Wednesday, June 6, 2012

PlotWatt? Plot What? Plot Everything!

To improve, you must first measure. And measuring is getting easier with cloud computing.

The problem with a lot of software solutions is that the medicine is often more toxic than the disease - implementing and maintaining the system is so ugly that the benefit never emerges.  The best applications are those that provide value without much investment in setup and inputs.  Witness the SaaS revolution and the market value of  And even their software requires quite a bit of “out of band” data entry to begin handing back benefits - but the friction for technology set-up is minimal.

With the cost of hard goods rising every day, and the cost of computing, data collection, and analysis getting lower every day, more and more applications are going to appear that use the power of the ever cheaper computer to lower the start-up barriers to software value.  I think PlotWatt is a great example of one of those new applications.

I met with Luke Fishback, the CEO of PlotWatt the other day, and we share a common mind on many topics regarding how to provide value with cloud computing.  The PlotWatt application uses the nearly free computing cycles and data storage at Amazon AWS to analyze time series data reported by smart power meters.  By watching the data and asking a few questions of the user, the application “learns” the profiles of the appliances that are consuming power.  With further watching and analysis, the user begins receiving intelligence on which appliances are consuming more power than they should.  The application essentially provides a Pareto analysis of opportunity for saving money on your power bills.

So what is special about PlotWatt?  Well, historically the customer would have to wire up every appliance with sensors (or purchase expensive “smart” appliances and building control systems) to measure current and report the information to the application.  Instead, PlotWatt is using computing power and lots of data storage and analysis to eliminate the start up costs and maintenance associated with all of those sensors and proprietary building control systems.  The customer can save capital on appliance deployment (i.e. deploy dumb appliances) and still get the benefit of smart power consumption.  Cloud computing once again replaces capital with an extraordinarily affordable, variable expense approach.

Now, for the really interesting part.  Imagine correlating all of that power consumption time series data with equipment service record time series data.  Yep, you guessed right.  The application can tell you when to call the service guy.  Or it can call the service guy for you.  It can tell you if the service guy did the job right (i.e. better performance/lower power after service).  Or not.  Imagine moving beyond power to gas and water.  The plumber shows up to fix the leaky toilet before you knew it was leaking (and washing your profits down the drain).

If you can see information and data without extraordinary costs (and aggravation) for collecting it, the potential for improvement in facility maintenance costs is amazing.  Stay tuned for lots more info on how the cost of plotting everything is going to go to zero.