Show HN: Open source Android app building a network of barometers | Hacker News Reader

Show HN: Open source Android app building a network of barometers

market.android.com

27 pointscryptoz16 comments

Cool idea. A couple questions I have about it:

Are graphs available on the website? (I tried clicking through to cumulonimbus.ca but it was not loading.)

Does it account for elevation? A change of elevation of 100 m is about 12 mbar of pressure, comparable to a modest weather event. You could use ground topography to correct, but people working in midtown Manhattan skyscrapers would be a noisy signal, for example.

Currently the data isn't available outside the app, but I'm working on that and should have that out within a couple of weeks. My focus so far has been growing the user base, so the web interface has lagged. Coming soon!

For elevation, we currently attempt no fixes. The attitude I've held so far has been that this is the primary reason for building a _network_, since a single barometer cannot be trusted. However, your Manhattan scenario is interesting: what if most people are constantly changing elevation? Hm. I'll work on that one...

This is a wonderful idea, especially since the area covered by cell phones vastly surpasses that of weather stations. Unfortunately, there's a very real privacy concern of which you must remain cognizant: storing a time series of the paired barometer reading and GPS coordinates. This information is personally identifiable, even if the device ID is just a unique string of numbers.

I'm not sure how you're storing the data, but I would suggest to store only the GPS coordinates and barometer reading, normalizing both to have the same approximation error.

> Unfortunately, there's a very real privacy concern of which you must remain cognizant: storing a time series of the paired barometer reading and GPS coordinates. This information is personally identifiable, even if the device ID is just a unique string of numbers.

Yes, this has been a complication from the start. Here's our current policy/methods: We delete all data after 1 month and we let users remove all of their submitted data completely. We never show anyone's actual location on the map (coordinates are fudged by a few hundred meters).

I do want to store some data longer than 1 month in the future as will be helpful to use for data mining, looking for patterns, etc. But first I will have to find a way to carefully anonymize it.

I'd surprisingly prefer storing all data indefinitely and not allowing users to delete what they submitted if and only if you also eliminate any way to group locations by a device.

This data is just too dangerous, even in spite of innocuous intentions. Data breaches happen, National Security Letters remain threats, and targeting ads based on a user's current location will tempt anyone. You will fill your users with confidence if you remove your ability to group the data points.

Moreover, you shouldn't discount the importance of historical barometric data. At some point, it might no longer be practical to store the data in its most atomic state, but maybe rolled up to a periodic snapshot (maybe hourly) of some geographic region.

What data would you be looking to store for longer than a month?

I'm the developer for this project. Currently only the Motorola Xoom and the Galaxy Nexus can submit readings, but all devices 3.1+ can download the app. Let me know what you think, any comments, suggestions, etc.

In the long run, we're working towards improving short-term, local weather prediction by adding this new pool of unprecedented data to the mix.

Edit: Source available at http://cumulonimbus.ca

More data is nice, but the experience with citizen-operated weather stations suggests that the data is useless without extensive quality control. Calibration is hard, and mobile sensors are particularly challenging. What happens when one of the sensors goes into a climate-controlled building, then takes an elevator to the 30th floor?

So, part of this is an experiment. I don't think anybody's done such fast and distributed crowdsourcing of weather data before, so I'm not sure how useful the data will end up being. As far as a single user entering a building, taking elevators, etc, I'm not concerned since those data points will stand out from the nearby readings and can easily be discarded. However, a point raised in another comment here is, what if most users in an area (say Manhattan) are constantly changing elevation. That very well may pose a problem. We'll see. :)

Thats great, surprisingly there are already a few submissions in my area (São Paulo / BR) even if the Galaxy Nexus was not launched here yet and the only device with a barometer is the Xoom. I think it is a revolutionary idea that can change the way we see weather forecasts and I am really excited about it.

This is a very cool idea. Sadly I don't have a Galaxy Nexus in order to participate (just bought a Galaxy S2, wondering if I will regret that choice) but I can't wait to see how this kind of thing turns out in a year or two when a ton of people have these devices.

Slightly offtopic - have you noticed anything lacking in GS2, compared to what newer handsets have? I was going to do the same and can't find anything apart from raw processing power that I'd miss.

Not really. I actually came from having tried a Windows Phone (Samsung Focus S) for nearly a month which was incredibly light and thin, so the GS2 actually feels quite bulky to me- the 4.5" screen is honestly a little too big for my liking, but given that the Galaxy Nexus is bigger still I don't see any very viable alternatives.

Another thing to note is that the AT&T SG2 (the one I have, the 'Skyrocket' LTE version) is very new, and really don't have any fully functional custom ROMs yet. I'm using a half-baked CyanogenMod version and there are still plenty of issues.

Slightly off topic: But why does those phones have barometers build in? What are those normally used for?

The altitude input helps speed up the GPS lock process.

http://androidcommunity.com/samsung-galaxy-nexus-barometer-i...

Is there any page that displays aggregate data without having to install the app?

Not yet, but I'm working on it! So far every line of code I've written has been in effort to directly increase barometer reading submissions. That's coming along nicely, so one of our next steps is to redo the website and include a map and graphs of the data there as well.

Reply on news.ycombinator.com