Phone Makers Could Cut Off Drivers. So Why Don’t They? | Hacker News Reader

Phone Makers Could Cut Off Drivers. So Why Don’t They?

nytimes.com

12 pointsmsabalau61 comments

It would be a massive breakthrough in machine learning to create software that figures out if the phone is being held by a driver or passenger, and unsurprisingly this article claims they "could" do it without presenting a method. What a load of rubbish.

Exactly. A related issue: the console in most cars disables many functions while in motion, such as the ability to dial a phone number or change settings. That prevents the passenger from doing so as well.

Especially annoying considering how all modern cars have a sensor to know if there's a passenger (so they can play a noise if they're not wearing a seatbelt). The console should unlock when there is.

You could even go one step further, and lock the console controls if the driver doesn't have both hands on the wheel. That would still allow a passenger to use the console.

What consoles do that? I never experienced that, even basic cars like a rental Ford Focus give you full control from the steering wheel including making a call.

You can still do things available from the steering wheel, but Toyota and Honda both lock out specific console controls, such as the dial pad to dial an arbitrary phone number, or most of the settings dialogs.

Could be a US thing, tho I can't remember ever trying to dial a number directly.

I do remember when some german cars used to come with a dial pad in steering wheel :P

My roommate's late-model Subaru Crosstrek. We seriously have to pull off to the side of the road so that I (the passenger) can change the BT pairing so that my phone plays music instead of his. The experience is so incredibly aggravating that it has put me off Subarus completely.

My Hyundai won't let you pair a phone while moving, but it will let you switch which previously paired phone it is connected to while moving. It's critical for couples that share cars.

As an example, now Jeep Cherokees don't let you use/reprogram the navigation while driving.

(This is the same car that can be hacked into and driven remotely through the entertainment system. Awesome that they are minding owners' safety)

Yeah. Typical user experience: https://twitter.com/supersat/status/763206342328758273

It seems GPS inside my office building is a bit haphazard. Just leaving my phone on my desk is enough to trigger it once every few minutes - eventually GPS will bounce you around enough that Pokemon Go thinks you're driving.

My phone can't even tell if I'm lying on my side and don't want my view to rotate.

What would you lie to your phone about?

Exactly. This is not something that we currently can do. Pretending that it is currently possible is disingenuous.

Well we could borrow from drunk driving regulation. In California, as a condition of getting your drivers license, you agree that a cop can check you for DUI at any time for any reason.

My proposal is to extend this slightly differently. Instead you have a UI in the phone that pops up a message saying something like "It appears that you are operating a motor vehicle. Under penalty of perjury [or whatever legalese should go here] you affirm that you are not operating a vehicle," you then either click "OK," or "cancel."

Then if you cause a car accident, or there's a "texting while driving" checkpoint that finds you clicked "OK" and there was no one else in your car, you get hit with severe penalties.

Waze will pop up a scary warning if you try to type while moving, and you have to press the 'passenger' button. I imagine that it will stop precisely zero people from driving while typing.

Which is why I suggest putting legal ramifications behind it. For example if you cause a fatal accident while operating a mobile phone after hitting the "passenger" button you should have a mandatory of some time in jail. I have no idea how that lady escaped with no jail time.

What an absolute waste of an article. It just repeats itself over and over and over quoting different people each time.

"They should fix it, blah blah blah. Make them do it, blah blah blah."

If you don't answer the question of HOW then you just completely wasted your readers time.

There is the tiniest mention in passing about something using sound waves to guess the location of the person holding the phone. It doesn't work well.

And that's it. What a useless article.

Gave up less than 3 paragraphs in...

"But if the behavior has addictive qualities, can drivers really be expected to police themselves?" the article asks, nearly at it's end. Not answering it, of course, because if drivers actually are responsible for their actions, the rest of the story in meaningless.

Alcohol also has addicting qualities. I'm pretty sure we allow alcohol-consuming motorists to consume their preferred beverages and just expect them to self-regulate. It would be extremely laughable if someone argued to ban bars and liquor stores, that are say within 500 meters of a road to prevent abuses.

Yea, that's just one step down the path to a world where meaningless and unscientific personality tests (which is all of them) are used to dictate who is allowed to do what, all in the name of safety. I would much rather have the liberty rather than the illusion of safety.

Another way to look at it is that, to quote Heinlein, ignorance is it's own death penalty. In principle the people dumb enough to drive while texting will kill themselves faster than people who aren't so dumb. On the other hand the risk of collateral damage is pretty high, so it could be ineffective. Having the police issue citations for it is not very effective, but probably worth it. Not an easy problem to solve, and one that's only going to get worse.

I think that the only technical solution that might fix it would be truly self-driving cars, and you have to expect that to be pretty far off. Even if they were available right now it would take at least a decade to get good enough market penetration to really reduce the risk.

Are what technology, pray tell, will prevent the driver from user their phone while driving, while at the same time allowing their passengers to use their own phones?

Also: a great way to put Uber and Lyft out of business.

Exactly. Even ignoring things like gps / maps, music, and audiobooks / podcasts, car passengers will not tolerate their phone being disabled. Or passengers on public transit.

What kind of infantilized nanny-state fingerpointing scapegoat culture are we living in here? It is beyond ridiculous to say phone makers should be responsible for fixing this problem. Not just because of the impossibility, but because of the slippery slope of impossibility for all the other things that all the hand-wringing in the world won't put under the control of major corporations' safety initiatives.

How about this: texting while driving is as dangerous as drinking and driving yes? How about giving it commensurate penalties and drivers license revocation?

Only it's /not/ the same.

A driver 'under the influence' can't sober up in ~5-10 seconds.

A driver 'under the influence' can't /plan/ when they're going to dedicate a few seconds to interacting with a radio, or a GPS, or opening up a voice input prompt to something.

What WOULD make it safer, however, is if the UI guidelines for a platform followed common sense. If they realized that OF COURSE drivers will click through some annoying popup. That it would be SAFER if they STREAMLINED interaction and gave a more astronaut glove friendly interface.

Thankfully, both of these are temporary problems: in another five years we will start transitioning large numbers of people to self-driving cars (whether privately owned or simply self-driving taxis), and this problem will go into decline.

No we won't. There will not be in-city self driving cars for decades.

There will be self driving cars on the freeway in specially marked and instrumented lanes made just for self driving trucks (cars will come later).

But not in-city. Not till there is general AI.

This is one of those comments that people will dig up and mock on HN in about 5 years.

I except reality will fall somewhere between their two responses myself.

Or maybe yours will be.

I give it 50/50.

Google started in 2009, it's now 7 years later. We (they) are no closer to success than they were when they started.

(The cars can only drive on pre-programmed roads with a human marking any changes such as temporary traffic lights.)

What makes you think that they will succeed in 5 years if they have not solved the hard problems in the last 7?

No. Self driving cars (trucks actually) will only drive on limited access roads AKA freeways. Those roads are controlled, and any changes can be easily programmed in.

Or we could have the phone makers fix it directly.

There. Problem solved.

And we didn't even need to get thousands of local government agencies involved in enforcement! We just fixed the problem at the source: the cell phone manufacturers.

Would you rather fix a million separate problems - the millions of texting drivers - or one problem at the cell phone manufacturer?

How much would your solution cost vs my solution?

Systematic solutions that remove human judgement from the solution will always be better than any solution that's dependent on human judgement to work.

The nanny state exists because people are dangerous idiots. And they need to be nannied.

The correct point to remove people from the equation because they're idiots is operating the car.

Fully automated driving, without human supervision being necessary.

Or better yet - a rail system.

You didn't solve anything. You did the equivalent of yelling and crying and saying "you fix it".

Unless you also have a suggestion of how to do it?

The solution is to force phone manufacturers to disable phones in moving cars.

So passengers phones are disabled, too?? No thanks.

How many dead souls per passenger phone call are you willing to trade off?

You say that in a shocked tone as if we don't already accept a huge number of motor vehicle deaths for the convenience of driving.

That doesn't exactly answer the question now, does it?

Tu quoque fallacy.

It also presumes my lack of awareness of the additional problems.

As I'd commented and cited in this thread: mobile-device distracted driving is a factor in 25% of motor vehicle accidents.

All of them. Making the world 100% safe is a fools errand. There are always risks. I think the current state of risk is just fine, and I'm not willing to lessen my convenience just to reduce a risk that I don't perceive to be very high.

How do they tell the difference between passengers and drivers?

How do they tell the difference between cars and buses?

> How do they tell the difference between passengers and drivers?

> How do they tell the difference between cars and buses?

You don't. You disable it for all people in moving vehicles.

And how do you propose to convince people to buy such crippled phones?

They won't have a choice, since all phones would implement this crippling.

The Apple patent mentioned in the article describes how it could be done. It wouldn't be as simplistic as disabling phones in moving cars, but would also sense whether the operator was sitting in the driver's seat, among other things.

That said, I would still rather police my own behaviour than have my phone do it for me. (I also would prefer phones have an official rooting procedure, for that matter.)

Today while I was driving I was waiting to make a right turn. I waited for a car to pass and then pulled out and followed the car. While the car was passing me I could see the driver texting. It was a teenager / early 20s. I followed him for 3 miles on my way to my destination. The speed limit is 50 mph. There are trees on either side of the road and the road is pretty narrow. There are hills and curves so visibility is pretty limited.

The teenager was all over the road. At one point he crossed into oncoming traffic while he was going up a hill with limited visibility. It's a marvel he wasn't killed. I vowed never to bike along the road in this community. This kid could have easily killed someone.

I was in a rural community. Lyft /uber aren't a solution. I'm not sure if the phone companies can solve this but man it's super dangerous out there if you aren't in a car.

In my neighborhood, we have had a lot of cars driving into fixed objects like trees and buildings, during clear weather. It feels like texting while driving.

Um because many people use phones in cars who are "not" driving. How are companies suppose to determine when a phone is being used by a driver vs the passenger right next to them or the people in the back seats? For someone not driving there are countless good uses such as taking photos/video of sights, games, directions and more such as calling someone for an address, or for reporting your own or someone elses emergency.

Trying to ban it will never work since people will always find a way around it or even worse alternatives, and any company that does, another company will exist or popup offering the ability.

Just like with the industry trying to stop illegal music downloads it was futile to stop them until instead the provided a better alternative (cheap streaming, easily accessible high quality music). In the case of cars, Apple and Google are hoping to entice users into using safer more convenient options that holding a phone in their hand, with CarPlay and Android. The downside being it will take a long time before the average person has a car with that available.

Same reason car radios don't lock out their controls when the car is in motion, and why the driver's side visor flap doesn't lock up the mirror.

some navigation systems will, can't look up an address while moving, even the passenger.

Cars do lock out everything on the console except maybe the volume control.

Some cars. All my Audi locks out is pairing a new phone. Most rental cars I've had are similar.

What car locks out, say, the frequency dial, or the track skip buttons?

Be an adult and don't use your phone while you're driving. Tired of this argument.

It's true. Better yet, disable cellular access for most apps when you don't have Wi-Fi - to your pleasant surprise you might actually start noticing the world around you, driving or not.

Matt Richtel, the author of this piece, is not unfamiliar with the subject of distracted driving. Indeed, he's written the book on the subject: A Deadly Wandering: A tale of tragedy and redemption in the age of attention

http://www.worldcat.org/title/deadly-wandering-a-tale-of-tra...

It's very simple: I will not purchase a phone with this feature. Period. I don't need technology nannying me.

Hold phone manufacturer, vendor, and carrier executives, engineers, and sales and marketing staff responsible for manslaughter charges for each death linked to phone use whilst driving.

I suspect a solution to the problem will be found rapidly.

This is a problem. It's a current problem. People die, at the rate of 350 per year: http://www.nsc.org/learn/NSC-Initiatives/Pages/priorities-ce...

Cellphone use causes over 1 in 4 car accidents: http://www.usatoday.com/story/money/cars/2014/03/28/cellphon...

An earlier study put cell phone distraction at 2,600 deaths, and another 330,000 injuries, in 2005: http://www.livescience.com/121-drivers-cell-phones-kill-thou...

Those are lives snuffed out or altered forever, plus the families, friends, neighbors, and co-workers affected. Hundreds of thousands directly, millions indirectly.

For the convenience of a text, phone call, or music?

If the issue is one of driver (or passenger) convenience, versus the lives of innocent drivers, cyclists, motorcyclists, or pedestrians unfortunate enough to share the road with those who are unable to resist the lure of devices designed to be as addictive as possible, then I'm going to side with safety, and against convenience. We've tried common sense. It's not working, people.

The largely irrational invective expressed within this thread against the suggestion itself strikes me as a very strong argument for technolgical blocks an very strong product liability, such as that which I lead off this comment with.

The argument that self-driving cars is an easier problem to solve than "is my cellphone being used in a car", or even "is my cellphone being used by a driver in a car", strikes me as ... less than convincing. I may be wrong.

The point that an immediate retroactive software or OS change can fix the problem, but self-driving cars will take a a minimum years, if not a decade or more, to emerge (if ever) seems to have escaped many as well.

Neither is is the job of an op-ed writer in a nontechnical newspaper to supply the technical solution. We don't require those calling for a public policy or international relations measure to specify turnkey solutions. Only to make a compelling case that the change ought be implemented.

Apple, Google, Microsoft, Verizon, AT&T, T-Mobile, LG, Samsung, and many others have been murderously irresponsible.

Including, I'm quite sad to say, far too many HN commenters here.

Reply on news.ycombinator.com