About

I am currently doing a PhD at Imperial College London's Computing Department on "Smart Meter Disaggregation"; I started in Oct 2011. Here's the abstract for my PhD:

    Improving electrical energy efficiency is desirable for both financial and environmental reasons.  Measuring consumption using a smart meter is a prudent first step towards managing consumption. The UK government has mandated that every house must have a smart meter installed by 2019.  Studies have demonstrated that feedback provided by smart meters can enable consumers to reduce consumption by 5-15%.  However, smart meters only measure the aggregate consumption of an entire building. Disaggregated, appliance-by-appliance information is more valuable than aggregate information for a variety of uses including behaviour change and load forecasting for the electricity grid.
    The aim of this PhD is to design, implement and evaluate a set of computational techniques capable of disaggregating smart meter signals into appliance-by-appliance information, and to do so given as little explicit information about the household as possible.  This aim can be broken into two halves: 1) a system which can automatically learn a model representation for each appliance, given data from a variety of noisy sources and 2) a system capable of searching through an aggregate data signal to infer which appliances are present in the signal and estimate the quantity of energy each device has consumed.  It is envisaged that a single, shared database of device representations will be developed such that the system only has to be trained once per appliance in order to recognise that appliance in the aggregate data stream of every user.

My PhD is funded for three years by an EPSRC Doctoral Training Account (DTA) Studentship (which I'm really greatful for!).