Boston startups merge to create driver-tracking giant for insurance industry | Hacker News Reader

Boston startups merge to create driver-tracking giant for insurance industry

bostonglobe.com

16 pointsDeinos31 comments

speaking as someone with a working knowledge of the technology used to track drivers for insurance purposes (OBD2 and CANbus) Ive always been remarkably suspicious of this sort of black-box "we will save you money" proposition.

tapping into the CAN (car area network) system on a vehicle lets you see things like timing retardation and advancement, barometric pressure, o2 levels, and throttle body positions. it tells you what these systems are doing in realtime but it fails to tell you why these systems did these things. at best youll have to make a guess, and insurance companies seem to have an incentive to lean into the "because you drive poorly" excuse in order to minimize risk to investors.

theres no target and no set of parameters that define what an altruistically good driver is in the eyes of these companies, only the companies insistence it can "save you money" without explaining what you must do other than "be a good driver" and install their widget, which runs mystery code that interacts with one of the most critical parts of your vehicle that may itself not even be DOT approved..

As someone working in the industry I can tell you that these insurance devices use precisely none of the sensors you mentioned. They use speed, odometer, and plug/unplug from the ECU, but that's basically it. They often collect time, GPS, and accelerometer, but that's from chips in the device not the car.

Also the OBD protocol is standardized and the device basically operates in a read-only fashion (cannot write data to the ECU). The worst problem that actually occurs with these is increased rates of battery drain while the car is off (though most devices have shutoffs in place to prevent the battery from getting too low).

I touch an OBD2 port every single day.

I should have been more specific. They may not use this information for their product, but they are privy to it. theres also no transparency to say they ever delete that data.

OBD may be read only but CANbus pins exist in the OBD2 connector, and many CANbus implementations can be accessed through the OBD2. steering, brakes, and throttle can all be remotely controlled through the CANbus in most vehicles.

there has been at least one PoC that can detonate your airbags from the OBD. https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2017-1493...

Insurers don't really care why some variable is correlated with increased claims as long as it is. They just want the maximum data at their disposal so they don't get outpriced by their competitors.

Weird, how did I miss that apps like this even exist? Seems a little creepy to me, though maybe it cuts premiums for some people?

"Both companies’ core products are apps on smartphones that use the sensors in the phones to collect data about how people are driving, such as whether they are speeding, frequently braking hard, or picking up their phones to text when they should be paying attention to the road."

https://archive.is/u0Qpu#selection-1417.0-1424.0

As someone who used to spend a decent amount of time driving overloaded trucks the speeding thing always pisses me off because it's lacking in nuance.

There is nothing safe about following the speed limit when the speed limit is substantially lower than what the rest of the traffic wants to do.

In the UK black boxes are almost a necessity for 17/18 year olds to be able to afford insurance, even on the cheapest cars.

Making dense urban centers AND cars unaffordable for young adults seems to be a great long term plan.

I wonder if the UK is getting a net gain of young, highly motivated and educated professionals or a loss.

No clue how the two could possibly be connected, though I doubt it.

If you're in california, it's because california regulation is very specific in what can be used for car insurance rating and those apps are not one of them.

If the premiums approach being sufficiently fine grained as to account for individual risk then what's the purpose? Why not just squirrel the money away in some sort of HSA-esque fund and cut out the expensive middleman entity? If the state is gonna make me pay for mandatory coverage why not just do it with taxes and have one less set of predatory companies trying to screw me?

Even with perfect information about the past/present, there is still uncertainty which can be insured against. Current models are nowhere near precise enough, and probably will never get there (self-driving cars will replace the need for insurance at the individual driver level before we have perfect models of driving risk)

How could there be certainty? Even a good driver can skid on ice or get hit by an uninsured driver running a light.

Indeed, how could there? There is inherent uncertainty to the future which cannot be known, and thus there is risk which can be insured against

If internet comments are any indication you just adopt a circular definition of "good driver" where all those things disqualify one from being a good driver.

At least in Ohio, you can put some amount of money on deposit with the state as a substitute for buying auto insurance.

Some level of car insurance is mandated by law in every state. Might as well make it as cheap as possible for yourself.

Additionally, many people can barely afford their car. It's crazy how many people are driving around in a car worth more than their annual salary. $60,000 vehicles and such, especially trucks. It would take them a while to "self insure" against the risk of totalling that.

I was thinking more about the people driving $2k-$5k cars.

The last thing they need is their insurer jacking their rates because they're ending too many trips in bad zip codes.

You can "incentivize them onto public transit" all you want but most of those people would already be on public transit if there was even a $40/mo difference.

> The last thing they need is their insurer jacking their rates because they're ending too many trips in bad zip codes.

Wait, the 'data' is based on ending a trip in a bad zip code!?

Man, and I'm over here thinking they map all the drivers over time and place to see who is normally on the road with each other, then basing those factors on courteousness and abilities with ambiguity.

>Wait, the 'data' is based on ending a trip in a bad zip code!?

These companies use cell phone and/or OBD2 data. The only reason they haven't changed your rate based on the fact that you go to a strip club or shopping at Walmart is because there hasn't been a business justification for operating on that level of detail yet.

Insurers have long used proxies for wealth in order to set rates. If your zip code is the hood you're gonna pay premiums like everyone else in the hood. The only difference is that now they'll have more detail so the guy who registers his truck to his apartment in the hood but is on the road 5day/wk for work will pay based on two days of ending trips in the hood and five days of whatever motel he's in happens to be.

>Man, and I'm over here thinking they map all the drivers over time and place to see who is normally on the road with each other, then basing those factors on courteousness and abilities with ambiguity.

They might be doing that too if it happens to be worthwhile for their business.

I have already such a tracker in my car that connects to the OBD2 and has its own 3G or 4G chip. I have an app and can track where my car is and could do some stats. Why did I get it for free? It was a joint deal between the insurance and the car garage and they probably can see where I drive - the garage wanted me to do all maintenance in their garage and the sensor/app company offered me to be my car insurance. So also the dealer or garage will be interested to know where you do your car maintenance. All in all I should probably take it out but I see a little benefit and do not care about that amount of data that I share (any cellphone manufacturer will have a million more data points).

New Hampshire doesn’t require car insurance. https://www.thehartford.com/aarp/car-insurance/new-hampshire...

live free or die

There would be some good reasons, for example regulations could mandate that only factors that one can change and are most directly related to driving are valid for the rate setting. i.e. field of employment or income cannot be used but speeding, hard breaking, cell phone use while driving, and other unsafe behaviors would be.

>If the premiums approach being sufficiently fine grained as to account for individual risk then what's the purpose?

It doesn't matter if there is a purpose when so many states mandate that you buy the product anyway.

Together with mandatory insurance laws, is the government literally forcing us to let companies spy on us? Unless you're wealthy enough to afford the higher rates (for as long as those are offered).

Well, the citizenry could elect people to pass laws to forbid such actions.

Yeah I mean, take California for example: they basically prohibits insurance telematics completely (these apps cannot operate on CA customers)

Metromile also runs a Boston office. Are they a competitor or higher up on the stack?

There are a few differences: Metromile is an insurance company that uses devices which plug into cars to track mileage and charges by the mile. CMT/TrueMotion aren't insurers, they provide apps/SDKs to insurance companies which track driving behavior. This tracking is used to decide on how much to charge for insurance, but this is just a factor applied on top of someone's traditional premium, that is, CMT/TM can't really support pay-per-mile insurance

Tesla has been doing this since 2019.

Reply on news.ycombinator.com