While reading about the ‘internet of things’ I’d come across a site called Pachube (how do you pronounce that?). It’s an open source type community site, that allows you to publish data feeds from sensors and the like. It then makes pretty graphs and charts and stuff for you to use and share.
This seemed great, but as it was pretty much a turn-key solution, I felt it didn’t quite offer me the tinkerability that I so crave. It was then that I discovered a bit of software called pChart, it’s PHP based and makes pretty nifty graphs, charts and progress indicators out of your data. It’s free, so I thought I’d give it a go.
It took a couple of hours, and I was pretty confused, but then the penny dropped. This thing was great. I had it showing me the temperature in my living room, over time! The ultimate plan is to build a dashboard for the whole house, with temperature data and a whole host of other things reported in one central location.
I felt I was getting ahead of myself, charting a load of data from a hastily assembled prototype. But when I started looking at the graphs of the temperature, I noticed anomalies I’d probably have overlooked in a table of data. Every few hours, the WiFly would report an errant result. It seemed the temperature in my house was around 18-19C during the course of the day (which seems about normal) but once or twice a day, it would have a single odd measurement at say, 11C. Which is glaringly apparent on a graph of that data. And then the subsequent readings would follow on as though nothing had happened, 18.482… etc.
So now I’m thinking of ways of dealing with these anomalies. I thought that perhaps it was some sort of radio interference due to the excess of wire connecting everything together. But as yet, no solution.