Tweet-A-Watt is an elegant solution to the problem of tracking electricity.  While the smart grid gathers steam, it's still a long-term project, and many electricity consumers won't see results for several years.  Tweet-A-Watt addresses that problem in the short term.  By combining the Kill-a-Watt power meter, the XBee wireless board, and the Twitter API, Tweet-A-Watt gives consumers a device into which they can plug their appliances, and tweets their energy usage to Twitter.

The basic Tweet-A-Watt system could be improved by making the data available in a semantically meaningful way, i.e. via some sort of structured database for electricity usage which is API-accessible.  With that kind of data infrastructure in place, it would be possible to develop interesting tools that allow consumers to see historical trends in their electric usage; break down usage by outlet and/or appliances; and get alerts whenever their usage exceeds a certain limit.

Fortunately, the Tweet-A-Watt system is very accessible.  Start-up Tweet-A-Watt hardware packs cost $90, and Wattcher, the Tweet-A-Watt software, is open source.  An interested hobbyist could be up and running with Tweet-A-Watt in a weekend, without much trouble.

For wider distribution, it will probably be necessary to make Tweet-A-Watt a bit easier to build, and a bit cheaper - perhaps through a partnership with Kill-a-Watt, or by using a cheaper meter.  There's also a need for software that's a bit easier to use, perhaps using a software-as-a-service (SAAS) business model.  Such a service would allow users to register their Tweet-A-Watt systems, describe the kinds of appliances plugged in to each Tweet-A-Watt, view charts and graphs of their usage, and then receive alerts whenever usage jumps beyond some absolute or relative limit.  The software could even integrate with users' Facebook accounts, to let their friends know when they saved electricity thanks to Tweet-A-Watt.

While I think a SAAS model would be the easiest way to create a sustainable, accessible data store of electricity usage, the software could also be distributed as an open source web application, which motivated users could install on their own private web servers.  Neither the code nor the database structure for this kind of software is likely to be very complex, and probably won't require very sophisticated maintenance.

I'm always eager to see low-cost, accessible solutions like Tweet-A-Watt emerge, and I'm even more excited about the prospect of derivative services that make those solutions easy-to-use.  Making energy efficiency tools readily available is an important step in reducing our overall energy usage.