EDF Energy has an R&D centre in the UK with around 50 researchers in situ. The centre is an integral part of the EDF Group R&D network and works across the whole value chain of electricity; successfully leading on millions of pounds worth of international projects. It is a centre of expertise for the group benefiting from the UK's strong experience in a number of topics including Offshore Wind, Intermittency or Smart Grids and works across all EDF Energy specialisms including Nuclear, CCS, Energy Efficiency, Smart Meters, Electric Vehicles, and Digital Futures.
The R&D Centre works in partnership with a number of universities across the UK to sponsor post-graduate research projects. Information will be provided here on EDF Energy sponsored research opportunities when they arise.
The CDT does not specify the degree subject, but for our research topic a degree in Mechanical Engineering or Physics would be appropriate at 2:1 level or above.
Using BIM, sensors and phone data to improve prediction of a building’s energy consumption
This project, run in collaboration with EDF Energy, aims to use fine-grained data on activity and sensors in a building— along with basic Building Information Modelling (BIM) data — to better predict energy consumption (and consumption needs), in a visualisation platform for building data.
A phone app would let people voluntarily share their activity (e.g. location/movement) when in the building, initially via the Bluetooth beacons installed in the Library by Chalmers’ group, and sensor packs (that record temperature/noise/light/CO2 levels) from the FRuIT project (PIs: Perkins and Singer, SOCS). Susan Ashworth, University Librarian is already supporting exploratory app work. Later, we aim to apply the project’s methods to the Learning and Teaching Hub. We note that Ashworth is on the project board for the L&T Hub, and that Frank Coton has expressed a desire to apply the findings from this kind of project to it.
This PhD would not develop new types of analytics per se (e.g. new forms of statistical model), but improving energy prediction by going beyond the simple linear methods (e.g. regression models) generally used in this context, to non-linear models, by selectively extending the set of features used for analysis. By making the collection of such data economical, we aim to make it more likely that the gains in optimising energy use would more easily outweigh the costs in collecting and analysing the data.
The student would most likely have a Computer Science background, being able to do app development, handle straightforward sensor setups, and to analyse data using contemporary toolkits such as Jupyter Workbench and IBM BlueMix/Watson — bringing to bear the many tools for predictive modelling in these tools and on the web. The co-supervision role of Fiona Bradley in Engineering would be vital, as we ensure that the modelling conforms to the needs and aims of civil engineers. We would also reach out to Estates, for advice and collaboration—as, ultimately, we aim to develop methods to be used by them to reduce and optimise the energy consumption of the Library and other university buildings.