Exploring the potential of generative AI based data synthesis for the future energy system: data, models, biases and implications (EPAi)

Description

Heating and cooling are central aspects of reducing energy use in the residential sector. However, predicting the energy performance of buildings is currently a difficult task, partly due to energy use data from end-users being inaccessible or sometimes missing. Therefore, this project aims to develop a framework for generating synthetic data for the Swedish built environment to improve energy predictions and peak load management in heating systems. In doing so, we aim to create a better understanding of how energy is used, by whom, and at what times. This will aid in planning for peak loads on the grid as well as improve the possibility of predicting end-users' energy demand and more accurate designs of heating and cooling energy systems.

The project will also contribute to future energy systems in a broader sense by developing a socio-technical and ethical methodological framework for working with synthetic energy data. The project will map available data together with stakeholders' needs. Data will be collected and prepared as training data to develop and evaluate a model for synthetic data. The project team is interdisciplinary, which will ensure the integration of socio-technical, ethical and gendered aspects of energy use and synthetic data.