After a year and a few months, my Raspberry Shake is continuing to operate well in my garage, sending a constant stream of seismic data to the online server. I have to say I have been impressed by the stability of this device. It has captured quite a few earthquakes (I'm in California) and has had no unexpected downtime that I know of. I work with a lot of homebrew projects and other commercial embedded devices, and I cannot say the same of any of them! The OSOP team in Panama that continues to develop the hardware and the global network has remotely updated the R-Shake several times during the past year, all hands-free on my end.
Just today, the same group launched a new kickstarter with a similar idea: this time called "Raspberry BOOM" which is an infrasound monitor. So it's like the previous "Shake" project, but using a microbarometer (air pressure sensor) instead of a geophone (ground vibration sensor).
One of my own motivations is that in my neighborhood we have an online forum "Nextdoor" and with some regularity I will see posts like "did anyone hear that sound just now, maybe a firecracker? ...or a gunshot?" And then a few more people will post who either did or didn't hear something, and someone else will say they hear that kind of sound all the time, and it's hard to say what the bump in the night was, if indeed it happened at all.
Now with something like this instrument, if you do pick up a signal you could say at least when it was, and quantitatively how loud compared with other such things previously observed, and from past records how often that type of sound happens. And if several neighbors had these and picked it up, (can you imagine, several neighbors, they might think it was a movement... oops, sorry) you might be able to tell something about where it came from, since NTP can keep the RPi clock aligned to UTC often within 10 msec or so, which is the time it takes sound to travel about 3.3 meters. The R-Shake is networked and I gather the R-Boom will be, so it may be possible for the remote server to determine the location of a sound, if enough sensors capture it. They are already doing this for earthquakes with the R-Shake network, however I gather in the case of the "Boom" device this capability is speculative and as yet untested.
I don't have this device yet but recently I played around inside with a regular microphone modified for lower frequency response and was interested to see I could tell when doors several rooms away open and close from the air movement, and when people walk over the wood floor which apparently flexes from the weight like a drumhead, causing low-frequency pressure fluctuations. My mic hack is only good to about 1 Hz but the "Boom" claims sensitivity down to 0.05 Hz which is a 20-second period.
Lower frequency sounds travel longer distances. One extreme example is the 2013 Chelyabinsk meteor
which emitted pressure waves detected around the world by the existing infrasound network intended to monitor nuclear tests. Meteors that large are rare, but if (when) another one like that hits, there's a good chance this device could detect it. Likewise with large volcanic eruptions.
FWIW I am not affiliated with the group doing the kickstarter, but I think these widgets are very cool, so I wanted to post here.