[ rss / options / help ]
post ]
[ b / iq / g / zoo ] [ e / news / lab ] [ v / nom / pol / eco / emo / 101 / shed ]
[ art / A / beat / boo / com / fat / job / lit / map / mph / poof / £$€¥ / spo / uhu / uni / x / y ] [ * | sfw | o ]
logo
technology

Return ]

Posting mode: Reply
Reply ]
Subject   (reply to 24638)
Message
File  []
close
telematics-infographic-2.png
246382463824638
>> No. 24638 Anonymous
5th September 2015
Saturday 3:39 pm
24638 spacer
In order to knock about a hundred and fifty quid off my car insurance I opted to install a telematics box for 3 months.

Information is sent remotely via mobile network to their server. Using the information, they can rank your 'driving skills' against a benchmark or the other policyholders and determine if you deserve a bonus discount on your insurance. I don't think they can penalise you outright if they decide you are a bad driver.

Do any of you have any experience with these devices or the implications of having my driving logged by an insurance company?

I've got some pictures of the inside of the unit and have some idea about how it all works if anyone is interested.
Expand all images.
>> No. 24639 Anonymous
5th September 2015
Saturday 3:48 pm
24639 spacer
The implications are that eventually people who refuse to use these things will be forced to pay significantly higher premiums. The recording of your behavioural pattern will be accessible by both the law and corporations, and eventually everyone with an internet connection. It will not only record how "safely" you drive but your entire routine and whenever you speed, pretty much everything.
>> No. 24640 Anonymous
5th September 2015
Saturday 4:37 pm
24640 spacer
>>24639
I've gone over the material for the policy and there is worryingly little smallprint about what they do with the data. The only relevant information I can find is this excerpt on their website.

>We only share the data with those companies required to provide the services under the policy, such as the unit provider. Data isn't sold or shared with anyone else.
>> No. 24641 Anonymous
5th September 2015
Saturday 4:40 pm
24641 spacer
>>24640
It doesn't really matter who they intentionally share it with. There's laughably little legislation on how carefully companies have to protect your data. Even if there was, it would be stolen sooner or later anyway.
>> No. 24642 Anonymous
5th September 2015
Saturday 5:18 pm
24642 spacer
>>24640
You can bet your bottom dollar that if the police decide one day that they need information on your driving and whereabouts for an investigation, the providers will hand it over without a fuss, since they know that, just as for you, nothing that they say to the police could ever help them, and if it comes down to choosing between the police and you they know that the law effectively forces you to do business with them.
>> No. 24644 Anonymous
5th September 2015
Saturday 6:18 pm
24644 spacer
>>24642
Is this much more than what they could get from your mobile, though?
>> No. 24645 Anonymous
5th September 2015
Saturday 6:25 pm
24645 spacer
>>24644
Unless your phone is already tapped into the car's systems, yes, it's considerably more.
>> No. 24646 Anonymous
5th September 2015
Saturday 8:51 pm
24646 spacer
>>24638 I've got some pictures of the inside of the unit and have some idea about how it all works if anyone is interested.

Definitely interested. I've designed one of these for an insurer (last year), and felt somewhat conflicted.
Then again, I also built a widget that fucks bus drivers off too, by limiting their throttle action when the bus is empty, so I'm an equal opportunity bastard.
>> No. 24647 Anonymous
5th September 2015
Saturday 9:11 pm
24647 spacer
I have one on my car but they've hidden it so well I can't find it. Is it connected to the OBD2 line?

My guess is that it's connected to that, and has a small antenna and some onboard instruments like accelerometer? It then collect that along with GPS data and sends it over mobile networks?
>> No. 24649 Anonymous
5th September 2015
Saturday 9:25 pm
24649 spacer
>>24647
I'd always assumed all they need is a power input, so there would just be a 12v tap coming off somewhere. I can't see any need to connect to the OBD as they should get all the data they need from just GPS + an accelerometer.
>> No. 24650 Anonymous
5th September 2015
Saturday 10:09 pm
24650 spacer
>>24649
Presumably by tapping into the on-board sytems they get access to information such as the pressure on the pedals and the engine speed so they can tell the difference between someone revving hard in a lower gear and someone driving gently in a higher gear without having to guess from the data. GPS signal is also poor in heavily built-up areas or in tunnels.
>> No. 24651 Anonymous
5th September 2015
Saturday 10:11 pm
24651 spacer
Can you mess about with it so that it shows that you are a competent 80 year old driver?
>> No. 24652 Anonymous
5th September 2015
Saturday 10:18 pm
24652 spacer
>>24651
You could drive like a competent 80 year old driver, but then that would be no fun. Especially since you can't rely on "but I'm 80!" when you crash into something at 15mph.
>> No. 24653 Anonymous
5th September 2015
Saturday 10:25 pm
24653 spacer
>>24650
An intermittent GPS signal isn't too much of an issue when the insurer is going to collect months worth of data. And I can't see why an insurer would be that interested in how hard you're revving the engine, a simple accelerometer is far more useful.
>> No. 24654 Anonymous
5th September 2015
Saturday 11:04 pm
24654 spacer
>>24653
>And I can't see why an insurer would be that interested in how hard you're revving the engine
The modern car insurance industry is heavily data-driven. Any additional data point that either gives them a new dimension to rate on or lets them increase the precision of their pricing is an opportunity for them to make more money. An accelerometer gives them an indication of how aggressive your driving is. An accelerometer plus mechanical data gives them an even more accurate indication.
>> No. 24655 Anonymous
6th September 2015
Sunday 10:01 am
24655 spacer
There are a few cases already in the US of these companies having security vulnerabilities in these devices, so be careful. Most of them plug into the OBD port on your car to monitor it, a couple of them essentially make the OBD port gateway into Wifi and so you can see how this might end..
>> No. 24656 Anonymous
6th September 2015
Sunday 11:48 am
24656 spacer
>>24655
The OBD port on my car is exposed and usable, so if it is connected it'll be tapped. I'm unsure if the OBD2 protocol allows multiple connections, though.
>> No. 24657 Anonymous
6th September 2015
Sunday 5:46 pm
24657 spacer

PCB1.jpg
246572465724657
>>24646
This is one side of it.

uBlox is the GPS IC. I reckon the one in the lower mid is the processor. As far as I'm aware the only other sensor is a six axis gyroscope which might be one of the chips on this side.
>> No. 24658 Anonymous
6th September 2015
Sunday 5:51 pm
24658 spacer

PCB2.jpg
246582465824658
>>24646
I think the big ceramic / metal thing is the GPS antenna. The other big chip is GSM for the mobile network. I assume it needs a SIM card but maybe that is concealed.

It's powered by the 12 V socket on the car with a pretty bulky plug and regulator. No connection to the CAN bus or anything like that.
>> No. 24659 Anonymous
6th September 2015
Sunday 10:15 pm
24659 spacer
Interesting, thanks!
For what it's worth, the SIM will probably be the SO-8 under the GPRS module (although I've only sued them in DFN style not SOP. Availability is patchy as hell, though.
I'd imagine there's an accelerometer in there, rather than (or as well as) a gyro.
Is this a puck that velcros onto the dash or something? Tried it in a Renault with the fast defrost stuff?
>> No. 24660 Anonymous
6th September 2015
Sunday 11:37 pm
24660 spacer
Is there a shielding can under that stamped GPRS antenna?
Trying to work out where the code's running. The GE865 seems to have a Python interpreter, but I'm not sure that'll be fast enough.
It's also interesting that it's 2.5G - last I heard, networks were far from keen on taking on bulk M2M endpoints on anything less than 3G.
>> No. 24661 Anonymous
7th September 2015
Monday 3:09 am
24661 spacer
>>24660

>The GE865 seems to have a Python interpreter, but I'm not sure that'll be fast enough.

Easily. The GSM and GPS modules are doing all of the heavy lifting. I can't imagine they're transferring much data, purely to keep their costs down. If they're doing what I expect they're doing based on their publicity (logging peak acceleration events, time, speed and location, sending a GPRS packet every few minutes) then you could run the show on an 8 bit MCU.

IIRC the most valuable data is simply time of day, as most major accidents involving young drivers occur late at night.
>> No. 24662 Anonymous
7th September 2015
Monday 7:40 am
24662 spacer
The two crystals on the back of the board make me think there's a micro on the other side (so that probably is a shielding can).
Not a particularly huge micro, though - there's no via field for a BGA.
Pulling 10Hz data from accelerometer and GPS, fusing them to try to work out distance travelled and generate some metrics of driving style, then compressing the results, managing the GPRS , inevitable shonky OTA updater / autobricker, all starts to add up to some real compute. If you're just trying to phone in with a 'driving, not driving' message, less so.
>> No. 24663 Anonymous
7th September 2015
Monday 8:08 am
24663 spacer
>>24662
Mine has ratings for Speed, Time, Acceleration and Braking, but it also doesn't just sit in the 12V port.
>> No. 24664 Anonymous
7th September 2015
Monday 9:34 am
24664 spacer
>>24663
Does it report when (it thinks) you're speeding? That strikes me as quite a big back-end job, although if it can cache the limits on your most-driven (commute) roads, that'd reduce the load, I guess.
Have you tried changing the gain on the accelerometer, and seeing if it throws a fit about impossible accelerations, or if it says you're a very smooth driver now...
I can envisage a lot of people getting GPS jammers, and crapping up the GPS band with them, in the near future.
>> No. 24665 Anonymous
7th September 2015
Monday 1:48 pm
24665 spacer

2015-09-07_13h47_58.jpg
246652466524665
>>24664
I'm not OP, I'm the lad who can't even locate his box. I've traced the OBD2 back and there doesn't seem to be anything there.

I speed regularly but still get 100/100 for the speed rating, so I'm not sure exactly how it reports it. When I cancel this insurance, I'll rip the box off and have a look -- I'm an electronic engineer so hopefully I'll be able to deduce something. This won't be for a while though. As one of you two already said, to constantly poll and report a full dataset would use an incredible amount of data. I don't imagine there would be any way to detect if the driver is speeding on the chip -- having a fully fledged GPS would be costly, computationally expensive and unreliable, as I imagine all of the reports are checked against a constantly updated central map. To work out if the driver is speeding, I reckon it would be necessary to send back location and speed data at a fair rate, which I don't think is done. I do, however, get pretty much random acceleration and braking reports, and that is something that the accelerometer/gyro could log and send back in batches.

Pic attached is my scores for last month. I'll probably solve it as a set of simultaneous equations to figure out the weighting and what they value more.
>> No. 24666 Anonymous
7th September 2015
Monday 2:37 pm
24666 spacer
>>24665
Having solved using data from previous months with scores less than 100 for speed, too, my calculations spit out the following weightings:

21% speed
37% time
20% braking
23% acceleration

Interesting.
>> No. 24667 Anonymous
7th September 2015
Monday 3:19 pm
24667 spacer
>>24666
I'm curious about what why the time-of-day gets such a high weighting. Particularly given all the fuss certain organisations make about how inappropriate speed causes a third or accidents, or summat.
>> No. 24668 Anonymous
7th September 2015
Monday 3:57 pm
24668 spacer
>>24667
It looks like they simply can't detect 'speed' based on limits - maybe there's a trigger at 80+ or something, which would be a bit crap.
Expect more and more scoring based on what's easy to measure, rather than what's relevant, until the car manufacturers start joining in, exposing some of their cameras, pedal & wheel positions, stereo and central locking status.
I'm not looking forward to this future.
>> No. 24669 Anonymous
7th September 2015
Monday 4:06 pm
24669 spacer
>>24668
You may as well learn to like the future now, because it's coming whether you like it or not.

Danny Boyle on some film he made:

>“These films have to be made,” he said. “Benign as they may seem, they have created forces that are more powerful than governments and banks. And they don’t seem motivated by money. I find that extraordinary. It’s a paradigm shift we seem blissfully unaware of. They’re not interested in money but in data. Our data.”

We're now in an age where companies can rival banks inside of five years, rather than over the course of five centuries. You don't find that exciting? Don't you want to sit atop the pyramid and rule the technologically illiterate?
>> No. 24670 Anonymous
7th September 2015
Monday 4:30 pm
24670 spacer
>>24669 Don't you want to sit atop the pyramid and rule the technologically illiterate?

No. I don't really want to rule anyone, just have fun and play with interesting stuff. Preferably without having anyone rule me to any degree that interferes with that. Not going to happen, of course, but a chap can wish.
>> No. 24671 Anonymous
7th September 2015
Monday 5:06 pm
24671 spacer
>>24670
Well guess what, the people who are going to rule us feel exactly the same way. They just work harder than us so their geekery ends up having global ramifications. Technology is technology. Only the luddite fears it. The geek embraces it for what he believes is right.
>> No. 24672 Anonymous
7th September 2015
Monday 6:04 pm
24672 spacer
>>24671
>Only the luddite fears it.

It's worth remembering who the luddites actually were before using the term as a catch-all insult for those who are skeptical about the consequences of new technology and boosterism of progress at any cost.

It's also probably worth realising that if you're on this forum you are highly unlikely to be one of those who gets to "sit atop the pyramid", so assuming that unrestrained technology continues its path towards, as you see it, of empowering a few to rule without any reasonable balance the rest of humanity, where does that leave you?

Brb, off to smash a loom and live in a cave with no medicine, etc.
>> No. 24673 Anonymous
7th September 2015
Monday 6:06 pm
24673 spacer
>>24667

Amongst young drivers, time of day is a huge risk factor. When driving late at night, they're disproportionately likely to be drunk, on drugs or showing off to a car full of mates. The accidents that happen under these conditions tend to be major high-speed crashes that cause life changing injuries, rather than the more minor shunts that happen during peak hours.

http://www.brake.org.uk/news/15-facts-a-resources/facts/488-young-drivers-the-hard-facts
>> No. 24674 Anonymous
7th September 2015
Monday 6:07 pm
24674 spacer
>>24672

>It's also probably worth realising that if you're on this forum you are highly unlikely to be one of those who gets to "sit atop the pyramid"

I'll be fine, I'm an electronics engineer and I know how to be obsequious in Mandarin. Our overlords will need lackeys like me.
>> No. 24675 Anonymous
7th September 2015
Monday 6:34 pm
24675 spacer
>>24672
>It's worth remembering who the luddites actually were before using the term as a catch-all insult for those who are skeptical about the consequences of new technology and boosterism of progress at any cost.

Not really, since that's not what the use of the word refers to.
>> No. 24677 Anonymous
7th September 2015
Monday 7:07 pm
24677 spacer
>>24672
It's obvious nobody has ever called you a luddite before. Sheesh.

Do say hello to Mister Ludd for me when you see him.
>> No. 24678 Anonymous
8th September 2015
Tuesday 3:56 pm
24678 spacer
>>24674
Why do you even want to be a lackey?
>> No. 24679 Anonymous
8th September 2015
Tuesday 4:07 pm
24679 spacer
>>24678
Because he lacks a spine.
>> No. 24680 Anonymous
8th September 2015
Tuesday 4:19 pm
24680 spacer
>>24678

It's my best chance of avoiding the death camps.

Return ]
whiteline

Delete Post []
Password