After completing the journey to PMSS, I have now experienced first hand what it is like to complete a project at the destination, on the deadline. While it was not the first time that I have helped to “macgyver” something at the last minute, and will definitely not be the last time, it was actually quite a rewarding experience. Our team worked together swiftly and successfully to solder together working sensors. We did this by following the schematic for the DHT11 and DHT22 and using a sort of solderable breadboard.
We soldered the sensors in place, as well as 2 eight-pin headers that would connect to the pins on the raspberry pi. Based on the placement of those pins, we were able to solder wires correctly connecting the legs of each of the sensors to their respective places (power, ground, an input pin).
While we were working on soldering together sensors, one of our colleagues and Dr. Jadud worked on getting our code built into the state machine. We had to modify our sensors to read in 5 “good” readings and average them. The term “good” refers to a value that is not None, because sometimes the sensors do not read. They were also able to get the code working to where it began upon boot up. Eventually we realized that we had left all of our raspberry pi’s for the deployment sitting on a partially out-of-sight table back in the lab. At this point, we almost panicked. Lo and Behold some of us had brought our personal pi’s on the trip. We were able to scrounge together and locate 5 raspberry pi’s that we could hook up the sensors to and run the code on.
We gave each pi a separate sensor ID (animal names of course) and tried to pull the code down onto them from the web. This gave us some issues because a few of them did not want to connect to the internet, and one of them completely quit working, which left us with 4. Once we had installed the code and connected the sensors, we ate dinner and commenced to deploy our prototypes.
We placed our sensors in 4 different locations. One was connected to a loose extension cord in the archives room. This pi is no longer speaking to the dashboard. Another was connected to a plugin controlled by a switch. This pi has strange gaps in reporting time (as in it is down for a couple of ours at a time). The other two seem to be going strong at the moment. Overall it was a successful adventure that I am happy to have been a part of.