PressureNET 2.1 Released: the Distributed Barometer Network For Android

cryptoz writes "Cumulonimbus has released a new version of their open source, global barometer network. The network is built around an Android app called pressureNET which uses barometric sensors in new phones (such as the Nexus 4, Galaxy Nexus, Galaxy S3, Note, and others) in order to build the comprehensive network. They plan to use the data to improve short-term weather prediction, and the gives a teaser of the new data visualization tool they are building."

Comments

neat

[Anonymous Coward]

Already installed. Are there any more of these distributed tools for phones?

Re:neat

[gbjbaanb]

there was a distributed "3g survey" mapping project that the BBC tried a while back and its still on-going using OpenSignalMaps which will no doubt come back into fashion as the 4G rollout starts up in earnest.

oh crap...

[Anonymous Coward]

now they can verify where we are based on how our phone sensors match up with barometric maps.

too...much...pressure...

Very nice

[medcalf]

I've been thinking that temperature and pressure sensors would be a great app enabler on cell phones. Kudos to Google and the Android device makers then.

Re:Very nice

[alexander_686]

I think it’s just pressure – and I am not sure how valuable a tempura gauge would be. I thinking my body temperature would throw off the readings to the point that they were meaningless.

Re:Very nice

[amRadioHed]

Pressure yes, but temperature would be really pretty useless. The reading would always be some random, ever changing mix of the heat generated by the phone itself, your body temperature, and the ambient temperature. What exactly could you do with that?

Re:Very nice

[Charliemopps]

Infrared temp sensors can read temperatures at a distance. Put one on the front and back, and you could easily figure out where the user is and use the temp on the other side. Of course, that'd only get the temp of the room your in. But if you compared multiple readings all day, and only took ones that were close to what the local airport said, you could then throw out the rest. Obviously it's not 72 in Canada right now... but then the user walks outside and it drops suddenly to 25 degrees, and you have a reading.

Re:Very nice

[19thNervousBreakdown]

A tempura gauge sounds awesome!

Warning! You are dangerously close to delicious food. Rapid cholesterol rises eminent.

Re:Very nice

[medcalf]

All sensors are subject to error, of course. Speaking of which, you neglected to mention instrument error from design and manufacturing as possible causes of it. I suspect that there are measures you could take to correct them, but I'd be far more interested in things like being able to detect differences than in absolute values. We're not really going to get a perfect distributed measurement system, and even if we tried, we'd be creating many of the same problems that exist with current measurement systems (where many of the stations are poorly sited or are simply not terribly accurate instruments). But wouldn't it be interesting to be able to do things like having your phone automatically adjust your electronic thermostat while you're on your way home, based on geofencing and the temperature it detects locally? Or wouldn't t be interesting to have your phone act as a night-time fire alarm by sounding at full volume if the temperature exceeds a certain threshold (which wouldn't have to be terribly accurate; within 10 degrees would be fine)? Or what about figuring out the optimal place to site a thermostat based on the current location of the thermostat, its setting, its location, the temperatures at other locations in the building and the desired temperature? Or what about being able to set the phone on a person's body, and get their temperature and (via the motion sensor) heart rate? Or being able to set it at various cracks and such like doors and windows to see if you have voids in your insulation, or other places you could fix your house to be more energy efficient?

Re:Very nice

[amRadioHed]

The temperature at my place is often 10-15 degrees off from the temperature at the airport which is on the coast. Are those temperature readings legitimate or not? The phone can only guess. It seems like way too much effort for something that would never work well enough to be terribly useful.

Re:Very nice

[nschubach]

Wouldn't pressure be useless when someone gets into their car and turns the A/C on high or is that not enough pressure difference to affect barometric readings?

Re:Very nice

[amRadioHed]

All sensors are subject to error, of course. Speaking of which, you neglected to mention instrument error from design and manufacturing as possible causes of it

These errors tend to be minor and can be largely corrected for with calibration. There is no calibration that can account for the differences from walking into a warm room, or running a game or navigation app that makes the phone run hot. And honestly none of the things you mention seem like they would be useful for the phone to have.

But wouldn't it be interesting to be able to do things like having your phone automatically adjust your electronic thermostat while you're on your way home, based on geofencing and the temperature it detects locally

Adjust my thermostat at home based on the temperature in my pocket? Or in my car? Either way how would that be useful? The local weather data the phone already has is far more likely to give a usable number. Or why not just skip remote temperature readings of all kinds and just use a temperature reading from the house itself?

Or wouldn't t be interesting to have your phone act as a night-time fire alarm by sounding at full volume if the temperature exceeds a certain threshold

By the time there is a substantial temperature increase at my nightstand I've probably already succumbed to smoke inhalation. My house already has smoke detectors which will go off way before my phone detects anything.

And pretty much everything else you mention are one off things that are far better done with a specialized tool that will give decent readings. IF the phone could get reliable temperature readings then I'm sure someone would find a use for apps that do any of these things. But it can't, so why bother.

Re:Very nice

[icebike]

I doubt the car fans would make a huge difference. MAX AC usually works with recirculated air anyway, rather than trying to push more air into the closed car.

But you could test it for yourself by taking you house barometer for a test ride.

The slamming of the car door might induce a transient pressure spike if the car windows were closed.
Accelerating, or Braking hard will create a pressure differential between the front and back seats of the car. You can sense that with nothing more complex than a tethered helium balloon. As you accelerate, the balloon floats forward, toward thinner air.

(I can foresee a CSI tv episode based on this where the hit and run driver claims he was a passenger in the back seat but was running this app on his phone which registered a pressure spike just before time of impact).

Re:Very nice

[ThePeices]

How do you measure different degrees of tempura?

Is it a 10-1 like-dislike scale?

Re:Very nice

[ThePeices]

I havent seen any Android phone that has an ambient temperature sensor. All current ones just have a battery temp sensor.

And an ambient sensor would mostly show the temperature of ones pocket.

Re:Very nice

[The Bean]

WOAH. Are you serious about the tethered helium balloon?

I HAVE to try this. I shall be angry if I discover I've been trolled :)

Re:Very nice

[icebike]

Would I troll you dude?

See demo on YouTuve: http://www.youtube.com/watch?v=XXpURFYgR2E

Re:Very nice

[blind biker]

I think itâ(TM)s just pressure â" and I am not sure how valuable a tempura gauge would be.

I like Japanese food a lot, tempura especially. I think a tempura gauge would be very useful.

Re:Very nice

[nschubach]

My older brother did his Science Fair project on this back in the 90s. It's very cool to see it happen and it explains why having a lot of helium balloons in your car can be a bad idea.

Re:Very nice

[amRadioHed]

Turning the fan on full in my car, with windows shut, made a difference of maybe .5 millibars. Almost negligible considering I'm seeing about a .2 millibar range of random fluctuation from the meter.

Surprise sensors.

[Anonymous Coward]

The most exciting part of this - my Galaxy Nexus has barometric sensors!?

Re:Surprise sensors.

[Anonymous Coward]

Pop on the Play store and look for an app called "Android Sensor Box". It shows which sensors you have and a "fun" version of the output, in the event you want to toy with it rather than see the raw data.

My phone (Kyocera Rise) has: Accelerometer, light sensor, orientation sensor, proximity sensor, sound level sensor, and magnetic sensor.
It doesn't have temperature sensor (though I know it's got a CPU temp sensor, go figure, just not ambient), gyroscope sensor, or barometric pressure sensor.

Have fun.

Re:Surprise sensors.

[RealUlli]

Install AndroSensor and check out what you phone supports.

https://play.google.com/store/apps/details?id=com.fivasim.androsensor&feature=search_result#?t=W251bGwsMSwxLDEsImNvbS5maXZhc2ltLmFuZHJvc2Vuc29yIl0.

(Hint: I own a Galaxy Nexus too, it does have a barometric sensor. Btw, the proximity sensor has only two values: 0 inch and 2 inch. It's used to lock the and unlock the screen when you put your phone on your ear during a call)

Why do the phones have barometric sensors?

[SuperKendall]

I thought it was pretty interesting that phones would include barometric sensors which I had not heard of before - are they just there in a package with other more commonly used sensors? How do the phones that have them normally make use of or present that data?

Re:Why do the phones have barometric sensors?

[Anonymous Coward]

I thought it was pretty interesting that phones would include barometric sensors which I had not heard of before - are they just there in a package with other more commonly used sensors? How do the phones that have them normally make use of or present that data?

Makes GPS system start up faster.

Re:Why do the phones have barometric sensors?

[Atticka]

They were originally added to assist with navigation (they double as an altimeter, sensing pressure changes due to elevation) allowing the phone to acquire GPS lock quicker by using the data in conjunction with latitude and longitude calculations.

Re:Why do the phones have barometric sensors?

[plaukas pyragely]

Altitude measurments / checks?

Re:Why do the phones have barometric sensors?

[babtras]

Would the manufacturers really incur the extra cost and extra power consumption of another component for this reason alone? I'm sure there must be more benefit than a couple seconds gain on GPS acquisition.

Re:Why do the phones have barometric sensors?

[interkin3tic]

More specifically, it confirms where you are faster. I've heard it's not quite sensitive enough to give you weather predictions based on your one phone (hence the need for distributed phones) but it IS sensitive enough to give an idea of altitude. So if your barometer shows you're at sea level, and the GPS starting up is saying you're in Denver, your phone can say "You sure about that, GPS?"

Re:Why do the phones have barometric sensors?

[bryantthesmith]

They also assist in Emergency (911 and such) location. I worked at a company that develops the MEMs ICs for the pressure sensors and one of the goals was to determine accurate enough altitude information to determine the callers location in a tall building. If you call 911 in a high rise it would be nice for the emergency personnel to know which floor you are are on.

Re:Why do the phones have barometric sensors?

[nschubach]

What if you are in Denver? ;)

Re:Why do the phones have barometric sensors?

[nschubach]

But wouldn't the fact that you are in an enclosed and possible pressure variable environment (A/C on high?) cause those readings to be inaccurate?

Re:Why do the phones have barometric sensors?

[shutdown -p now]

Then it adjusts the date to somewhere circa 2200, to account for the apparent ocean level difference.

Re:Why do the phones have barometric sensors?

[bryantthesmith]

I wondered that too. I asked those involved with the design and they said that according to their tests, nearly all buildings are quite leaky and that the ventilation systems don't effect pressure enough to throw off the readings.

Re:Why do the phones have barometric sensors?

[interkin3tic]

Presumably, in such a situation, your phone being wrong is less of a concern than you being a mile underground.

Re:Why do the phones have barometric sensors?

[Belial6]

Has 911 ever used phone tracking to send emergency personnel? Anyone got a link to show that this has ever been done? Every time I have called 911 from a cell phone the operator always insists that I give them an accurate location by voice. This is even when I was on a long dark road without any signs.

Re:Why do the phones have barometric sensors?

[godel_56]

Would the manufacturers really incur the extra cost and extra power consumption of another component for this reason alone? I'm sure there must be more benefit than a couple seconds gain on GPS acquisition.

I've heard barometric sensors may be used in currently-experimental navigation apps inside buildings, such as large shopping malls.

The air pressure tells you which floor you're on and other methods, such as triangulation of signal strength from wi-fi sources, give your X,Y location.

Re:Why do the phones have barometric sensors?

[Agripa]

It will not matter in all but a very few specially designed buildings but the difference sure shows up in a pressurized aircraft. Either of my sensor equipped GPS units will show a GPS elevation of say 28,000 feet and a barometric elevation of 5 to 6 thousand feet when cruising.

Who knew?

[Anonymous Coward]

Who knew the Android was aneroid?

Needs isobar lines

[hey]

Then you can see what's happening

Re:Needs isobar lines

[cryptoz]

I agree completely! We are adding those as soon as we can. In the early days of the project, there was not enough data to build isobars. But now, we have enough and we are determined to add that feature. The project is built by me and volunteers in our free time, so it'll probably be a couple weeks or months before we get isobars in. Of course, pressureNET is fully open source and so if anyone feels like writing the isobar code and getting that feature done faster, we will welcome that too! Code is on github: https://github.com/JacobSheehy/pressureNET, https://github.com/JacobSheehy/pressureNETServer and https://github.com/JacobSheehy/pressureNETAnalysis.

Re:Needs isobar lines

[Anonymous Coward]

Rather than open source, do you mean "free software" as in freedom?

The next weatherbug?

[Anonymous Coward]

I hope not...Always felt like they hijacked what would have otherwise been a nice effort. Lets hope the marketing folks are still more interested in branding er..I mean naming storms ;)

Just like the "Dark Knight"

[Thud457]

Now we have a distributed sensor net to pinpoint where things go "boom". Also this. (WTF, ACM, you want people to pay for that?!!!)

Does it really improve local forecasts?

[sunking2]

What are you really going to get out of it compared to set calibrated weather stations that are probably within a dozen miles of where you are. Other than the fact that someone has their phone on the 25th floor throwing things off. Or driving in a car with the heater/ac on throwing things off, etc. Fun thing to play with, but to think it's going to improve on what we have available is a bit much for most people.

Re:Does it really improve local forecasts?

[cryptoz]

You might be right, but I don't think so. If you look at the screenshot of the analysis tool in the blog post, you can see that even through all of the noise there is a *very* clear curve for Hurricane Sandy. Although, we do know that the data is very noisy and we are working with some professors who are researching this very topic (calibration of phone sensors for weather data collection). I think we will have useful data for short-term prediction. But if not, that's okay too. Science!

Re:Does it really improve local forecasts?

[SuricouRaven]

Data mineing. The value of data from an individual phone is low due to those noise factors, but if you've got a sufficiently huge number of them then you can still extract useful information via statistical processing.

Re:Does it really improve local forecasts?

[Anonymous Coward]

Greater resolution with accuracy that improves as number of participating devices increases.

Re:Does it really improve local forecasts?

[MightyYar]

Yes, how do they handle corrections for people inside of buildings kept at a positive pressure differential from the outside? People leaving their sunroof open on their car? Seems like very noisy data, but I guess if you have enough it will be statistically possible to pick out the right pole and use that average. I guess you could use the more established sensors to pick the correct pole.

Re:Does it really improve local forecasts?

[na1led]

The pressure in my home is different than outside, especially during winter months. I have a digital weather station on my wall, and the sensor has to be placed outside, so fail to see how useful this is going to be on phones.

Re:Does it really improve local forecasts?

[Immerman]

A heater/AC won't throw off pressure readings, and the guy on the 25th floor will have his altitude properly reported by GPS once a lock is achieved (the better you can estimate such things a priori the easier it is to get a lock, but once you have a lock you can calculate them more accurately).

As for the usefulness, which do you suppose is more informative, especially in an urban environment where micro-climate effects can be substantial? A calibrated weather station twelve miles away, or that same station augmented by thousands of auxiliary barometers scattered throughout your area? At first it's going to be little more than a novelty to the average user, but it's building up an immense data set that could help drastically improve our understanding of how ground-level structures affect localized weather patterns.

rubbish source of data

[SuperBanana]

Cell phones are often:

-In cars, which have varying interior pressure levels depending on design, speed, and other conditions (for example, I had a car where putting the sunroof in the "vent" position would result in a noticeable change in air pressure)

-In buildings, which can have wildly different pressures floor-to-floor or even between areas depending on a variety of factors

-In hyper-localized high pressure areas (for example, ever been caught in a severe wind gust between skycrapers? How about subway entrances and exits?)

-At different heights. Barometer readings are useless without knowing your altitude, and GPS is extremely poor at moment-to-moment altitude data; you have to collect a fair number of points over at least a couple of minutes. Do they perform this calibration?

A+ for the idea, C on evaluating the likely accuracy of the data...

Re:rubbish source of data

[Exoman]

I imagine they've considered outliers in aggregation of the data. You particular phone may be all that, accuracy-wise, but with sufficient data points, one could readily get to a point near the precision & accuracy limits of the phones' sensors.

Re:rubbish source of data

[roninmagus]

I'll take your word on the accuracy of barometer readings as I don't know much about the science there. But I can add that I believe this to be a fact with pretty much any crowd-sourced data. Your algorithms will have to allow for these sorts of things, and you are sacrificing an exact single reading for an approximation based upon multiple slightly inaccurate readings.

Re:rubbish source of data

[Anonymous Coward]

If you have a large enough sample, based on their locations, then you can remove the anomalies and get some very accurate readings. It only involves even more private data changing hands, but most people don't care about their privacy, so I don't know why I care.

Re:rubbish source of data

[jelle]

"C on evaluating the likely accuracy of the data..."

Don't forget that currently available barometric data from other sources is quite sparse in most locations.

You have to start somewhere. Apps can be updated with improvements...

A lot of the data issues you mention occur more in urban environments than they do in suburban environments or out in the country, and the currently available barometric data from other sources is probably a lot more sparse outside of urban environments than it is in areas with subway systems and tall buildings (people do have smartphones with enough data coverage for sensor networks well outside of large cities).

The app can certainly also look at other sensors such as the acceleration, positioning, and motion sensors to help with data issues. The 'network' can find bias or accuracy issues from nearby measurements of other devices. The 'In a car with open windows' situation is easily detectable by the high rate of variation in pressure even for a short measurement duration (and even so, a long term average might still yield useful information, bias 'when in motion in the car' can be measured when the vehicle stops (using the motion sensors)). The app could even learn, f.e. "When near this wifi, report the measurement with a potentially previously learned pressure bias", or "When near this bluetooth, then it's in a car". With the light sensor and positioning/acceleration sensors, the app can even be quite aware of various state of the phone (sitting motionless on a flat surface, or being carried inside of something, or being carried open, or being held up to an ear, etc). I'm pretty sure it is possible to make an app that collects a tremendous amount of useful data for weather prediction that would otherwise be unavailable (actually, temporal variations from motionless devices in a location with unknown bias can be very useful raw data, and the strengths and positions of the windgusts around buildings may even be useful raw data).

Also don't forget the power of large numbers, you don't have to use all the data: Outliers or inaccurate measurements/devices can be detected as such when surrounded by enough other measurements, and snr improves with filtering of multiple measurements (also from multiple devices).

Re:rubbish source of data

[AmiMoJo]

In cars

Detect movement > 10kph, reject reading.

In buildings

Barometers work fine in most buildings, outliers are rejected.

In hyper-localized high pressure areas

Results are averaged, outliers rejected, several phones data required for confirmation etc.

At different heights

Few people live/work high enough for it to make much difference, and ground level is easy to adjust for based on approximate location (1km).

This problem was solved decades ago.

Re:rubbish source of data

[Dishevel]

As far as I know GPSs ability to get your altitude is just as accurate as getting your lat/lon.

Re:rubbish source of data

[Fnord666]

-At different heights. Barometer readings are useless without knowing your altitude, and GPS is extremely poor at moment-to-moment altitude data; you have to collect a fair number of points over at least a couple of minutes. Do they perform this calibration?

Could you do a Lat/Long lookup to obtain a "good enough" altitude for the measurement? Hopefully GPS is better at these readings.

Re:rubbish source of data

[gtirloni]

In my area I can safely assume >50% of people are in air conditioned environments. How would they detect it and ignore our data? Very tricky.

Re:rubbish source of data

[MozeeToby]

Yes, but sometimes the outliers are the data you want. I work in a building complex with several thousand people in it that covers nearly a square mile, compared to probably a handful on the grounds outside at any given moment. A statistical analysis would throw the outdoor data out and keep the indoor.

Re:rubbish source of data

[SuperBanana]

"Barometers work fine in most buildings, outliers are rejected."

Barometers work in buildings where the barometer is in a set position, the barometer has been calibrated by the owner (who fetched the most recent barometric pressure reading via weather radio and set theirs to match), and the building's ventilation system doesn't change.

"Few people live/work high enough for it to make much difference, and ground level is easy to adjust for based on approximate location (1km)."

Newsflash: many people live in areas where the elevation difference within 1km can be hundreds of feet. Just because YOU live in a flat area doesn't mean everyone else does, child.

Re:rubbish source of data

[tlhIngan]

Few people live/work high enough for it to make much difference, and ground level is easy to adjust for based on approximate location (1km).

Could you do a Lat/Long lookup to obtain a "good enough" altitude for the measurement? Hopefully GPS is better at these readings.

Let's see, all around me I see what qualifies as high-rises (10-20 storeys, though a few are higher). If you're not properly taking altitude into account (and these pressure sensors are usually good to +/- 20 feet vertically) then there are pressure differences. Heck, my aircraft altimeter even shows a difference, and it's precision is only +/- 50 feet on a good day (it's older, so its mechanics do get "stuck", but still compares against my Gnex.

The whole reason to do this project is to get an idea of the variations in air pressure over a region - which will be fairly minor (altimeter settings don't change that much - usually less than 0.1"Hg over 40 miles or so). If you're wanting small localized pressure variation data, you'll want finer data and altitude must be taken into account (0.1"Hg is roughly 100 feet).

Plus, buildings can be pressurized.

Oh, and in his area, there are LOT of people, so there will likely be a higher percentage of people who would participate in this project. Plus, places like New York City I'm sure only have "few" people working in high rises (and are ripe for very widely varying micro-climates and weather cells).

Re:rubbish source of data

[PPH]

- All the inmates who have to keister their contraband cell phones during a shake-down.

Re:rubbish source of data

[quarrelinastraw]

Your points are all valid, but the problem isn't as bad as it seems. If the noise in the errors are uncorrelated (e.g. all cars throw off readings, but in a random way), then the noise disappears in the average (by the Mean Value Theorem). Otherwise, you just model the correlation and account for it (e.g. with Bayesian methods). More generally, there are numerous algorithms specifically designed to extract useful aggregate data from a large number of networked noisy/low quality sensors.

Re:rubbish source of data

[bill_mcgonigle]

This problem was solved decades ago.

If only the people who put thousands of hours into developing this system had spent the five minutes a random Slashdotter did criticizing the system they could have given up already!

Re:rubbish source of data

[swillden]

With sufficiently-good location data, that could easily be addressed by using available map data (Google is extending the coverage of its building footprint data, for example) to identify people who are almost certainly indoors vs outdoors, and using those as references to help determine which to discard and which to use. Of course, location data indoors is often very poor, because GPS signals are badly attenuated -- but that is also a clue. Android also uses Wifi signal strength as well as GPS and cell network triangulation to refine position, which can make indoor location quite accurate, but as long as the barometric app can distinguish tell the source of the error margin reported on the location, that can be handled.

As for phones in cars, velocity is a good clue.

Finally, it could also be correlated with barometric data from weather stations.

There's no doubt that extracting really meaningful data from the torrent this app could generate is non-trivial, indeed finding ways to do it is almost certainly fodder for a good number of PhD theses, but I'm fairly confident it's within the power of advanced statistical analysis. It may take a few years to make it really good.

Re:rubbish source of data

[swillden]

No, vertical location is about an order of magnitude less precise than horizontal location.

Re:rubbish source of data

[Anonymous Coward]

The only thing rubbish is your understanding of and assumptions about data collection and inteperatation.

Infrequent readings of individual devices are going to be crap, yeah. But that would be stupid.

What is a smartphone? A computer. A computer, always on, always charged, always connected to a network, and with GPS it always knows where it is /and/ what time it is..
You can poll the sensor as often as you want. Properly programmed, you could probably take a reading every minute and have a negligible impact on battery life.(Probably batch upload all collected data once a day)

You could take all the collected barometric readings an as an aggregate and easily get more than enough information to negate any noise caused by the fact you're collecting data from someone's smartphone. As more users join the project, more information enters the pool and your useful accuracy goes up even further.

Re:rubbish source of data

[spire3661]

Hell, open just the back windows in my car at 40 MPH+ and a pulsing shockwave forms in the interior. I have to open the fronts slightly or the pressure waves become unbearable.

Re:rubbish source of data

[spire3661]

Bicycles go faster then 10 kph. You parameters just eliminated a sizable chunk of useful data.

Re:rubbish source of data

[Anonymous Coward]

A statistical analysis would throw the outdoor data out and keep the indoor.

Yeah, I guess a brain-dead statistical analysis would. Real statistical analysis isn't a process of saying, "oh, those are outliers, just discard them." It's a process of studying data, determining what it says, and then using statistical tools to improve the S/N ratio.

Re:rubbish source of data

[ThePeices]

Yes cell phones often are.

Which is one reason why we use filtering algorithms from a large number of sensors to remove these artefacts from the data.

Re:rubbish source of data

[knarf]

Ah, another naysayer who knows best, typing away from the comfy confines of his home/basement/cubicle.

Tell me, oh know-it-all, have you considered the possibility of using software to *gasp* filter out those anomalies? After all, given a dense-enough sensor network it becomes easy to sort out the outliers since air pressure does not vary that much in a given area.

As to the different heights, this is also rather easy. The location of the sensor is known within a few hundred meters. The height can be correlated from map data. Yes, in some areas people tend to climb tall buildings, but those areas are well-known and rather concentrated. Also, there is that software filter you can use to normalize the data.

To conclude this rant, maybe you should think a bit before starting to spout criticism. You might find out that, in fact, you did *not* know better than all those other people who have spent time and energy to rig up an interesting project.

Old news

[technomom]

Old news. I have had this on my S3 since Day 1. I also have Barometer Monitor, which generated this pretty cool graph on Monday and Tuesday, October 29-30, the days Hurricane Sandy came to town and then left. http://i.imgur.com/tuM8x.png

Re:Old news

[dotancohen]

I think that the real news is that your phone has more vertical pixels than my desktop. Nice screenshot!

This is great news!

[ducomputergeek]

I can finally build that Bull Halsey Don't Sail Your Fleet into A Typhoon app I've always wanted to create!

Oveerpressurised

[Teun]

I usually work in over pressurised labs, how's that for nice local weather?

Won't Work?

[Lord Byron II]

According to this article (http://us.gizmodo.com/5851288/why-the-barometer-is-androids-new-trump-card), a guy from The Weather Underground says it won't work. He says that the pressure gradients are too flat and the sensors are too imprecise to be able to accurately measure local pressure any better than the existing network.

If everyone in a crowded room ran the app

[Ukab the Great]

You could theoretically pinpoint SBD malefactors in real time.

Variations from natural vs. unnatural causes

[SuperKendall]

I agree with all your points.

But are the relative changes from weather significantly higher than from building caused conditions? If so the building factor would only be a constant filter atop changing data, and also no matter what relative drops over time would be detected.

Also you could say you would only sample data when you were moving at walking speed and not joined to a WiFi network. That way you'd mostly be sampling outside, not work or home or a car.

Excellent

[Trogre]

This should lead to another patch for the all-important Tricorder app.