Overview We've built a platform to both help consumers save on their energy bills and help energy providers predict load but leveraging social media. Through another project, we have access to a sample size of homes in Massachusetts. These homes are reporting their energy use to our servers, in real time using one minute intervals. Around this data we have built a challenge platform where towns can complete with each other for a prize, in our case winning a pot of money to have potholes fixed in their town. User can view in real time, on a map, by town whether a town is currently over or under their baseline. There is a slider so they can scrub back in time. Underneath towns can compare their usage in more detail. Tweets referencing the competition appear on the left as well as the current weather (which impacts usage).
For energy providers, which historically use modeling and weather patterns to predict load, we propose using real time social streams to predict future energy patterns. As social software reaches a critical mass (see Facebook's mood analyzer) this will be more and more possible. What can upcoming events, Flickr photos and localized tweets tell us about a space's future energy consumption.
Technical: YUI for interface, YQL for a number of the data sources, BOSS search for popularity rankings, Yahoo! weather API, upcoming API, Google maps, Twitter widget, Flex, Rails, jQuery