Data Interpretation

Energy Disaggregation

Jack Kelly jack.kelly@imperial.ac.uk(Swipe or press arrow keys on your keyboard to change slides)

What is Energy Disaggregation?

Aggregate Energy Bill

Itemised Energy Bill

Why bother with disaggregation?

GB Smart Meter Roll-out

• All homes to have a smart meter by 2020.
• The business case assumes that smart meters will drive savings of £4.6 billion due to reduced energy consumption (across both electricity and gas).
• Aggregate data alone unlikely to lead to such savings.
• Disaggregation has been demonstrated to drive savings more effectively than aggregate data alone.

My Work

• First to apply deep neural nets to energy disaggregation

Recurrent Neural Nets

Example Output from Deep Neural Net

My Work

• First to apply deep neural nets to energy disaggregation
• Lead developer on open source disaggregation tool NILMTK
• Collected & released disaggregated energy dataset: UK-DALE
• NILM Metadata: metadata schema for energy data
  • Used by > 9 datasets

Questions?

Deep Neural Nets

ImageNet Large Scale Visual Recognition Challenge (ILSVRC)

From: Krizhevsky, Sutskever & Hinton. ImageNet Classification with Deep Convolutional Neural Networks. NIPS (2012)

Image from devblogs.nvidia.com

Krizhevsky et al.'s DNN Results on ImageNet 2012

Krizhevsky, Sutskever & Hinton. ImageNet Classification with Deep Convolutional Neural Networks. NIPS (2012)

Recurrent Neural Nets

Denoising Autoencoders

Bounding Rectangle

Example Output from Deep Neural Net