[HN Gopher] Android Earthquake Alerts: A global system for early...
___________________________________________________________________
Android Earthquake Alerts: A global system for early warning
Author : michaefe
Score : 130 points
Date : 2025-07-22 18:23 UTC (4 hours ago)
(HTM) web link (research.google)
(TXT) w3m dump (research.google)
| fusionadvocate wrote:
| >"[...] 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.
| homebrewer wrote:
| 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.
| ianburrell wrote:
| 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.
| homebrewer wrote:
| 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.
| Aachen wrote:
| 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
| el-salvador wrote:
| 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.
| bitpush wrote:
| 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
| baxtr wrote:
| But isn't that also what makes us special? Like not everyone
| is the same and stuff?
| transcriptase wrote:
| 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.
| thezilch wrote:
| 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.
| transcriptase wrote:
| 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.
| thewebguyd wrote:
| > 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.
| thezilch wrote:
| But how boring and unhelpful to have someone post it on
| every product that Google builds.
| Aachen wrote:
| > 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
| kccqzy wrote:
| 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.
| kobalsky wrote:
| 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.
| seydor wrote:
| 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.
| IvyMike wrote:
| 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.
| ianburrell wrote:
| 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.
| Robelius wrote:
| 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...
| nemomarx wrote:
| 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
| jerojero wrote:
| 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...
| irvymike wrote:
| 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"
| underdeserver wrote:
| I heard it was the cell broadcast which caused the phones to
| vibrate at the same time, not people picking them up.
| ls-a wrote:
| 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
| kylecazar wrote:
| Better yet, he put them in a child's wagon and carted them
| around Berlin
| Miraste wrote:
| 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.
| irvymike wrote:
| They fixed this bug, we had plenty of emergency cell
| broadcasts since the false alarm.
| Aachen wrote:
| 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
| CGMthrowaway wrote:
| I was thinking the same thing. A taylor swift concert where she
| tells everyone to sway their phones in unison might trigger
| this
| srameshc wrote:
| 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.
| Aachen wrote:
| 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
| perihelions wrote:
| > _" 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.
| mnky9800n wrote:
| You could probably also use this system for seismic imaging too.
| londons_explore wrote:
| 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!
| duskwuff wrote:
| 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).
| marcsto2 wrote:
| It only runs when a phone is plugged in and stationary.
| irvymike wrote:
| 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.
| CommenterPerson wrote:
| Company that tracks people to send them ads wants you to sign up
| for some 10 year old technology for .. saving lives!
___________________________________________________________________
(page generated 2025-07-22 23:00 UTC)