Recently they had a significant country wide false alarm in Israel at 3AM... There was a emergency alert cell broadcast (similar to amber alert), which caused everyone to move their phone at the same time, which was falsely detected as an earthquake, which caused an Android earthquake alert to be sent to all phones in Israel 30 seconds later. I guess they didn't plan for this scenario
Edit: Arstechina article seems to mention this: "only three were false positives. One of those was triggered by a different system sending an alert that vibrated a lot of phones"
That would be quite an implementation flaw if it didn't account for the phone's own vibrations. Lots of countries use widespread emergency alert messages frequently.
It's obvious to anyone that's written software for hardware sensors to ensure the hardware itself is not going to interfere with readings of other sensors. Don't tell the motor to run if the motor is already running type of issues. It's a fairly common bozo check
Typical Google product. Reminds me of a person who put a bunch of phones in a car and drove which caused Google maps to wrongly show traffic in that area. It was deliberately done though as an experiment
Note that those are three completely false events. The survey results Google published show 15% of people not feeling any shaking (neither strong nor light). That's still a good figure, but reading there were only 3 false positives gave me the impression that you're basically always in for a ride when you get the alert and it's not that miraculously accurate either
Surprised they don't do some signal processing on all the IMU signals to see if they correlate (to within a rotation matrix), and if the timing of shaking at different locations is consistent with the distance to some (solvable) epicenter.
The whole country moving phones in random directions at exactly the same time isn't what an earthquake signature should look like.
I don't get it. I thought earthquake alerts were meant to trigger _before_ the earthquake arrives. If it happens 30 secondes after detecting vibrations, not considering the false positive, it can only mean "hey, you just felt, or are feeling an earthquake, hope you're sheltered".
This is how I experienced the Google Earth Alerts in Wellington, NZ which experiences frequent earthquakes. Earthquake happens. 30 seconds to five minutes later I get an alert from Google.
The best earthquake indicator is an old house with wooden windows that you'll hear rattling five seconds before you get to feel it.
A faster moving small quake (p-wave) will precede the bigger, more damaging quake. This system detects the p-wave and alert people hoping they can get out in time before the big quake hit.
This is when the earthquake is a few hundred kilometers away, such as in Mexico City, where most earthquakes occur off the coast and the waves take a few minutes to reach the city.
Somewhat relatedly, I support a service with global scale traffic. Whenever there's an earthquake in APAC, we get a traffic spike, like 100x normal for that time of night. I'm pretty sure it's incoming alerts waking people up / checking where the epicenter is and if they need to run from any tsunami or flooding, but it's still really hard to scale that kind of thing up for regional demand when it spikes that hard!
This is really cool, and it smells like old-school Google, in a good way, like "let's do this because we can". It feels like it's been a while since something coming out of Google Engineering is meaningful and not designed to unlock new existential creeps, so, well done I guess.
No ads, no creepy monetization angle (at least not yet), just a genuinely useful system that leverages something only Google could realistically pull off. Feels rare these days, but really nice to see.
Nowadays I just assume these "nice to have" features exist in order to get users to enable location services so that they can be permanently tracked. Very cynical of me maybe. But there's never an option to "enable location sevices for this use only"-kind of setting. It's always the globally enabled one.
Cynical: People who makes decisions at Google push for features that trick people into enabling system-wide always-on location services that they otherwise wouldn't. For functionality that shouldn't actually require system-wide always-on location services. The primary goal being to sell more ads, not to help with earthquakes.
A conspiracy: Governments push or force or encourage Google to do it in order to have access to location data on citizens.
A few years back I was woken by a shake in HongKong. To confirm was a quake I found my Android phone and sure enough Google had registered a quake. It was one far inland and <5 IIRC. Creepiness or not as someone who helped after Sichuan 2008 quake these kind of systems can save lives.
> "Of those roughly 1,300 events that triggered alerts, only three were false positives. One of those was triggered by a different system sending an alert that vibrated a lot of phones, something that should be relatively easy to compensate for in software. The other two were both due to thunderstorms, where heavy thunder caused widespread vibrations centered on a specific location. This led the team to better model acoustic events, which should prevent something similar from happening in the future."
Do the range of detectable acoustic sources include military jets, drones, and bomb blasts (i.e., gauging effectiveness of targeting?) I don't know what I'm supposed to think of tech companies turning gadgets into remote-root physics sensors without user consent. Maybe I'm reflexively cynical; I can't trust a FAANG with yet another side-channel attack, *even if* the first (public) application is, on appearance, a life-saving unalloyed good.
I have received a few earthquake alerts (Greece). one was for a significant 5.2 earthquake about a month ago, and the notification arrived about one minute earlier or so. It woke me up , and i was able to experience the entire duration of the earthquake. Pretty cool if they were using the new system and i was impressed at the time.
Last time we had a reasonably powerful one in Portugal, I grabbed my phone the house was still shaking and it already had the Android warning!
I was surprised as I didn't even know it was a thing.
I was also a bit spooked as it was in the ocean, near the coast, and when I turned on my FM radio as we were always taught in school, all I heard were pre-recorded music programs.
Turns out it didn't meet the threshold for a warning so the authorities didn't issue any message about tsunami danger. I think they should've anyway, as I wasn't the only one that had that thought.
yeah I've ran into bunch of people who were scared for real and drove up to higher elevation in the middle of the night.
If you'd search online you could have known quickly there was a negative Tsunami notice, but I get that this is just not feasible for everyone (or everytime).
Feel like the severe weather SMS etc are working quite decent, I wish they'd expand that for those sort of things as well (like there was an earthquake, this is what you should do next).
The Android notification was a bit odd, because it's not trivial to get back to that notification if you've just skipped/acknowledged it.
Few months back we experienced an earthquake. I got an alert on my Android, which at first I was confused about but took me a second to process that there is a possible earthquake and then we ran out and it was a 5.2 magnitude earthquake. So it is much improvement over the last time I experienced an earthquake and only knowing later that it was one about 3.5 or so.
Pretty impressive use of existing infrastructure for public safety. Turning billions of smartphones into a global seismometer network is one of those "why didn't we do this sooner?" ideas. Sure, it's not a replacement for dedicated seismic systems, but when most of the world doesn’t have access to those anyway, this feels like a huge leap forward
- Aug 2020: "Starting today", if the accelerometer shows a trace that "may be an earthquake, it sends a signal to our earthquake detection server, along with a coarse location". "we’ll use this technology to share a fast, accurate view of the impacted area on Google Search". Alerts were additionally issued in part of the USA based on government data --https://blog.google/products/android/earthquake-detection-an...
- Mar 2022: up to three USA states now with government data, rest of the world gets alerts based on crowdsourced data. Article mentions "2+billion Android phones in use around the world" (I take that to mean "2.1 billion Google Play Services devices"). If the quake is expected to be heavy, it "Will break through Do Not Disturb settings, turn on your screen and play a loud sound" --https://crisisresponse.google/android-alerts/
- Jul 2025 (this submission): nothing seems to have changed (still govt data for the same subset of the USA), but some stats on how it's going and that accuracy is improving. It notes that, to receive alerts, users must have "location settings enabled"¹ (and internet of course). About 1/3rd of the alerts are true positives that are also received before the shaking, but 85% of people found it 5/5 very helpful
¹ This confuses me. Surely Google doesn't get your location every ~10 seconds to know whether to send your device an alert; that's too battery-draining. Maybe it sends your location a few times per day~hour and they'll just use that? Because the alternative option, if the server sends "earthquake in {geojson polygon}" to all devices, the OS could just check your (last known) location without having to care about whether you want to provide location info to apps. I have the user-level location setting turned off whenever I'm not routing/mapping because why'd I want GNSS to be running... well, for this apparently, but it never told me this
There is also the Earthquake Network (EQN) app that works on very similar principals to Google’s system - phones monitor their MEMs accelerometers, when they are left charging with their screen off, and when enough neighbouring phones detect vibrations simultaneously an earthquake is detected and apps nearby are alerted. It’s been running since 2012.
> Surely Google doesn't get your location every ~10 seconds to know whether to send your device an alert; that's too battery-draining. Maybe it sends your location a few times per day~hour and they'll just use that?
Why not?
You can use router IP address for location and WiFi to send. Minimal power consumption.
Then you don't need location services on. That only gatekeeps WiFi and Bluetooth scan results as well as GNSS. It doesn't gatekeep Internet access so they always get your IP when you talk to the server
Sadly my IP is routinely from the wrong country and I don't use a VPN
As someone who lives in an earthquake prone area it's hard to explain the spooky feeling of receiving a message about an impending earthquake 2-3 seconds before it hits. To be honest it doesn't feel helpful. There's never enough time to react properly.
I've only ever experienced a big earthquake once, which was in Bangkok a couple of months ago. And if I had known it was an earthquake, I probably would've reacted differently. Not knowing what was happening, and genuinely thinking my building was about to collapse on top of me was one of the scariest feelings in my life.
It would, because then at least you know it's an earthquake, and you kind of know what to expect. Not know it was an earthquake, and wondering why your building is shaking and making so much noise, thinking you're about to die in some freak building collapse is pretty terrifying.
So the use cases like "get down from a ladder" aren't really likely to be achieved in real life? When I'm working on a ladder, I'm not going to be checking phone notifications anyway.
Yes. The 2 types of earthquake alerts each have a unique sound that was designed specifically for earthquake alerts. They are available in the supplemental data folder that's associated with the paper (you can also hear one of these by going into the android earthquake settings and clicking on the Demo feature).
This relies on the accelerometer being always turned on - which it typically isn't when the phone screen is off.
Thats a decent amount of extra energy being used globally! And also everyone's batteries dying a little sooner.
I wonder what sample rate they have the accelerometer running at, and if it is just one axis to save power? Typically 8 bit single axis 1Hz sampling is ~10 microamps, but full 10 kHz 3 axis sampling could be 10 milliamps = 1000x more power use!
All your questions/assumptions are answered in the linked paper supplementary material: 50hz, 3 axis, only when charging. Accompanied with actual sample plots for various distances from epicenter, showing p waves and s waves.
Most MEMS accelerometers have low-power modes to generate an interrupt when movement is detected. That's probably what Google is using here (and only switching to higher-power modes when there's movement).
Had a fun experience of receiving "EARTHQUAKE ALERT. THERE WILL BE SHAKING. IT WILL BE STRONG" jolting me out of a deep slumber in Japan at 3am on a recent vacation. My wife asks the question on both of our minds: "here? back home? is there anything heavy over the bed?"
I think we both would have assumed it was for back home in California if not for the fact that we were in a country that ALSO has earthquake and tsunami history.
By the time she asked I just said "well, either way, if we don't feel anything in the next few seconds, we're fine." and then went back to sleep.
In retrospect, I didn't think to wonder where the alert came from. I suppose it must have been from the MyShake app on our iPhones which are connected to the ShakeAlert system. But maybe it was the Wireless Emergency Alert system? I'm not sure how that works overseas.
I'm still hoping someone makes an earthquake detection system where the data is just derived from people posting "Earthquake?" on Twitter/Threads/Facebook/Etc. Plot the geotagged tweets and it seems easy to get both the location and magnitude.
This reminded me about an old blogpost I read. This linked post may not be the one I remember, but it's close[1].
Back in 2011 there was an earthquake that New Yorkers felt. There were New Yorkers who read tweets of people further south on the East Coast posting about feeling an earthquake, and then the New Yorkers feeling the same earthquake a few seconds later.
There were some news outlets that picked up the story which you can find, but not exactly what OP was discussing.
The USGS created a system to do exactly this about 15 years ago. I’m not sure whether they’re still running it but at the EMSC, we've been running a similar system for many years to highlight earthquakes important to the public and improve our messaging. Twitter doesn’t give access to geotags anymore but we do manage to roughly estimate an earthquake’s location by analysis of the tweets. Estimating magnitude is much more difficult. Naturally there are some false positives but it works well overall.
We actually use the twitter detections to launch analyses of the seismic data in order to get confirmed results for events that aren’t reported yet [1] but there are some statistics for the twitter detections in the supplementary material of that article [2]. Basically, in 2016-2017 (wow, so long ago), we detected 893 earthquakes via twitter, with a median delay of 67s and a median separation from the published epicentre of 94km. Note that estimating earthquake epicentres is nontrivial anyway and so, for comparison, 10km accuracy would often be considered ok. So the twitter, I mean X, method isn’t optimal but it gets you down to the right region. Partly it’s because geocoding the tweets is inaccurate and partly it’s because people live clumped together in cities rather than smoothly spread over the surface of the earth.
When Twitter had an open API, some tech teams actually used it as an additional source for detecting incidents that internal monitoring missed (similar to how electricity grid operators watch TV to understand when demand surges are going to occur due to half time in sports games, etc).
The most useless detection system because you are either fine or buried under rubble at that point. Every real detection system attempts to catch the p-waves to warn users in real time ahead of shaking.
Certainly it's not going to compete with real time systems, but much like computing the speed of light using your microwave and a chocolate bar, it's just kind of neat to see how accurate such a system can be.
Google has this. I remember recently feeling a minor earthquake, and googling it. The message that came up said that others had felt it too in my area, and then it showed up on official databases a few hours later.
Receiving alerts should be on by default as long as you're in one of the supported countries. You can verify by going to settings, safety and emergency, earthquake alerts.
Google "AI Search" claims that the Canadian Government's early warning alert system provides the same function and that is why the Android Earthquake Alert system isn't available in Canada.
I live in a seismically active (and poor) area. Dunk on Google all you want, they're the only organization who provide earthquake alerts in my area. The government has better things to spend money on (like pervasive corruption), but Google usually sends a notification 30-60 seconds before a perceptible earthquake happens.
Note that they're also one of the only ones who can unilaterally choose to preinstall this on a majority of devices around the world. Of course I agree that it's good that they do it, for free and all, but to put it in perspective it's either each government for themselves or one of the two global superpowers that have devices with accelerometers and constant internet connectivity on every square kilometer of this oblate spheroid
Google (and Apple) has been partnering with ShakeAlert from USGS for quake reporting on west coast of US. But that takes network of seismometers and detection system.
I could see smartphone seismometers being useful for areas that don't have all that. OTOH, if phones are useful seismometers, it should be possible to make cheap, dedicated ones.
The closest one (that I know of) is approximately 1000 km away, and it does provide data about earthquakes, but only after the fact. They already have some information sharing, because I usually look up the info on usgs.gov, but almost certainly not in real time.
We’ve had this Google service in El Salvador for a while, and it’s really cool. The first time we received an earthquake alarm we felt like we were living in Japan. I never thought we would have Japanese-style earthquake alerts here.
iPhone users were a bit annoyed though, because it only worked on Android phones.
Google alone tackling this problem for 10 years and then killing it is still better than no one solving this problem and no one getting 10 years of free earthquake alerts.
big companies doing stuff for free can kill industries.
10 years is enough the ensure that any professional and company trying to make a living from earthquake early detection systems is working on something different.
yeah, someone will pop up after they inevitably kill it, but this stuff can end up delaying progress.
Case in point: Google providing Android for free killed Windows Phone, Symbian, PalmOS, Blackberry, and several other attempts to create a mobile device OS
Which was among the reasons that Google did that, not that the company would say so.
Microsoft ran Windows Phone into the ground. I could rant, but WP7 was interesting, WP7.5 was usable, WP8 and WP8.1 were pretty good, and then WM10 was very late and pretty meh. Had WM10 shown up on time and with quality, we would have a different discussion. Lots of OEMs were making phones for WP8, not many for WM10.
Symbian had flirted with Open Source, but IMHO, Nokia's fight with US carriers over shipping a SIP client and the resulting disappearance from the US market of most Nokia phones doomed Symbian... Tech journalism is dominated by US outlets and nobody reported on Symbian phones because they weren't there.
I assume you mean Palm's WebOS, cause PalmOS was pretty stylus driven and ux expectations had moved on. WebOS was neat, but Palm didn't have the corporate resources to support it.
Blackberry certainly had time and resources to compete. BB10 was supposed to be quite good, but maybe a little late.
All of these had what I like to call early mover disadvantage. Being first or at least early to a market makes it hard to adapt when the market changes. Coming in ten years late to smartphones was great for Apple and Google.
Local governments are unable to do it right because, well, they are bureaucrats who cant differentiate the left hand from their right hand. Governments IT systems are eternal money sink holes only producing power points and weekly status reports to be forever forgotten.
I assume you’ve never had the delightful experience of relying on a product Google built or acquired then let decay or killed outright because it doesn’t contribute to ad revenue and the people who cared leveraged it in their promo packet to go elsewhere.
No business has the obligation to keep running what you find useful. If it was that useful, someone else will make it.
If no one is doing it or well, I see no reason to just complain and offer no solution. If there are other solutions and Google is going to hurt or destroy "competition", that's what should be discussed.
You’re allowed to just say I’m right. When Google puts enormous amounts of resources into something like Google Home, then acquires Nest, haphazardly merges the two ecosystems while co-opting the Nest brand for unrelated products, then effectively abandons it except for the occasional update that breaks prior functionality? Sure, no obligation. But I’m not being some sort of whiny brat by pointing out that your experience with Google is driven by what will eventually become forced updates meant to drive you away from a product so they can kill it, because the internal culture/incentives promote launching and not maintaining or improving.
Someone else won't make it because they already have this for free. Well for this particular case, I think having Google product might just be better than none. But for some more critical things like typhoon warning, when people actually make decision based on data from the system, relying on a product that might get killed 5, 10 years later is worse.
If your infra is critical enough, they should nationalized, publicly owned.
> Always curious why people comment like this when they have a choice to, you know, not do it
Not OP, but it's still an important consideration - one can be both glad Google is working on this, but also cautiously optimistic given Google's history. IMO it's right to be wary of private entities taking care of what should effectively be a public service.
Did you mean cautiously pessimistic? Or maybe that's my bias from reading HN threads where this is a reliable theme in Google product threads, as well as seeing the list of killed products, while not seeing a list of kept-alive products
Recently they had a significant country wide false alarm in Israel at 3AM... There was a emergency alert cell broadcast (similar to amber alert), which caused everyone to move their phone at the same time, which was falsely detected as an earthquake, which caused an Android earthquake alert to be sent to all phones in Israel 30 seconds later. I guess they didn't plan for this scenario
Edit: Arstechina article seems to mention this: "only three were false positives. One of those was triggered by a different system sending an alert that vibrated a lot of phones"
I heard it was the cell broadcast which caused the phones to vibrate at the same time, not people picking them up.
That would be quite an implementation flaw if it didn't account for the phone's own vibrations. Lots of countries use widespread emergency alert messages frequently.
What if it's actually another phone's vibration laying next to each other? Me and my wife leave our phones together when going to sleep.
not only that but my android phone sleeps together with my android reading tablet
Awww that's cute. Did they produce little baby Android devices?
They fixed this bug, we had plenty of emergency cell broadcasts since the false alarm.
if (!isShakerRunning) { sendAlert() }
seems like a pretty obvious thing to add
Only obvious once you've thought of it I suppose...
It's obvious to anyone that's written software for hardware sensors to ensure the hardware itself is not going to interfere with readings of other sensors. Don't tell the motor to run if the motor is already running type of issues. It's a fairly common bozo check
Doesn't seem that obvious to me
Typical Google product. Reminds me of a person who put a bunch of phones in a car and drove which caused Google maps to wrongly show traffic in that area. It was deliberately done though as an experiment
Better yet, he put them in a child's wagon and carted them around Berlin
One of the last before-times human interest stories: https://www.wired.com/story/99-phones-fake-google-maps-traff...
Note that those are three completely false events. The survey results Google published show 15% of people not feeling any shaking (neither strong nor light). That's still a good figure, but reading there were only 3 false positives gave me the impression that you're basically always in for a ride when you get the alert and it's not that miraculously accurate either
Surprised they don't do some signal processing on all the IMU signals to see if they correlate (to within a rotation matrix), and if the timing of shaking at different locations is consistent with the distance to some (solvable) epicenter.
The whole country moving phones in random directions at exactly the same time isn't what an earthquake signature should look like.
I don't get it. I thought earthquake alerts were meant to trigger _before_ the earthquake arrives. If it happens 30 secondes after detecting vibrations, not considering the false positive, it can only mean "hey, you just felt, or are feeling an earthquake, hope you're sheltered".
Earthquake early warning only works because the internet is faster than seismic waves. It only works for people who are far enough from the epicenter.
This is how I experienced the Google Earth Alerts in Wellington, NZ which experiences frequent earthquakes. Earthquake happens. 30 seconds to five minutes later I get an alert from Google.
The best earthquake indicator is an old house with wooden windows that you'll hear rattling five seconds before you get to feel it.
Here in San Clemente, CA the one alert I've had (earlier this year) actually gave me about 10 seconds warning for some pretty strong shaking!
A faster moving small quake (p-wave) will precede the bigger, more damaging quake. This system detects the p-wave and alert people hoping they can get out in time before the big quake hit.
To clarify: pressure waves travel faster than shear waves. Both start at the same time and the same place (epicenter).
Earthquake early warning only works because the internet is faster than seismic waves. It only works for people who are far enough from the epicenter.
This is when the earthquake is a few hundred kilometers away, such as in Mexico City, where most earthquakes occur off the coast and the waves take a few minutes to reach the city.
The first alert made the phones vibrate (or people touching them), which got falsely detected as an earthquake.
Yes, and they sent a notification apparently after the detection. That is not a warning system
Earthquake early warning only works because the internet is faster than seismic waves. It only works for people who are far enough from the epicenter.
Somewhat relatedly, I support a service with global scale traffic. Whenever there's an earthquake in APAC, we get a traffic spike, like 100x normal for that time of night. I'm pretty sure it's incoming alerts waking people up / checking where the epicenter is and if they need to run from any tsunami or flooding, but it's still really hard to scale that kind of thing up for regional demand when it spikes that hard!
I was thinking the same thing. A taylor swift concert where she tells everyone to sway their phones in unison might trigger this
I would assume the detections would need to come from a larger area than that to trigger it, like city sized.
From memory I think earthquake oscillation profiles are higher frequency then swaying side to side is likely to be.
Kind of ironic that the alert caused the earthquake warning
Maybe it was due to the blasts from their own ordenances
This is really cool, and it smells like old-school Google, in a good way, like "let's do this because we can". It feels like it's been a while since something coming out of Google Engineering is meaningful and not designed to unlock new existential creeps, so, well done I guess.
No ads, no creepy monetization angle (at least not yet), just a genuinely useful system that leverages something only Google could realistically pull off. Feels rare these days, but really nice to see.
Nowadays I just assume these "nice to have" features exist in order to get users to enable location services so that they can be permanently tracked. Very cynical of me maybe. But there's never an option to "enable location sevices for this use only"-kind of setting. It's always the globally enabled one.
I agree and wish it would be a separate app you can install or not instead of part of the system.
For Android's Earthquake Alerts System to work, Google requires your phone to "always" send location.
If government requests access to this location data, Google is compelled to provide it. There is no conspiracy or anything cynical about it.
Cynical: People who makes decisions at Google push for features that trick people into enabling system-wide always-on location services that they otherwise wouldn't. For functionality that shouldn't actually require system-wide always-on location services. The primary goal being to sell more ads, not to help with earthquakes.
A conspiracy: Governments push or force or encourage Google to do it in order to have access to location data on citizens.
I no longer live near any epicenter but this was the biggest feature I admire as a non-Android user.
A few years back I was woken by a shake in HongKong. To confirm was a quake I found my Android phone and sure enough Google had registered a quake. It was one far inland and <5 IIRC. Creepiness or not as someone who helped after Sichuan 2008 quake these kind of systems can save lives.
> "Of those roughly 1,300 events that triggered alerts, only three were false positives. One of those was triggered by a different system sending an alert that vibrated a lot of phones, something that should be relatively easy to compensate for in software. The other two were both due to thunderstorms, where heavy thunder caused widespread vibrations centered on a specific location. This led the team to better model acoustic events, which should prevent something similar from happening in the future."
https://arstechnica.com/science/2025/07/how-android-phones-b...
Do the range of detectable acoustic sources include military jets, drones, and bomb blasts (i.e., gauging effectiveness of targeting?) I don't know what I'm supposed to think of tech companies turning gadgets into remote-root physics sensors without user consent. Maybe I'm reflexively cynical; I can't trust a FAANG with yet another side-channel attack, *even if* the first (public) application is, on appearance, a life-saving unalloyed good.
I have received a few earthquake alerts (Greece). one was for a significant 5.2 earthquake about a month ago, and the notification arrived about one minute earlier or so. It woke me up , and i was able to experience the entire duration of the earthquake. Pretty cool if they were using the new system and i was impressed at the time.
Curious if the alert mentioned how strong the shaking would be, or was it more of a general "earthquake incoming" type message?
there is something like "moderate shaking expected" in the alert
Last time we had a reasonably powerful one in Portugal, I grabbed my phone the house was still shaking and it already had the Android warning!
I was surprised as I didn't even know it was a thing.
I was also a bit spooked as it was in the ocean, near the coast, and when I turned on my FM radio as we were always taught in school, all I heard were pre-recorded music programs.
Turns out it didn't meet the threshold for a warning so the authorities didn't issue any message about tsunami danger. I think they should've anyway, as I wasn't the only one that had that thought.
yeah I've ran into bunch of people who were scared for real and drove up to higher elevation in the middle of the night.
If you'd search online you could have known quickly there was a negative Tsunami notice, but I get that this is just not feasible for everyone (or everytime).
Feel like the severe weather SMS etc are working quite decent, I wish they'd expand that for those sort of things as well (like there was an earthquake, this is what you should do next).
The Android notification was a bit odd, because it's not trivial to get back to that notification if you've just skipped/acknowledged it.
Few months back we experienced an earthquake. I got an alert on my Android, which at first I was confused about but took me a second to process that there is a possible earthquake and then we ran out and it was a 5.2 magnitude earthquake. So it is much improvement over the last time I experienced an earthquake and only knowing later that it was one about 3.5 or so.
Even a few seconds of warning can make a big impact, especially if you're indoors
Pretty impressive use of existing infrastructure for public safety. Turning billions of smartphones into a global seismometer network is one of those "why didn't we do this sooner?" ideas. Sure, it's not a replacement for dedicated seismic systems, but when most of the world doesn’t have access to those anyway, this feels like a huge leap forward
I thought this was ancient but apparently not. Searched back a bit:
- Feb 2016: third-party app starts doing it, so you had to go out of your way to install it but it may hit critical mass at some point. This is probably what I was thinking of --https://earthquakes.berkeley.edu/blog/2016/02/11/seismic-sen...
- Aug 2020: "Starting today", if the accelerometer shows a trace that "may be an earthquake, it sends a signal to our earthquake detection server, along with a coarse location". "we’ll use this technology to share a fast, accurate view of the impacted area on Google Search". Alerts were additionally issued in part of the USA based on government data --https://blog.google/products/android/earthquake-detection-an...
- Mar 2022: up to three USA states now with government data, rest of the world gets alerts based on crowdsourced data. Article mentions "2+billion Android phones in use around the world" (I take that to mean "2.1 billion Google Play Services devices"). If the quake is expected to be heavy, it "Will break through Do Not Disturb settings, turn on your screen and play a loud sound" --https://crisisresponse.google/android-alerts/
- Jul 2025 (this submission): nothing seems to have changed (still govt data for the same subset of the USA), but some stats on how it's going and that accuracy is improving. It notes that, to receive alerts, users must have "location settings enabled"¹ (and internet of course). About 1/3rd of the alerts are true positives that are also received before the shaking, but 85% of people found it 5/5 very helpful
¹ This confuses me. Surely Google doesn't get your location every ~10 seconds to know whether to send your device an alert; that's too battery-draining. Maybe it sends your location a few times per day~hour and they'll just use that? Because the alternative option, if the server sends "earthquake in {geojson polygon}" to all devices, the OS could just check your (last known) location without having to care about whether you want to provide location info to apps. I have the user-level location setting turned off whenever I'm not routing/mapping because why'd I want GNSS to be running... well, for this apparently, but it never told me this
There is also the Earthquake Network (EQN) app that works on very similar principals to Google’s system - phones monitor their MEMs accelerometers, when they are left charging with their screen off, and when enough neighbouring phones detect vibrations simultaneously an earthquake is detected and apps nearby are alerted. It’s been running since 2012.
[1] https://sismo.app
I wouldn't be surprised if coarse location on Android doesn't even bother turning on the GPS. You can do a lot with cell towers and wifi ssids
> Surely Google doesn't get your location every ~10 seconds to know whether to send your device an alert; that's too battery-draining. Maybe it sends your location a few times per day~hour and they'll just use that?
Why not?
You can use router IP address for location and WiFi to send. Minimal power consumption.
Then you don't need location services on. That only gatekeeps WiFi and Bluetooth scan results as well as GNSS. It doesn't gatekeep Internet access so they always get your IP when you talk to the server
Sadly my IP is routinely from the wrong country and I don't use a VPN
As someone who lives in an earthquake prone area it's hard to explain the spooky feeling of receiving a message about an impending earthquake 2-3 seconds before it hits. To be honest it doesn't feel helpful. There's never enough time to react properly.
I've only ever experienced a big earthquake once, which was in Bangkok a couple of months ago. And if I had known it was an earthquake, I probably would've reacted differently. Not knowing what was happening, and genuinely thinking my building was about to collapse on top of me was one of the scariest feelings in my life.
> Not knowing what was happening, and genuinely thinking my building was about to collapse on top of me was one of the scariest feelings in my life.
This what earthquake causes sometimes. So knowing that it was an earthquake would not change it, would it?
It would, because then at least you know it's an earthquake, and you kind of know what to expect. Not know it was an earthquake, and wondering why your building is shaking and making so much noise, thinking you're about to die in some freak building collapse is pretty terrifying.
So the use cases like "get down from a ladder" aren't really likely to be achieved in real life? When I'm working on a ladder, I'm not going to be checking phone notifications anyway.
Does the phone make a different kind of noise, so you'd potentially associate that with "get off the ladder quick"!?
Yes. The 2 types of earthquake alerts each have a unique sound that was designed specifically for earthquake alerts. They are available in the supplemental data folder that's associated with the paper (you can also hear one of these by going into the android earthquake settings and clicking on the Demo feature).
Hmm, Earthquake alerts aren't available: Location switch is off.
But I like it that way.
This relies on the accelerometer being always turned on - which it typically isn't when the phone screen is off.
Thats a decent amount of extra energy being used globally! And also everyone's batteries dying a little sooner.
I wonder what sample rate they have the accelerometer running at, and if it is just one axis to save power? Typically 8 bit single axis 1Hz sampling is ~10 microamps, but full 10 kHz 3 axis sampling could be 10 milliamps = 1000x more power use!
All your questions/assumptions are answered in the linked paper supplementary material: 50hz, 3 axis, only when charging. Accompanied with actual sample plots for various distances from epicenter, showing p waves and s waves.
Most MEMS accelerometers have low-power modes to generate an interrupt when movement is detected. That's probably what Google is using here (and only switching to higher-power modes when there's movement).
It only runs when a phone is plugged in and stationary.
Had a fun experience of receiving "EARTHQUAKE ALERT. THERE WILL BE SHAKING. IT WILL BE STRONG" jolting me out of a deep slumber in Japan at 3am on a recent vacation. My wife asks the question on both of our minds: "here? back home? is there anything heavy over the bed?"
I think we both would have assumed it was for back home in California if not for the fact that we were in a country that ALSO has earthquake and tsunami history.
By the time she asked I just said "well, either way, if we don't feel anything in the next few seconds, we're fine." and then went back to sleep.
In retrospect, I didn't think to wonder where the alert came from. I suppose it must have been from the MyShake app on our iPhones which are connected to the ShakeAlert system. But maybe it was the Wireless Emergency Alert system? I'm not sure how that works overseas.
Last time I received it I was on a 14th floor and I was terrified. Longest seconds of my life while I waited the P wave to arrive.
Is there a way to get alerts or contribute data with IOS?
I'm still hoping someone makes an earthquake detection system where the data is just derived from people posting "Earthquake?" on Twitter/Threads/Facebook/Etc. Plot the geotagged tweets and it seems easy to get both the location and magnitude.
I remember many years ago seeing exactly this project being led by a researcher in Chile [1].
It's not really a new idea, i don't know what happened to this project though.
[1] https://portaluchile.uchile.cl/noticias/119844/twitter-ayuda...
I don't think that is fast enough since the window for alert is seconds to minute. The alert lets people get to safety and stop systems like trains.
Tracking social would be useful for plotting where quake was felt.
This reminded me about an old blogpost I read. This linked post may not be the one I remember, but it's close[1].
Back in 2011 there was an earthquake that New Yorkers felt. There were New Yorkers who read tweets of people further south on the East Coast posting about feeling an earthquake, and then the New Yorkers feeling the same earthquake a few seconds later.
There were some news outlets that picked up the story which you can find, but not exactly what OP was discussing.
[1] https://www.ralphehanson.com/2011/08/25/earthquakes-social-m...
The USGS created a system to do exactly this about 15 years ago. I’m not sure whether they’re still running it but at the EMSC, we've been running a similar system for many years to highlight earthquakes important to the public and improve our messaging. Twitter doesn’t give access to geotags anymore but we do manage to roughly estimate an earthquake’s location by analysis of the tweets. Estimating magnitude is much more difficult. Naturally there are some false positives but it works well overall.
[1] https://www.usgs.gov/media/audio/shaking-and-tweeting-usgs-t...
We actually use the twitter detections to launch analyses of the seismic data in order to get confirmed results for events that aren’t reported yet [1] but there are some statistics for the twitter detections in the supplementary material of that article [2]. Basically, in 2016-2017 (wow, so long ago), we detected 893 earthquakes via twitter, with a median delay of 67s and a median separation from the published epicentre of 94km. Note that estimating earthquake epicentres is nontrivial anyway and so, for comparison, 10km accuracy would often be considered ok. So the twitter, I mean X, method isn’t optimal but it gets you down to the right region. Partly it’s because geocoding the tweets is inaccurate and partly it’s because people live clumped together in cities rather than smoothly spread over the surface of the earth.
[1] https://www.science.org/doi/10.1126/sciadv.aau9824 [2] https://www.science.org/action/downloadSupplement?doi=10.112...
I swear Twitter or Google was working on this?
https://scistarter.org/the-twitter-earthquake-detection-prog...
I did find this and some papers that seem related
When Twitter had an open API, some tech teams actually used it as an additional source for detecting incidents that internal monitoring missed (similar to how electricity grid operators watch TV to understand when demand surges are going to occur due to half time in sports games, etc).
Not just electricity, water too.
Many people go to the bathroom when their favorite show ends or when it's halftime during a sports match.
The most useless detection system because you are either fine or buried under rubble at that point. Every real detection system attempts to catch the p-waves to warn users in real time ahead of shaking.
Certainly it's not going to compete with real time systems, but much like computing the speed of light using your microwave and a chocolate bar, it's just kind of neat to see how accurate such a system can be.
Google has this. I remember recently feeling a minor earthquake, and googling it. The message that came up said that others had felt it too in my area, and then it showed up on official databases a few hours later.
Did people opt into having their data analyzed by Google? They seemed to have skipped over that is super important part about consent.
As always.
I wonder if this is why the service isn't available in Canada.
Google "AI search" results claim that because Canada has its own early warning system, Android Earthquake Alerts aren't available in Canada.
Call me a hater but I think Google has too much might already. Give em more leads to... you know the story.
It's just a matter of time until that warning is followed up by ads for generators
I was about to install the app to help... but I guess this is part of the general dragnet so no need?
> To receive alerts, users must have Wi-Fi and/or cellular data connectivity, and both Android Earthquake Alerts and location settings enabled.
Enabling location settings is a global configuration. Can’t this be configured just for this purpose?
From the coverage area map, it seems japan and indonesia (two countries with lots of earthquakes) aren't covered by this service. I wonder why.
Is this app native to Android OS xx ??
Is it by default on ? How can i check if its actually working ?
So many questions! None detailed in the article...
Receiving alerts should be on by default as long as you're in one of the supported countries. You can verify by going to settings, safety and emergency, earthquake alerts.
Thanks, looklike i'm not in supported region :(
If you're ok with sharing, could you say which country?
It's not supported in Canada. I don't know why.
Google "AI Search" claims that the Canadian Government's early warning alert system provides the same function and that is why the Android Earthquake Alert system isn't available in Canada.
Vietnam, tbf we don't have a lot of earthquake here, but still, it would be nice to have.
thanks!
You could probably also use this system for seismic imaging too.
What exactly is their prediction model?
Really curious to know.
Is there an iOS app for it?
Obligatory XKCD: https://xkcd.com/723/
Company that tracks people to send them ads wants you to sign up for some 10 year old technology for .. saving lives!
It would be great, if I had any reason to trust google.
It'd be cool if they'd partner with governments to send the alerts out through the same means as amber alerts and such.
WEA has higher latency IIRC
> Android Earthquake Alerts: A global system for early warning
Fixing bugs in Google products is no fun. Better make something else what nobody needs.
>"[...] and builds user trust with each successful alert"
So the company notorious for killing projects is going to tackle infrastructure grade systems? I don't trust Google to tackle this problem.
I live in a seismically active (and poor) area. Dunk on Google all you want, they're the only organization who provide earthquake alerts in my area. The government has better things to spend money on (like pervasive corruption), but Google usually sends a notification 30-60 seconds before a perceptible earthquake happens.
Note that they're also one of the only ones who can unilaterally choose to preinstall this on a majority of devices around the world. Of course I agree that it's good that they do it, for free and all, but to put it in perspective it's either each government for themselves or one of the two global superpowers that have devices with accelerometers and constant internet connectivity on every square kilometer of this oblate spheroid
Google (and Apple) has been partnering with ShakeAlert from USGS for quake reporting on west coast of US. But that takes network of seismometers and detection system.
I could see smartphone seismometers being useful for areas that don't have all that. OTOH, if phones are useful seismometers, it should be possible to make cheap, dedicated ones.
The closest one (that I know of) is approximately 1000 km away, and it does provide data about earthquakes, but only after the fact. They already have some information sharing, because I usually look up the info on usgs.gov, but almost certainly not in real time.
We’ve had this Google service in El Salvador for a while, and it’s really cool. The first time we received an earthquake alarm we felt like we were living in Japan. I never thought we would have Japanese-style earthquake alerts here.
iPhone users were a bit annoyed though, because it only worked on Android phones.
Google alone tackling this problem for 10 years and then killing it is still better than no one solving this problem and no one getting 10 years of free earthquake alerts.
big companies doing stuff for free can kill industries.
10 years is enough the ensure that any professional and company trying to make a living from earthquake early detection systems is working on something different.
yeah, someone will pop up after they inevitably kill it, but this stuff can end up delaying progress.
Case in point: Google providing Android for free killed Windows Phone, Symbian, PalmOS, Blackberry, and several other attempts to create a mobile device OS
Which was among the reasons that Google did that, not that the company would say so.
That and the fact that all those other OS were a complete disaster.
Microsoft ran Windows Phone into the ground. I could rant, but WP7 was interesting, WP7.5 was usable, WP8 and WP8.1 were pretty good, and then WM10 was very late and pretty meh. Had WM10 shown up on time and with quality, we would have a different discussion. Lots of OEMs were making phones for WP8, not many for WM10.
Symbian had flirted with Open Source, but IMHO, Nokia's fight with US carriers over shipping a SIP client and the resulting disappearance from the US market of most Nokia phones doomed Symbian... Tech journalism is dominated by US outlets and nobody reported on Symbian phones because they weren't there.
I assume you mean Palm's WebOS, cause PalmOS was pretty stylus driven and ux expectations had moved on. WebOS was neat, but Palm didn't have the corporate resources to support it.
Blackberry certainly had time and resources to compete. BB10 was supposed to be quite good, but maybe a little late.
All of these had what I like to call early mover disadvantage. Being first or at least early to a market makes it hard to adapt when the market changes. Coming in ten years late to smartphones was great for Apple and Google.
Local government is disincentivized to build a system because they already has something up for free from the Big Tech.
Case in point: No one built a group-chat for our government officials cause Telegram is free.
Local governments are unable to do it right because, well, they are bureaucrats who cant differentiate the left hand from their right hand. Governments IT systems are eternal money sink holes only producing power points and weekly status reports to be forever forgotten.
Are you always this salty, so it is only certain topics that make you do this?
Always curious why people comment like this when they have a choice to, you know, not do it
I assume you’ve never had the delightful experience of relying on a product Google built or acquired then let decay or killed outright because it doesn’t contribute to ad revenue and the people who cared leveraged it in their promo packet to go elsewhere.
No business has the obligation to keep running what you find useful. If it was that useful, someone else will make it.
If no one is doing it or well, I see no reason to just complain and offer no solution. If there are other solutions and Google is going to hurt or destroy "competition", that's what should be discussed.
You’re allowed to just say I’m right. When Google puts enormous amounts of resources into something like Google Home, then acquires Nest, haphazardly merges the two ecosystems while co-opting the Nest brand for unrelated products, then effectively abandons it except for the occasional update that breaks prior functionality? Sure, no obligation. But I’m not being some sort of whiny brat by pointing out that your experience with Google is driven by what will eventually become forced updates meant to drive you away from a product so they can kill it, because the internal culture/incentives promote launching and not maintaining or improving.
Someone else won't make it because they already have this for free. Well for this particular case, I think having Google product might just be better than none. But for some more critical things like typhoon warning, when people actually make decision based on data from the system, relying on a product that might get killed 5, 10 years later is worse.
If your infra is critical enough, they should nationalized, publicly owned.
But isn’t that also what makes us special? Like not everyone is the same and stuff?
> Always curious why people comment like this when they have a choice to, you know, not do it
Not OP, but it's still an important consideration - one can be both glad Google is working on this, but also cautiously optimistic given Google's history. IMO it's right to be wary of private entities taking care of what should effectively be a public service.
But how boring and unhelpful to have someone post it on every product that Google builds.
[flagged]
> cautiously optimistic given Google's history
Did you mean cautiously pessimistic? Or maybe that's my bias from reading HN threads where this is a reliable theme in Google product threads, as well as seeing the list of killed products, while not seeing a list of kept-alive products