Uber's Smartphone-Based Gyrometer Monitoring Seems To Be the Future of Driving (thestack.com) 85
An anonymous reader writes: Uber has announced that it has been conducting trials in Houston, Texas, since late 2015 which use data from the gyrometer in drivers' smartphones, combined with accelerometer and GPS data, to perform forensic analysis on Uber journeys where the customer flagged up errant driving behavior such as speeding or tailgating. Uber's post also indicates that talking on a phone whilst driving may be included as a factor in safety-oriented trials. The auto-insurers' move from dedicated telematics technology to smartphone-based data provision was spearheaded by British insurer Aviva in 2012, with massive U.S. insurer Progressive now actively pursuing driver monitoring. However the premium reductions are diminishing as the practice heads from experimental, to default, to obligatory — or so many believe.
Re: (Score:2)
Tracking drivers doesn't seem progressive (Score:5, Insightful)
I've watched a few episodes of Cops, and the roadside interviews on camera are astonishingly more polite than some I've experienced.
Re:Tracking drivers doesn't seem progressive (Score:5, Insightful)
Do insurers really believe people drive the same way when they know the monitoring device is plugged in?
Of course people drive differently when being monitored 100% of the time. That's kind of the point here, behavior modification to make driving safer.
I've watched a few episodes of Cops, and the roadside interviews on camera are astonishingly more polite than some I've experienced.
Polite or not, every time I see a "roadside" interview I want to drive a sign into the ground that reads "PLEASE DO NOT FEED THE NARCISSISTS".
Pretty much sums up the benefit of that activity.
Re: (Score:2)
Re:Tracking drivers doesn't seem progressive (Score:5, Interesting)
No this doesn't work as intended.
Progressive penalizes you for what's called a 'Hard stop' which is 7 Miles per hour of deceleration per second. Living in Austin Texas at the time taught me a few things to avoid the penalty.
- Run most if not all stop signs in parking lots,
- Not give any shits about yellow lights
- Run some red lights if the yellow was too short.
- Switch lanes rapidly to avoid cars slowing down for *anything*
- Increased my anger at other drivers for forcing me to hard stop, ( eg: playing it safe and avoiding a head on as someone gets into the wrong lane )
I had to effectively dive my vehicle like a golf cart to avoid the hits to my discount which translated into poorer driving habits that have persisted long after we switched insurance companies.
I am not alone in this either. You can spot a progressive driver a mile away.
Also a good write-up by a blogger here:
https://blog.joemanna.com/prog... [joemanna.com]
Re: (Score:1)
It will be much easier if they offer a discount to drivers who install software on their cell phone than if they try to mandate it.
Re: (Score:3)
The insurance company pays out less money. They pass on part of the savings.
Speaking as someone who has just had to shop around for their car insurance... no they don't. The one doesn't automatically follow from the other.
Re: (Score:2)
"since obviously people who commit driving-related crimes are a higher risk."
That might less obvious than you think. Drunk drivers are a class of their own, I don't want to talk about them. But, people with a lot of tickets aren't necessarily any more dangerous than you, or your Grandma, or the idiot who drives ten mile below the speed limit all the time, or any number of other fools.
makes there drivers more like W2's and not 1099's (Score:2)
makes there drivers more like W2's and not 1099's. So how far do they want to push this?
Re: (Score:2)
Forces 2 phone workaround (Score:5, Insightful)
Re: (Score:1)
I can see an app that hooks in so the gyroscope data and such are not handed over, or faked. An XPrivacy, or a PMP (protect my privacy), perhaps.
Re: (Score:3)
I don't think I've been in an Uber where I've seen the driver with less than two phones.
I'd always assumed one was personal and one was their Uber phone, but maybe for some one is Uber and one is Lyft so they can pick and choose based on where they are.
Re: (Score:2)
maybe for some one is Uber and one is Lyft so they can pick and choose based on where they are.
My sister drives for Uber and Lyft. She has both on the same phone, and that has never been a problem. She started out just driving for Uber, but added Lyft for the extra business as soon as they stopped requiring that stupid pink mustache on the front of the car.
Re: (Score:2)
So, then Uber drivers and anyone else being monitored for insurance etc, will resort to carrying 2 phones. One tied to their car, one for calling/texting so they can do both simultaneously and not get dinged.
Speaking of dings, feel free to elaborate how an Uber phone addict is going to use 2 phones to text their way out of the accident they caused due to distracted driving.
Then elaborate as to how this texting addict is going to explain to Uber who monitors this kind of illegal activity why they should not be fired for causing an accident doing something they agreed not to do while driving for the company.
Not get dinged? Yeah, fat chance of that.
Re: (Score:2)
No, most people do NOT do it. A large percentage, probably. Most? No.
But tell my last truck that it is OK to text and drive. When the 17 year old came strolling through the intersection at 55+, broadsiding me and rolling it on the roof...That was not a fun way to spend my lunch hour.
Re: (Score:2)
Well considering that most texting and driving does NOT cause an accident (source: most people do it, and we haven't of 80% of driver fatalities) then yes, avoiding the monitoring solves the biggest risk, namely getting caught.
If you're going to try and use this argument, then I see seat belts as not only a pointless thing to wear, but should be an optional accessory in cars.
After all, it's not just "most" of the time you don't need a seatbelt when driving. It's more like 99% of the time.
We're approaching 20% of all accidents with injuries in the US being related to distracted driving. How large does that number need to get for people to realize it's a problem, especially when running a company where you are not merely responsi
Easy (Score:2)
Re: (Score:2)
Future? (Score:2)
Regardless, this sounds less like the future of driving and more like the future of getting boned on your insurance rates.
Re: (Score:2)
The day this shit becomes compulsory for Progressive is the day I find a new insurance company and/or buy 7 more cars (junkers) so I don't have to carry auto insurance. Uhhh, well, maybe not that second option. Regardless, this sounds less like the future of driving and more like the future of getting boned on your insurance rates.
As car ownership declines into nothingness as we face the inevitable future of autonomous driving with vehicles we're not even allowed to own, expect the insurance bullshit to get worse.
Much fucking worse.
Here's an example...you act like you're going to have a fucking choice of who you run to when you leave Progressive. You won't, once human monitoring becomes a Federal mandate.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Quite the contrary. Red Barchetta predicted cars so safe and crash-proof that they could be used as battering rams against older cars. The aggressor vehicles were definitely being driven by people, however.
Song's based on "A Nice Morning Drive."
http://www.mgexp.com/article/n... [mgexp.com]
Re: (Score:1)
It seems unlikely. I do not know as fully autonomous vehicles are anywhere near as close as you seem to think they are, either. For a small example; There are people like me. I've not actually insured a single vehicle that I own (and I own quite a few) in a very long time. Instead, I just insure my license. Any vehicle I drive - even if I do not own it, is insured. And, to touch on the second part, I own a significant number of automobiles - as an aficionado and a collector of sorts. I doubt they'll prohibi
Gyro data is misleading (Score:2)
Say you break heavily because some arsehole cuts you up and hits his breaks. Happens a lot in the UK, because if you leave adequate stopping distance in front of you some prat will pull into it, realize he is going to read end the guy in front and stomp on his brake. According to data it looks like you are a bad driver who brakes heavily sometimes, even though you are avoiding someone else's mistake.
Similar issues exist with telematics that uses the ODB-II port to monitor the pedals. In cars with small engi
Re: (Score:3)
Re: (Score:2)
The driver who is aware of their surroundings but drives in places with crazy drivers will break a lot and sometimes hard, but not cause accidents
The driver who does not care about the car will break softly and ding other cars, and cause far more accidents ...
Re: (Score:2)
Citation needed, or are you just making stuff up because it 'sounds right' to you. Our drivers do the best part of a million hours driving a year so I'll base my position on the results of that data, and the fact that our insurers discount based on exactly the factors you're claiming would increase accidents thanks.
Re: (Score:1)
They're full of shit. Not only am I an automobile aficionado but I worked in the traffic modeling industry, have driven professionally, and have taken countless advanced driving courses over the years. Hell, I've even gone on vacation (at considerable expense) to take special driving lessons, including classroom time and time spent with a professional coach in the passenger seat or in the driver's seat, while renting exotics and I spent *days* doing laps at Nurburgring. That's just *one* of many examples. I
Phone on dashboard/in glovebox? (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
I'm almost tempted to sign up, hire myself to drive, than spend the day at a track burning up a set of tires.
Just to fuck with you.
Re: (Score:1)
That would be awesome! I believe that, during the week, I can get onto the track at Daytona for just a few bucks during the week. I've got my 6 series with me. Hell, want to take turns? We can even buy some slicks and see if we can how high we can get the LA to trip. I'm game for something like that - I'll even let you drive my car for a few laps (it's insured) if you want. It's "that" 6 series. As in dual turbo and ~450 ponies. It's a bit heavy feeling but actually very poised. I've not put it on a track b
Re: (Score:2)
I was thinking of just piggy backing on a SCCA track day and bringing my mild Mustang (300HP ballpark, just installed yellow Konis and polyurethane). The quarter mile car won't be any fun at whatever they call sears point these days.
I might just fly to Florida to drive a M6 around Daytona...you've got me thinking. But I couldn't really run someone else's car really hard, like my own. Also super speedways are just insanely dangerous. I was thinking more road course, hard braking and cornering. Spin withou
Re: (Score:1)
It's got insurance and I own it to be driven. When it breaks, I'll fix it or get a new one. It's just money, they'll make more. Yes it's "bespoke" but that just means I've got some extra sport package features, unusual trim and tires, and some different colors. Oh, and a nice stereo with a fancy HUD that's a little different from the standard model. ;-)
So, drive it like you stole it. 'Cause if my insurance company asks, that's what I'm gonna tell 'em you did. I kid... I kid... My license, thus any vehicle I
Re: (Score:3)
Also, I have not seen any links between acceleration and accidents.
Cars accelerating up to speed isn't when accidents typically happen. If anything, those who merge onto a road without having stepped on it are far more likely to cause an accident. Even if they're not always involved.
Avoiding using more than the first tenth of the gas pedal isn't defensive driving, it's creating dangers for everyone else who already are driving at legal speeds.
Re: (Score:2)
It's less that hard acceleration per se is an issue (or hard braking), but rather a pattern that combines a lot of the two.
Re: (Score:2)
It's less that hard acceleration per se is an issue (or hard braking), but rather a pattern that combines a lot of the two.
I still don't see it. Is there any reason to believe that hard braking is more dangerous when also combined with a pattern of hard acceleration, than hard braking alone?
Re: (Score:2)
Bottom line, it doesn't matter WHY there's a connection, only THAT there's a connection. If a given driving pattern is correlated to higher claim expenses, then it's entirely reasonable for an insurer to charge people who display that pattern a higher premium, and vice versa.
Obligatory my ass ... (Score:5, Insightful)
I don't have a smart phone, and I don't want one.
Unless they plan on making ownership of a smart phone as a mandatory condition for providing insurance, which I question the legality of, they simply can't make this obligatory.
Consenting to being tracked at all times for the benefit of an insurance company? Yeah, go fuck yourselves.
Re: (Score:2)
You could always go with a Windows smart phone... nobody will ever develop an app for that... ;)
Years ago I did the data port device thing in my '98 Jeep Cherokee. I was using Progressive at the time and they were rolling this thing out with the promise of significant savings on monthly premium costs.
I had it plugged in and working for almost a year and was also committed to proving what a safe driver I was. I would never exceed the speed limit, always brake as gently as possible, leave 1 car length for eve
Re: (Score:2)
My money is on tailgating as being the best thing to measure. It astounds me how close others drive to the car in front.
Also, lateral g-force would be very good to know -- a sign of a bad driver, in my books. Whereas acceleration
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
How do you propose to do that? Radar vans don't ticket vehicles unless they are doing 10mph or more over the limit. So *this* should be the measure of what true speeding is.
BTW, I'm not saying I'm an exceptional driver. I'm saying something very different -- that I'm typical, and that typical is good.
Monitoring should only trigger on extremes. Drivers who never go over the speed limit should be ticketed -
Re: (Score:2)
Re: (Score:1)
You've got it backwards.
You are being monitored because of your decision to put others at risk by accelerating several tons of metal to fatal velocities.
Unless you're doing it in a private track, you are endangering the public and they have a stake in how you do it,
Re: (Score:2)
Unless they plan on making ownership of a smart phone as a mandatory condition for providing insurance, which I question the legality of, they simply can't make this obligatory.
They will make it the usual reverse way. If you give them permission to monitor your driving, they will give you a big number of percents off - about the same number that they've just raised their premiums.
So basically, you have a choice - you can pay more, or you can be monitored.
Re: (Score:2)
What if it's unavailable? (Score:1)
What if they don't own a smartphone? What if their smartphone doesn't have gyro bullshit? What if they don't have a phone or data plan, just a phone itself?
This shit doesn't make sense.
Re: (Score:1)
Re: (Score:2)
Or what if the gyro in their smartphone is unstable?
I have a Nexus 5 and an Asus tablet. The level and compass apps I've tested on them should use the internal gyro data... Level is reasonably stable but compass is anything but stable. Especially when moving wildly, the compass goes everywhere but in the right direction...
I'm not against monitoring driver behaviour for drivers transporting passengers on a commercial basis or insurance premium cuts for 'model' drivers. But I do want to see some certified har
Re: (Score:2)
employees? (Score:2)
Re:employees? (Score:5, Insightful)
Why wouldn't a company want to be able to assess the quality of its contractors, and decide if a complaint was valid or not? If you work for an outsourced call center company, and the company gets a call saying "I just spoke to CSR Fermion, and he was unhelpful and swore at me," if I'm the company, I'd want to know if the complaint was legit, rather than just having a policy of "drop anyone who gets a complaint."
The customer is not always right (or sane).
Invalid readings (Score:1)
If it came down to monitoring your phones gyro, then I would let my phone dangle from a string on the mirror, and let it swing around while driving. The readings would so wild and out of parameter they couldn't possibly be counted as valid.
Re: (Score:1)
The fine print will say that invalid readings default to a penalty most likely.
Re: (Score:2)
Only idiots have things dangling from their mirrors.
They are training their cranial collision detectors to not work.
Think about it. What's the easiest way to 'see' a collision course? (Constant vector.) What does a pendulum do at the end of each swing? (present a constant vector)
simple: require Uber (progressive )snapshot (Score:1)
Great. Another data hogging app (Score:2)
Web engineers, meet control engineers (Score:2)
use data from the gyrometer in drivers' smartphones, combined with accelerometer and GPS data,
Also know as Inertial Measurement and Inertial Navigation techniques. Drones, [and some hoverboards] use this all the time (it's their core)...