I wish the journalist had asked more about this seemingly alarming statistic:
"Zendrive now has its monitoring technology on 60 million phones, roughly one of every four U.S. drivers"
For example:
How did this "monitoring technology" get on 60M phones? Do the owners know it's there? Is it hidden in other apps? Is there clear disclosure of its existence in those apps? What is done with this data? Who has access? Is it anonymized? Is it sold to third parties? Is it used for advertising or other non-safety-related purposes? How do they distinguish between drivers and passengers? What do they do about COPPA if it's on the phone of an under-13-year-old? Etc.
Indeed! Unless this is a part of Google Maps, Waze, Uber, or Lyft, some extremely popular app on our phones is doing location and activity monitoring no one was aware of. It's especially frightening because this data could be used to deny people reasonable insurance rates in the future and they weren't even aware they were being logged.
I'm curious how they know a person is definitely driving and not a passenger.
> It's especially frightening because this data could be used to deny people reasonable insurance rates in the future and they weren't even aware they were being logged.
More expensive insurance is getting off easy. After getting caught two or three times people should lose their license over it, IMO. It's as dangerous as drunk driving.
That said, it is concerning how they're collecting this information.
If I'm driving down the road with a phone held to my ear, a giant soda in the other hand, and steering with my knees, and I happen to pass a police officer, I hope they'd be allowed to stop me. I shouldn't be able to hide behind "I was in my private car, you had no right to look in on me and observe my behavior there."
If I'm driving down the road with a phone held to my ear, a giant soda in the other hand, and steering with my knees, and I happen to pass a police officer, I hope they'd be allowed to stop me
Some expectation of privacy probably applies to conversations inside the car. Also, there might be a different way to think about the entire record of behavior of where one drives. Neither of those would interfere with policing of the behavior you cite.
There should be a lot of signals to perceive where the user is sitting. If the phone is resting in the center console, the accelerometers can tell whether it is being lifted up to the right or to the left of the car whenever it is being used. It could tell whether it is resting in the center console or the left or right side of the car depending on the shape of arc, distance travelled, centrifugal forces whenever the car follows the road around a left vs. a right turn.
If they care about accuracy to put in the engineering investment, it could be right a very high percentage of the time.
My phone sometimes insists I'm driving on water. I don't believe you can tell the difference between drivers and passengers seats without in car sensors.
They can tell by which way you enter the car. This is problematic since women commonly put their purse in the passenger seat sometimes via the passenger door.
You can easily be off by meters or even 10s of meters in terms of the position of the phone.
You have no way to know the position of the car in relation.
You would need car sensors and per car interior data and even then it would likely confuse a driver holding a phone in their right hand and a passenger holding it in their left.
You talk about the phone keeping track of where the phone not only is but was historically in order to guess if you are driving and texting and I hear you want to design a broken system that is not even deterministically broken.
It also fails to account for people passing a phone around to someone else in the car.
I wouldn't be able to give my phone to a passenger because it would already have decided it is a a driver's phone.
Your users will only be able to use your phone if they can reverse engineer what their phone thinks they are doing.
Anyone who does not charge their phone will see car travel destroy their battery life.
One can trivially avoid such a feature by disabling GPS but most people won't have to worry about it because they just won't pay hundreds of dollars more for a phone that randomly doesn't work even in the passenger seat because it's incorrectly second guessing you.
Being able to detect which lane of the road you are on is still challenging enough for self-drive technology. A cheap phone sensor most likely doesn't have such spatial resolution.
There are probably simpler ways - for example, if an accident report states there was only one person in the car, it was probably the driver using his phone.
My wife and I have different phones and have consistently found the map experience better on mine. When we travel I drive and she uses my phone to navigate...
As a matter of practicality using this data adversely against customers is unlikely.
1. This is literally almost everyone on the road (that should frighten you more than insurance companies having this data)
2. Insurance is a highly regulated industry. Each state has a department of insurance that regulates how insurers apply rate and underwriting.
3. There are many voluntary opt-in programs out there that reward good behavior and marginally penalize bad behavior. Those programs focus on a few behaviors, none of which are distraction at the moment and they are all opt-in.
Full disclosure - I work at a large insurance company and am also an EFF donor. This is near and dear.
It is virtually guaranteed that this data will be used adversely against customers. Insurances want to find reasons why not to pay out, and will find those reasons in the data they collect.
Insurance may be "highly regulated" but regulation on what data can be collected is underdeveloped.
Helping them and everyone else find behavioral solutions to avoid using their phone while driving, for example rewarding good behavior with discount, gift cards, etc.
There are a number of insurance companies literally trying to help adjust behavior in order to save lives of those distracted and the innocent bystanders and other drivers on the road.
Honestly they could start by building voice controls that people want to use rather than the voice assistant bullshit. Idk what the google one is like, but siri is a toy.
You don't even need general purpose speech recognition to call someone on your mobile, just enough to recognize a trigger sound and the name of the person you want to call. This is how I could use voice recognition on my Nokia N73 to call anyone in my address book.
As I understood it the hard thing isn’t converting speech to text, it is understanding which text should lead to which actions.
Strangely enough in command line interfaces this works perfectly, so maybe we need just a more speech friendly way to call commands?
And why you’d need the cloud is beyond me. A speech assistent that fails once you don’t have a internet connection is not only annoying, in some cases it could become outright dangerous.
With a limited set of commands and fairly strict user training, you reduce this problem to "parsing a limited grammar" which is significantly easier. It's more or less what the current top-tier chatbots are doing. You don't need an outrageous amount of processing power for this.
The reason everything currently runs "in the cloud" is very simple; it binds you to the vendor and prevents anyone from reverse-engineering the software in any sort of usable form. It's essentially DRM gone wild.
I agree. I suspect most people would be a lot happier with a dozen or two commands customized to their personal use cases.
Instead we have companies shuffling data back to their servers attempting (and usually failing in my experience) to handle arbitrary commands mostly for the company's benefit
Google Assistant DID figure out which case should lead to which actions. It knew that it was supposed to call home.
Instead of performing the action, it refused to perform the action and demanded that I use my hands to unlock the phone in precisely the situation where doing so was both illegal and dangerous.
You're right about the spurious cloud connection requirement. There is no way that we need to make a cloud connection for this or many other functions, yet it seems to be the default architecture for almost everyone these days. Just because you can does not mean you should.
It is then refusing to do the action because they decided that I need to unlock my phone, PRECISELY in the situation where I CANNOT use my hands to perform that action.
Also, I last summer had an opportunity/need to use an obsolete DragonDictate from circa 2009. This old software was FAR better at recognition and well-thought-out command flow than any of Microsoft's or Googles current offerings. Yes. it is hard, but doing better than a decade behind is not hard.
So, both your general premise and your specific characterization are wrong
Why would they ever not use this data adversely against customers? Insurance companies already have offered an OBD tracker that would supposedly be used to prove you were a safe driver and therefore lower your rates. Why on earth would they not use this data to say "well you check twitter on the highway every morning, therefore we are raising your monthly payments by 20%." This is the entire insurance industry business model.
It seems like Zendrive is something you can integrate on your apps. Their target market seems to be companies that have fleets to manage, like Uber or DHL.
From Zendrive's FAQ page[1]:
> How does it work?
> Zendrive measures driver safety using only phone sensors. Integrate the Zendrive SDK into your driver app, and we’ll measure your drivers’ Caution, Control, and Focus while on the road, as well as detect collisions. This is done by measuring a wide variety of safety factors, like speeding, hard brakes, sharp accelerations, phone use, swerving, length of time driving, time of day, and many many more.
> Who’s it for? How are people using this?
> We built Zendrive for any On Demand company with a growing fleet, to help them manage their growth. With in-depth analysis showing any particular driver’s safety, or fleet-wide safety, it gives fleet managers the tools they need to improve their fleet’s driving to ensure safety, reduce risk and liability, and increase savings.
> Since launch we’ve found additional types of companies getting value from Zendrive, from fleet management platform companies, to activity tracking companies and expense tracking companies who use the driving detection in the technology, to hardware companies looking for new safety features.
Well, I doubt they'll reject clients based on the use they intend to give to their product, but I can't really fault them for that. It seems like a legitimate business. Hardly unethical.
Perhaps we could imagine better privacy laws. The fact that something is legal, and a legitimate business, doesn't mean that it's good or should be allowed.
Most (all?) of their examples in the FAQ are related to corporate drivers, not individuals. Something like Uber would be a bit of a gray area since you use your own vehicle, but all the others are more like DHL and FedEx, where you're driving a company-provided vehicle and there should be no expectation of privacy in regards to how you operate that vehicle.
Those are the examples given, but that wouldn't explain the huge numbers claimed elsewhere - nearly a quarter of phones in the US. Pardon me for not assuming that a data gathering corporation has my best interests at heart.
This is the closest the article comes to answering that:
> The reason for all this data is that at least one in five U.S. auto insurance policies now offers a potential discount if the customer consents to a vehicle monitor.
But I would also like some kind of affirmation that the software isn't being backdoored onto everyone's phones.
I thought those were typically GPS loggers that were powered by (and pulled data from) the OBDII port[1]? An app-based alternative could have some of the same info, though presumably less accurate acceleration and braking data.
I immediately wondered that as well. In the best case, insurers are giving people a discount to install an app, like health insurers handing out connected fitness monitors. That's creepy, but only mildly so by today's standards. However, I recently renewed my car insurance with a popular company, and they didn't ask me to install an app, so that seems unlikely.
I wouldn't be surprised if it were silently bundled into a bunch of apps that have location access and run in the background. It gives insurers vaguely-useful data, but mostly exists to hoover up sweet, sweet location data.
EDIT: See below[1,2]. It's the extra-creepy version.
It seems like another good follow-up question might be: how do we know zendrive's characterization of phone-use-while-driving is accurate and represents danger?
E.g., how does the thing know if you're driving or simply a passenger (especially on mass transit, like a bus)? And even if you are the driver, checking your phone at a light is probably not especially unsafe (though it could be rude or illegal). Finally, audio output app use like Maps or Spotify that doesn't require input is probably acceptable. So without knowing more about the methodology, the conclusions of the article don't logically follow.
I believe that this is the same app I encountered on my mother's phone. She got it from Verizon and got a discount on either mobile service or the phone. It has a bluetooth dongle that plugs into the cars OBDII port. I tried talking to her about the privacy implications though I don't think I got through to her. From what I could see the app was slurping the location of the phone all the time, even outside the car.
That sounds more like the car insurance schemes to give a discount by sharing your data. Are you sure she didn't opt into that? Just curious as first I've heard of such a thing outside of auto insurance.
I suppose the safest method would be to switch off a phone, or switch it to airplane mode, while driving. Would a requirement to use such an app while driving prevent that approach?
You can use an app like Offline Maps & Navigation that doesn't require the internet (after downloading the required maps)... and I think that Google Maps allows you to bring navigation instructions offline.
Sure, that's possible. But that means foregoing traffic alerts and redirections.
And it means you can't share your location with other people so they can track your progress. I do this on long trips so the people I'm driving to can see where I am without having to contact/disturb me.
Sure there are trade-offs, but anyways, I guess this app could be recording all the tracking data onto internal memory until you switch airplane mode back off and then transfer it to the insurance company.
"Zendrive now has its monitoring technology on 60 million phones, roughly one of every four U.S. drivers"
For example: How did this "monitoring technology" get on 60M phones? Do the owners know it's there? Is it hidden in other apps? Is there clear disclosure of its existence in those apps? What is done with this data? Who has access? Is it anonymized? Is it sold to third parties? Is it used for advertising or other non-safety-related purposes? How do they distinguish between drivers and passengers? What do they do about COPPA if it's on the phone of an under-13-year-old? Etc.