IDEAL

Description: Intelligent Domestic Energy Advice Loop

PI: Dr Nigel Goddard, University of Edinburgh

Fund: £1.7m BuildTEDDI

Project lifespan: March 2013 to March 2017

Contact: nigel.goddard@ed.ac.uk

Website: http://www.energyoracle.org/

Aims

We will create the following Intelligent Domestic Energy Advice Loop (hence the acronym IDEAL): a) detailed in-home sensing, sufficient to b) infer specific demand-related behaviours, enabling c) timely personalised behavioural feedback. We hypothesise that this loop can be tuned to improve reduction in energy demand from dwellings compared to the state-of-the-art consumption feedback methods, especially where behaviours are not easily identified from consumption feedback by non-experts (heating, cooling, etc). An exciting aspect of this loop over the length of the study is that we will be able to modify the feedback we give householders, using the behaviour changes that we observe and infer as due to our feedback-we will be getting feedback about our feedback!

Methodology

Our survey design includes 576 houses. The design will include households from demographics of significant social and environmental interest: 1) affluent/technically able, with the potential for high carbon savings, 2) low-income/fuel-poor, with the potential for significant social benefits as well as some carbon savings. The outcome will be a set of conclusions useful for policy setting, for commercial offerings targeted at different market segments and households, and for further studies. The study will involve Recruitment and Householder Journey; Installation and initial data acquisition; Real-time Sensors; and Post-installation data acquisition. We will develop methods for delivering timely personalised, multimodal feedback about energy related behaviours based on the description of these behaviours and other data produced by the Probabilistic Modelling component. It will also produce the feedback for controls, modelled on that provided by standard utility Smart Meters, developed in the Pilot Phase.

Key outputs

Machine Learning and Multimedia Content Generation for Energy Demand Reduction

Nigel Goddard, Johanna Moore, Charles Sutton, Janette Webb, Heather Lovell.

(available at: http://www.energyoracle.org/publications.html)