iPhone 14 car crash detection triggered by roller coasters Perhaps "a million hours of crash data, real-world driving and crash test labs" aren't quite enough for Apple's car crash detection feature on the iPhone 14 amid reports that it is being triggered by roller coasters. According to The Wall Street Journal, 39-year-old Sara White was caught out last month when she decided to …
Well of all the testing they did, apparently none of the people doing it rode a roller coaster. At least not a properly wild one.
Roller coasters have limited G force - in order to maintain consciousness at 9g fighter pilots need to be in tip top shape and wear a pressurized suit to keep the blood from pooling in your extremities. So they can probably assume roller coasters don't go above 5g, and increase the limit for what triggers the "crash" reporting.
Or maybe a bit of 'AI' can tell "this looks like I'm on a roller coaster" and disable the crash reporting for a few minutes.
What I wonder is how they tell the difference between "I was just in a crash" and "I just dropped my phone on the floor". Obviously they have that sorted or false alarms on roller coasters would be the least of their problems, so I'm sure there's a software solution at hand for roller coasters that won't involve anything crazy like geofencing amusement parks.
It's possible the iphone is measuring the rate of change of acceleration rather than the absolute value. Roller coasters can be designed to give rapid changes in acceleration rather than massive peak accelerations, although if you go too extreme, rapid changes can be a lot less pleasant than a continuous raise G level.
> how they tell the difference between "I was just in a crash" and "I just dropped my phone on the floor"
They probably don't, but that's what the preliminary warnings the article spoke about are there for: You can tell the phone that no, this wasn't a crash, don't call 911.
The problem only arises when you can't hear the warning because you're in a wild ride full of screaming people. Or obviously if your phone drops (off your bike, out of your plane) and you can't get it back to cancel the alert before it goes off.
As an experiment, I dropped my iPhone 14 Pro Max onto the floor from head height. The floor is carpeted (I'm not stupid) but that didn't trigger any sort of alert. The carpet is pretty thin and it is a concrete slab underneath in my basement where I did this so still should cause a fairly rapid deceleration.
Regarding sustained acceleration, the most on a coaster of the last 50 years has ever hit is about +6. Thriller, which ran on the German fairs a few years was famous for producing that much.; but is very much the exception.
+3.5 to -1 is more or less the nominal range. US legislation touted by Senator Markey in the early 2000's put limits on thrill rides that stifled the use of sustained positive G at levels above that, though it did not require older rides to be shut down for exceeding limits. The US being the coaster capital of the world means that those limits (indirectly) feature on most attractions elsewhere now too. The early 90's saw rides being built up to 4, maybe 4.5G; and it is no accident that those rides consistently feature in fan-favourite lists over more recent additions subject to limits on design.
Jerk (rate-of-change-of-acceleration), as opposed to acceleration is much more variable. There are rides out there that are borderline criminal for how much they generate.
There are plenty of real 911 worthy accidents in amusement parks, so unless the 911 call is able to somehow indicate "this was generated by an iPhone's automated crash detection" the authorities will have to respond. Given that it took years and lots of effort just to add mapping capability to 911 calls unless the iPhone is basically providing the audio to tell them 911 operators that there is no way for them to know.
You can disable this feature permanently without turning the phone off or putting it in airplane mode.
I can't wait for the ski season to kick in - many of the slopes are actually roads during the summer; you can easily hit the same speed as a car and I'm able to confirm the possibility - or likelihood, depending on your approach - of some serious shock loading on both the phone and the body.
In my case that would work, and should. I haven't been on skis for a while since my last time when I did not have myself properly checked for concussion after a fall. The next trip after that fall had a turn to avoid a rock wall. Well, I didn't - I blanked out just before there.
Waking up a few days later wasn't the scary bit, even though I was surrounded by machines that did more than just 'ping'. You know the warnings you find in the bits of paper that you throw away, that you ought to replace a helmet after a collision? When I found my helmet, the state it was in made that hint entirely superfluous. Long story short, I took the hint. Skiing is probably not my thing :).
On the plus side, I now have a documented medical reason to 'forget' my PIN if anyone asks..
Well done for being smart enough to wear one! Sounds like it did the job it was designed for. I've had two friends concuss themselves despite being virtually born on skis (Alaskans), so don't feel too bad. When they came around, they both bought helmets :-)
When I first got on a board 20 years ago it was pretty uncommon for skiers but now it's just ESF instructors, the old guys still in the same one-piece they bought in the 70's, and the few rookies who missed the memo and are perhaps not long for this earth. Although there are still plenty of folk strapping a camera on top, neatly turning a device designed to distribute load into one that will channel the impact into a point on the top of their head. Sometimes I despair.
replace a helmet after a collision?
same thing applies to unauthorised horse dismounts, mine had a 1.5 inch crack up one side where my head said "hello ground!" (which was not friendly at all) after going out the side door (horse spooks left and like cartoon style you hang in the air and drop!). Thankfully no memory of the fall, all I got was the group of us yelling "lets go" as we cantered off then lying on my back in the mud.
Yeah and head mounted camera's, a nice bulky object just waiting to punch a hole through the top of the helmet when you fall on it!
same thing applies to unauthorised horse dismounts
.. and motorbike ejection events...
I've seen people riding around (in the days when I rode as well) in helmets that had clearly been dropped and/or potentially crashed in, judging by the scuffs and impact marks. They obviously don't value their heads..
(Even though the contents of such heads have themselves taken a physical and chemical battering over the years!)
I like to do some serious downhill MTB, there are plenty of jumps and many harsh landings - wouldn't want to have to stop after each to reset the alarm status. (there is the occasional fall - but then that's a feature that might be welcome)
Serious question though, is it only the phone or do the iWatches have the same sensors? If so, judo practitioners might be in for a slamming surprise!
No, the (even more) expesive iWatches merely have a G force sensor that can handle 250G.
As I said before, that's why they were made extra water tight because after that amount of force the only way to remove you is by jet cleaner. And there's no need to call am ambulance then.
Anyway, a more serious answer is yes, the Watch had fall and crash detection quite early on, but its ability to subsequently call out depends on it either being equipped with phone capabilities or it's near a mobile it can route calls for it.
I can't wait for the ski season to kick in
Damn you. As soon as I read your comment, the classic-era "Ski Sunday" theme (official name "Pop Looks Bach" in case you're wondering, trivia fans) started playing in my head, and now it won't stop...
This post has been deleted by its author
And sat there in eager anticipation, wondering which exotic location it'd be coming from that week - bonus points if it was Garmisch Partenkirchen, because with a name like that you just KNEW it had to be special!
Same with the international athletics coverage during the summer months - for a child of late 70s/early 80s north-east England, where if you were lucky you might have travelled beyond the UK shores for a week in Benidorm, and things like Google Earth/Street View were just a dream in the minds of those with particularly vivid imaginations, to be whisked all over the world almost every week by the magic picture box in the corner of the living room was beyond awesome. And if the sporting coverage wasn't enough to slake your thirst for travelling the globe from the comfort of the settee, there was always things like Wish You Were Here or the epic Whickers World. God, sometimes I *really* miss those simpler times...
It's bad enough that Apple copy features from the competition and call them innovative* but why is the Apple version of the feature always worse than the feature they are copying?
*Apple buyers believe that every new iFeature is innovative because Apple buyers think that only Apple exisits
It does show the power of marketing though. Apples crash detection was all over the news media as the new iPhone was launched, even before the latest news of the false positives on roller coasters etc. And yet I have no memories of crash detection being a big selling point in the media when it was introduced on the Pixel (or any other phone that may have it).
Apple release new device - hugely newsworthy (do the media all use Apple gear?)
Samsung release new device - moderately newsworthy (The media have heard of Samsung, but don't use their kit much)
Google release new device - <tumbleweeds>, media response, Google make phones? Who new?
"turn your new shiny off" - haha, as if that were an option.... Not! I seem to recall a university experiment where the lecturer wanted to see if the students would learn better with their mobiles switched off. Except the students were so freaked out they didn't learn anything at all.
Can't even convince folks to "take the stuff out of your pockets so it doesn't hit the people on the ground when it falls".
I worked on an amusement park ride one summer. Most folks were good about emptying their pockets into the plastic bin - in plain sight of everyone within 50', so hard to steal from - before getting on. (Two riders at a time, so easy to tell whose stuff it was.) But occasionally some folks would accuse us of trying to steal from them. No, it's for our protection. Getting hit by a quarter that fell 100' is painful, but thankfully the wallet missed. A Bic lighter hit the concrete and exploded.
I attended a facility a while ago where anybody entering a particular area was asked to leave any electronic devices with the guard at reception. The devices were bagged labelled and customers given a receipt before the devices were locked away securely. However they had a problem with people trying to smuggle phones in. Only phones.
People would hand over all sorts of devices but not phones. Even when the guard asked "do you have a mobile phone" visitors would reply no, even thought the signs clears stated that electronic devices could be damaged if taken into the facility even if switched off. After finding contraband the guard told me they had to start sweeping folks with a detector and even found people with phones stuffed inside clothing.
Some people are so addicted to their phones that they would rather risk them being damaged than be without them for half an hour.
It's not just the iphone doing this. My android brick of some description pocket called the emergency services for me at Alton Towers on Wickerman. I had a private call from the police asking the sounds of screaming afterwards. Obviously, apologies to wasting emergency services time.
What was perhaps slightly odd is that the car-crash of a ride that is the Grand National in Blackpool didn't trigger it. But a modern, smooth wooden coaster did.
"a million hours of crash data, real-world driving and crash test labs"
That said, I suspect if they were able to persuade the various car maufacturers to provide them with the data from all of the airbags and suchlike they've individually captured as part of crash investigations, they'd have a pretty sizeable dataset from which to learn about the sorts of sensor inputs you'd expect to see in a crash.
Maybe all the crash test dummies for NTSHA tests are being equipped with iPhones?
I'm only half facetious here, NTSHA may want to conduct crashes with phones in the car in places where people might leave them in the cabin (i.e. hands free mounts on the dash or whatever) to see how they go flying about in the cabin during a crash.
Each time the emergency services get called by this feature but it's a false alarm, fine Apple $100. If it's a handful per year, they'll just pay the fine, no problem. If it's hundreds per day, they'll be much more likely to try to fix it. (Size of fine up for debate.)
Apply the same logic to telcos letting through spam calls and texts. Give them some freebies - first 100 per month are free, which means they still have to block 99%+.
Emergency services in some places will fine people for too many false alarms - mostly businesses to 'encourage' them to get their equipment and procedures in order - but I imagine some people will fine people who e.g. have kids that accidentally call 911 too often.
You'd have to be in the roller coaster of the week club for this to impact you though.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
