" at a higher/detailed rate and ho-hum - down the highway
> >> at a lower rate - or not at all).
>
> > Usually they record 5Hz GPS measurements and 100Hz acceleromenter measurements.
> > But the accelerometer is recorded only for few seconds before/after a crash.
> > The GPS is sampled at 5Hz, but usually not all the data is trasnferred: the device autonomously evaluates if a certain manouver is over the "limits" and send a notification to the server.
> > For tracking, a lower rate is used (1Hz or less) to avoid sending too much data.
>
> First off this not a crash recorder. Secondly, this device has to
> record in a way that is useful to insurance companies. Thirdly, as I've
> also mentioned this can be done intelligently by sampling at high rate
> but only storing what is needed to illustrate high dynamic events.
The one I described, is actually the behaviour of a "insurance box"
> >> order of 5 Hz or more. (note that an intelligent recorder can record
> >> "special events" at a higher/detailed rate and ho-hum - down the highway
> >> at a lower rate - or not at all).
>
> > Usually they record 5Hz GPS measurements and 100Hz acceleromenter measurements.
> > But the accelerometer is recorded only for few seconds before/after a crash.
> > The GPS is sampled at 5Hz, but usually not all the data is trasnferred: the device autonomously evaluates if a certain manouver is over the "limits" and send a notification to the server.
> > For tracking, a lower rate is used (1Hz or less) to avoid sending too much data.
>
> First off this not a crash recorder. Secondly, this device has to
> record in a way that is useful to insurance companies. Thirdly, as I've
> also mentioned this can be done intelligently by sampling at high rate
> but only storing what is needed to illustrate high dynamic events.
used to monitor crashes and driver behaviour.
Anyway, the problem is to understand what a "high dynamic event" is.
Insurance company does not disclose how they evalute.
Nor they say that you can have the evidence of the behaviour. You just
get the result of their evaluation
Alan Browne
Feb 16, 2012, 9:16:12 AM2/16/12
to
>
> [cut]
>>>> Depends on recording rate. For these purposes I would hope it is on the
>>>> order of 5 Hz or more. (note that an intelligent recorder can record
>>>> "special events" at a higher/detailed rate and ho-hum - down the highway
>>>> at a lower rate - or not at all).
>>
>>> Usually they record 5Hz GPS measurements and 100Hz acceleromenter measurements.
>>> But the accelerometer is recorded only for few seconds before/after a crash.
>>> The GPS is sampled at 5Hz, but usually not all the data is trasnferred: the device autonomously evaluates if a certain manouver is over the "limits" and send a notification to the server.
>>> For tracking, a lower rate is used (1Hz or less) to avoid sending too much data.
>>
>> First off this not a crash recorder. Secondly, this device has to
>> record in a way that is useful to insurance companies. Thirdly, as I've
>> also mentioned this can be done intelligently by sampling at high rate
>> but only storing what is needed to illustrate high dynamic events.
>
> The one I described, is actually the behaviour of a "insurance box"
> used to monitor crashes and driver behaviour.
> Anyway, the problem is to understand what a "high dynamic event" is.
> Insurance company does not disclose how they evalute.
> Nor they say that you can have the evidence of the behaviour. You just
> get the result of their evaluation
movement of the car in detail. When going down the highway at steady
state, you don't need to record much. When turning a sharp corner, even
at low speed, it is a high dynamic as the car is changing direction
(heading) at a high rate. A high dyncamic does not mean bad driving.
But the ability to record detail of a high dynamic is needed to detect
poor driving habits.
How the insurance company _hopefully_ evaluates the data, and
accumulates a data base, would smooth out random events (a swerve to
avoid a dog, a rejected lane change, and so on). But someone who
continuously drives in a dangerous manner would stick out. Surely
someone who always drives 70 - 80 km/hr in a city posted for 50 km/hr
should stick out as not having a prudent approach to city driving.
* * *
I recall a test done in the US in the 90's. Without telling anyone, US
manufacturers installed voice recorders in all cars. They would store
the last 30 seconds of conversation in the car and freeze the recording
if the airbag deployed.
In 48 states, the recorded voice just before a crash was invariably:
"Oh SHIT!" or other expletive.
In Texas and Alabama the recorded voice was: "Hold my beer and watch this!"
claudegps
Feb 17, 2012, 6:23:07 AM2/17/12
to
[cut]
say what they do and how they evaluate "high dynamics"... I'm not sure
they are doing it rigth.
And, believe me, sometimes I heard about some requirements about how
to detect events.... and are totally non-sense!
> * * *
>
> I recall a test done in the US in the 90's. Without telling anyone, US
> manufacturers installed voice recorders in all cars. They would store
> the last 30 seconds of conversation in the car and freeze the recording
> if the airbag deployed.
>
> In 48 states, the recorded voice just before a crash was invariably:
> "Oh SHIT!" or other expletive.
>
> In Texas and Alabama the recorded voice was: "Hold my beer and watch this!"
LOL
It's a joke, isn't it?
> > The one I described, is actually the behaviour of a "insurance box"
> > used to monitor crashes and driver behaviour.
> > Anyway, the problem is to understand what a "high dynamic event" is.
> > Insurance company does not disclose how they evalute.
> > Nor they say that you can have the evidence of the behaviour. You just
> > get the result of their evaluation
>
> High dynamic event - the way I use it - is simply to allow recording the
> movement of the car in detail. When going down the highway at steady
> state, you don't need to record much. When turning a sharp corner, even
> at low speed, it is a high dynamic as the car is changing direction
> (heading) at a high rate. A high dyncamic does not mean bad driving.
> But the ability to record detail of a high dynamic is needed to detect
> poor driving habits.
>
> How the insurance company _hopefully_ evaluates the data, and
> accumulates a data base, would smooth out random events (a swerve to
> avoid a dog, a rejected lane change, and so on). But someone who
> continuously drives in a dangerous manner would stick out. Surely
> someone who always drives 70 - 80 km/hr in a city posted for 50 km/hr
> should stick out as not having a prudent approach to city driving.
Hopefully they will handle the data correctly, but since they don't
> > used to monitor crashes and driver behaviour.
> > Anyway, the problem is to understand what a "high dynamic event" is.
> > Insurance company does not disclose how they evalute.
> > Nor they say that you can have the evidence of the behaviour. You just
> > get the result of their evaluation
>
> High dynamic event - the way I use it - is simply to allow recording the
> movement of the car in detail. When going down the highway at steady
> state, you don't need to record much. When turning a sharp corner, even
> at low speed, it is a high dynamic as the car is changing direction
> (heading) at a high rate. A high dyncamic does not mean bad driving.
> But the ability to record detail of a high dynamic is needed to detect
> poor driving habits.
>
> How the insurance company _hopefully_ evaluates the data, and
> accumulates a data base, would smooth out random events (a swerve to
> avoid a dog, a rejected lane change, and so on). But someone who
> continuously drives in a dangerous manner would stick out. Surely
> someone who always drives 70 - 80 km/hr in a city posted for 50 km/hr
> should stick out as not having a prudent approach to city driving.
say what they do and how they evaluate "high dynamics"... I'm not sure
they are doing it rigth.
And, believe me, sometimes I heard about some requirements about how
to detect events.... and are totally non-sense!
> * * *
>
> I recall a test done in the US in the 90's. Without telling anyone, US
> manufacturers installed voice recorders in all cars. They would store
> the last 30 seconds of conversation in the car and freeze the recording
> if the airbag deployed.
>
> In 48 states, the recorded voice just before a crash was invariably:
> "Oh SHIT!" or other expletive.
>
> In Texas and Alabama the recorded voice was: "Hold my beer and watch this!"
It's a joke, isn't it?
Alan Browne
Feb 18, 2012, 12:55:02 PM2/18/12
to
On 2012-02-17 06:23 , claudegps wrote:
>> I recall a test done in the US in the 90's. Without telling anyone, US
>> manufacturers installed voice recorders in all cars. They would store
>> the last 30 seconds of conversation in the car and freeze the recording
>> if the airbag deployed.
>>
>> In 48 states, the recorded voice just before a crash was invariably:
>> "Oh SHIT!" or other expletive.
>>
>> In Texas and Alabama the recorded voice was: "Hold my beer and watch this!"
>
> LOL
> It's a joke, isn't it?
Of course. I get great laughs. Esp. in states surrounding the 2 above.
>> I recall a test done in the US in the 90's. Without telling anyone, US
>> manufacturers installed voice recorders in all cars. They would store
>> the last 30 seconds of conversation in the car and freeze the recording
>> if the airbag deployed.
>>
>> In 48 states, the recorded voice just before a crash was invariably:
>> "Oh SHIT!" or other expletive.
>>
>> In Texas and Alabama the recorded voice was: "Hold my beer and watch this!"
>
> LOL
> It's a joke, isn't it?
Alan Browne
Feb 18, 2012, 1:00:56 PM2/18/12
to
On 2012-02-17 06:23 , claudegps wrote:
>> How the insurance company _hopefully_ evaluates the data, and
>> accumulates a data base, would smooth out random events (a swerve to
>> avoid a dog, a rejected lane change, and so on). But someone who
>> continuously drives in a dangerous manner would stick out. Surely
>> someone who always drives 70 - 80 km/hr in a city posted for 50 km/hr
>> should stick out as not having a prudent approach to city driving.
>
> Hopefully they will handle the data correctly, but since they don't
> say what they do and how they evaluate "high dynamics"... I'm not sure
> they are doing it rigth.
> And, believe me, sometimes I heard about some requirements about how
> to detect events.... and are totally non-sense!
Ins. co's are generally good at statistical analysis. Hopefully they
>> accumulates a data base, would smooth out random events (a swerve to
>> avoid a dog, a rejected lane change, and so on). But someone who
>> continuously drives in a dangerous manner would stick out. Surely
>> someone who always drives 70 - 80 km/hr in a city posted for 50 km/hr
>> should stick out as not having a prudent approach to city driving.
>
> Hopefully they will handle the data correctly, but since they don't
> say what they do and how they evaluate "high dynamics"... I'm not sure
> they are doing it rigth.
> And, believe me, sometimes I heard about some requirements about how
> to detect events.... and are totally non-sense!
will come up with analysis that is appropriate and relevant w/o leading
to unwarranted bias.
The other day I went to the bank to make a deposit. I counted my bad
driving mistakes as 11 over a 10 km trip. Counting 11 means I probably
made 22. Some were deliberate (boulevard pass on right because the 3rd
lane converged into the middle lane at that point, was the worst).
Others were just plain bad driving of various sorts including a very
deep yellow light.
OTOH, I'm not sure at all that the ins. co. systems would have picked up
any of them other than driving 120 on a 100 km/hr highway.
0 new messages