Learning about power-usage data from my home

I recently moved to a new home, and I got a freshly installed new electricity meter that’s a little bit more intelligent. It was now possible for me to sign up for slimmemeterportal.nl, that let’s you access the power-usage data from the meter by a 15 minute window. That’s quite some data to play with!

Being interested in the data, I wrote a little scraper that collects it for me and stores it in a SQL database, from there on, with Flask and Peewee it was quite simple to create a little API for my own dataset. Currently I’m playing with react.js and angular.js to learn them both a bit and explore what it possible. Here’s one of the first graphs:
Power Usage of my own home over the past 9 daysI’ve put the scraper-code on Github for everyone to use! I suggest if you want to liberate your data, that’s the way to go!

What I want to do? We’ll there are several things:

  • Crossfilter the data to pick up any interesting data points.
  • Explore machine learning if I can predict my power-usage more
  • Find correlations based on temperature, activities or weather
  • What are those mystery peaks during the night?