My SURF took quite a bit of work at the beginning, getting everything working. We built our whole system from scratch, so nothing actually worked until several weeks in. Now we have a working system! Currently, we only have one working client, a laptop in my SURF mentor’s office. We have ten of the accelerometers, though, so we’ll soon have more clients. The day after I installed everything for my mentor and got it running he came back with a list of improvements he wanted, so that will keep me busy! Our server had a bit of trouble towards the end of last week. Because of the construction in Annenberg, the power was going to be switched off, so we had to shut it down overnight. Unfortunately, the only people with root access to the server were in Norway for a conference! So, we had to turn the machine back on first thing in the morning, then wait for them to finish eating dinner (since they’re nine hours ahead), sign in, and set everything back to rights. Then the next day, do the same thing again! What a hassle! No matter. The SURF students working on the server side of the project whipped up a snappy little application that puts the data received by the server onto a Google map. Now we can see it all the way from thumping on the table to showing up as a dot on the map. Very gratifying.
I have data from the experiment with the sledgehammer, too. I overlaid the data received by the professional sensor and the data received by our sensor. Our data is quite a bit worse, which is to be expected. Since our device is so small it has very little inertia and ended up bouncing around after each event, which added lots of noise to the data. Apparently, I was too close to the sensors when I was swinging the sledgehammer, so the seismic activity was not quite right, either. Plus, we’ve switched to a different brand of sensors since then for programming reasons. Oh well, I’ll just have to do it again! You might notice that the times don’t match up quite right. That’s because I entered the time difference between samples on my device as 0.01 seconds, when it was actually 0.015 seconds. Oh well. I’ll fix it next time.