Energy metering infrastructures, i.e., smart meters (SMs), are a key component in the transition towards a more sustainable and intelligent electricity grid. Nevertheless, they have received very little attention as an enabler of user centric energy services (ES) such as demand response (DR) strategies, a key objective of the Clean Energy for all Europeans package (2019).
Achieving this goal, however, is linked to a series of challenges. First, it involves collecting data from SM systems that are not standardised among the EU Member States. Subsequently, a large amount of data must be processed to provide useful ES while also allowing for scalability. Furthermore, energy data comprise sensitive personal information, so privacy must be guaranteed at all stages.
This project is built around these elements with the following objectives: (i) the evaluation and definition of mechanisms to access SMs data; (ii) intelligent analysis of SM data for ES utilising hierarchically scalable data-driven techniques which combine non-intrusive load monitoring (NILM) for local asset identification and cluster analysis for top-level aggregation; (iii) implementation of these algorithms in a privacy-friendly manner, using computation over encrypted data (COED) techniques. This combination of research fields (ES, data science and privacy) aims to demonstrate the untapped benefits of using SMs to provide data-driven and privacy-friendly ES.
This project will be carried out at the Department of Electrical Engineering (ESAT) of KU Leuven, within the research group on Electrical Energy Systems and Applications (ELECTA) and under the supervision of Prof. Deconinck. It will also involve close collaboration with the Belgian company AE. Moreover, the research group on Computer Security and Industrial Cryptography (COSIC), also located at ESAT, will provide advice on COED techniques.
Marie Sktodowska-Curie Actions - Seal of Excellence of the Research Foundation – Flanders