The fact that they can use the delay of the satellite signal to predict rainfall, is pretty impressive. It immediately links this app to more data for weather prediction, which might lead to better weather predictions.
I wonder how they would measure that on an android phone. Does android expose such low-level delays to the user? Is this not handled by a GNSS chip in hardware, or the OS driver/kernel?
The idea is interesting, but there are some disadvantages of the app itself (see below). Maybe developers can read this and implement some changes.
- the phone should have non-obscured view of the sky. Thus having a phone charging and logging the data while it is on the desk inside a house doesn't work that well (red or orange indicator of the measurements quality, even near a big window).
- speaking of charging, there are no settings which allow to instruct the app to be dormant and automatically start recording when phone is charging.
- Account and login process is needed to upload the collected data. Why is it even the case? Can't the data be uploaded anonymously with just some unique phone identifier, or without one just relying on coordinates and other GNSS related measurements. Data can be cross-checked with others nearby and outliers can be removed just from that. There is no real need to know my name, email and create a leader-board or at least have an option of anonymous upload.
Hypothetically they could be using an account / login process to allow them to filter out measurements from devices that for whatever reason (hardware / environment / etc) to be sending low quality data. Alternatively they could have used device id without having the user setup an account but that actually feels more intrusive.
Not saying that is a reason but it could be one of the non-user visible rationales.
Here is an example of pretty stunning amount of visual detail provided by one of popular GPS data app on Android: https://play-lh.googleusercontent.com/lT3TpHarUx89Z7HIf043aU...
At least I can confirm it works will all navigation constellations on Samsung S10.
I had a lot of fun with that gps status app when I first got a smart phone - you could hold your phone up to the window while flying (to get gps signal) and see the 630mph or whatever speed you were going.
I think those are the old COCOM limits which are no longer in effect. I believe that currently the only limit is 600 m/s (which is nice, since when there were both speed and altitude limits some manufacturers OR'd them and some manufacturers AND'd them, so it wasn't clear if altitude alone would be enough to disable a device until you found out the hard way)
Seriously, that was an interesting question to research for a moment - is it really an iPhone (from the visual) and if: what type? I think its the version 6 regular size or an SE 2020, because the button until version 5 had a symbol on it and some like the 6s seems to have a different postioning of the power button, others don't seem to have the proportions. And then again, its probably a digital rendering (search for "mockup iphone"). Last but not least the visuals of apple iPhones have been copied countless times to cheap android knockoffs.
Interesting, I'm wondering why they didn't even mention this availability in their website, I would have turned on that setting just to farm more data for their app.
> The application works by detecting lit pixels in the phone’s camera when no light is entering. These pixels are lit as result of cosmic rays, local background radiation, or sometimes just noise.
Interestingly the app is made in Unity. A couple of oddities, but having this leader board function was definitely an interesting idea.
Sadly either my old phone's GPS is not great or my windowsill's viability to satellites just isn't great. It'll be very interesting to see the data pipelines they have to clean up the mountains of data they're going to be receiving.
This reminds me of something I read long ago when I was just a child. A computer magazine had a project where you used an FM radio, a computer probably a Commodore, and a plotter (who had one of those?!). The FM signal could detect meteor strikes in the atmosphere, you wrote a bit of code, the plotter mapped the strikes.
Interesting project, can't wait to help ESA out a bit.
Seems like the phone will just track the satellites via GPS, so it's not even that battery intensive. Overnight left charging could provide a lot of data.
The fact that there's a leaderboard makes it even more gamified.
Don't phones use a combination of various global positioning systems these days, including glonass and galileo?
Actually how much detailed GPS information can phones access? Most 'common' apps will use the wider 'location services', which combine GNSS data with things like known wifi points and 3/4/5g radio towers to provide better accuracy.
Only on Google store, and an open source google store client (Aurora) just says "failed to fetch app details"... why is it so hard to just put an apk on your website if you actually want people to use your app?
> As well as helping to create new Earth and space weather forecasting models, participants are also in with the chance to win prizes
They seem to be quite keen on getting users and I'd be interested in the data myself (don't care for prizes), but then they make it a Google ecosystem exclusive?
Edit: sent them an email using the address on the contact page. Let's see.
Or you just put the apk behind a big obvious button saying "Advanced users" and a checkbox for users to affirm they are comfortable installing/working with APKs themselves, which is required before the download can start.
You might still get the questions but at least then support teams can legitimately say that kind of advice is not provided as the users agreed to handle such matters.
You put a GIANT Google Play link , and a smaller link for alternative stores that will open a plain html page dense of information and with big warnings to scare away the non-tech people.
Basically the aab is supposed to be smaller based on utilizing shared resources. But really it seems like Google is using aab to force devs into using Google's key/ signing management. I have no idea if aab format prevents an app from being in Aurora. It was just my random guess that the format might be causing an issue.
A small personal note: I totally favor scientific crowd-experiment, but ONLY if done in FLOSS and public terms. SmartPhones these days are surveillance devices maintained and paid by the formal owner while they serve far more their vendor and other player behind, with the formal owner as the last in the pyramid.
In that sense I have to refuse because I have to refuse the device used even by a formally FLOSS and public experiment. Of course asking to buy or buy and offer sensors devices to the masses is unfeasible BUT it's perfectly feasible, just, needed, asking to IMPOSE open hardware without lock-ins, FLOSS code on them and services with public APIs as a State law, gradually growing to exit actual extremely dangerous and sorry situation. Scientific institutions are among those best qualified to took such public statements. Avoiding them means avoiding part of the Scientific duty, witch is doing their best to improve the society.
This is why Nokia was gutted. Sony phones with Sailfish is the furthest separation you can achieve from the freaks responsible for this situation and still remain a participant in the information age.
People don't understand the cruelty we have grown accustomed to.
Anything "new" and "that demand a substantial social change" is hard, at least for start and for a not so short period of time... If no one start change never happen, if someone start...
Software developers having been calling for FLOSS for decades at this point, and it's clear that the average person will only use FLOSS if it's mandated or clearly a better product in some way. Posing a fringe ethical dilemma alone will not get it done... the vast majority of people (including people collecting data for research) do not even care a little bit. They have a long list of other problems to solve first, and if FLOSS gets in the way of any of those they'll gladly forgo it.
Casual people and scientist are different cohorts of population, casual people do not have the culture to comprehend, most do not even understand the difference between a third-partly hosted web-app and a local one just because they look at the same screen, most do not understand why it's absurd and bad print a document, scan in to an image and send the image by mail etc BUT those normally are not scientist.
Scientist normally like to learn, so if someone explain them something interesting they learn it in means.
Population always follow.
Just look at the French Revolutions the Sans-culottes was driven by bourgeois, similarly on the other sides soldiers are peoples/"commoners" driven by other bourgeois and aristocrats. The people have chosen a side or another, and they are pushed to chose by both sides. So far I do not see much on the FLOSS side since there is no community anymore and most just live on someone else computer. Who better than Scientist can correct that aim? Peoples in Humanities surely are more listening but they mostly lack enough skill to bridge the gap between the philosophy (in the classical sense of "the why") and the practice, scientist normally can.
Effectively the problem is that the atmosphere (mostly ionosphere) distorts the ranges a bit. Different frequencies distort differently so you can use data from two to compensate.
IIRC, used to work in a gps accuracy firm many many years ago
I had an idea a few months ago that this just made me remember. I was thinking that I could come up with some way of mining crypto but instead of guessing hashes I would be proving that I was actively sending weather/sky data. So people actively participating in the sensor network would get rewarded. Proof of… sky?
There have been tons of fake apps that claim to do everything from microwaving food to measuring blood pressure. Not everyone knows everything about how things work in principle. I could imagine that GP doesn't know whether this is anywhere near realistic and just assumes it's a joke, maybe?
Sigh. You're putting your processing power to work for a wealthy corporation in the hopes they will give you scraps in return, and / or that those scraps will become worth more in the future.
Crypto idealism is long dead, it's all about making money now.
We don't even have apps to turn an Android/iOS device into a proper native webcam using UVC over USB2.0 & all attempts at RTSP/other sttreaming protocols just feel hacky & laggy.
afaik newer device kernels in android implements f_uvc function for it's usb gadget interface but i'm not sure you can use this without root.
so, apps can't create this functionality without root.
also i'm not sure it's works even apps can get root.
i didn't tried yet because my phone not support f_uvc and released kernel sources sucks.
That's great for monitoring & setting up shots but still doesn't show up as a native device. What I usually do is hook a stream upto OBS or join via video on the phone & stay muted+silent on my laptop.
I wonder how they would measure that on an android phone. Does android expose such low-level delays to the user? Is this not handled by a GNSS chip in hardware, or the OS driver/kernel?