Jul 29, 2024

Announcing the Launch of the OpenSynth Model Repository: Revolutionizing Access to Energy Demand Data

LF Energy is thrilled to announce that the OpenSynth Model Repository is now live! This marks a significant milestone for the energy industry and the global community of researchers, modelers, and policymakers dedicated to advancing energy transitions.

What is OpenSynth?

OpenSynth, originated by Centre for Net Zero (CNZ) and now open sourced under LF Energy, is an open community initiative aimed at generating and sharing synthetic energy demand data. In a world where access to detailed energy demand profiles is crucial for optimizing the grid in real-time, OpenSynth provides an innovative solution to the restrictive access to smart meter data due to privacy protections.

The OpenSynth Model Repository

You can now explore and contribute to the OpenSynth Model Repository at the following links:

Our repository includes algorithms and a common evaluation framework to benchmark the performance of various synthetic data generation algorithms.

Why OpenSynth Matters

Access to smart meter data is essential for understanding and adapting to changing energy demand profiles. Traditional global energy modeling relies on static and highly aggregated data, which fails to capture the dynamic and bidirectional nature of modern energy flows. With the rise of variable renewable energy sources and behind-the-meter technologies such as heat pumps, electric vehicles, and batteries, the need for granular, real-time demand data has never been greater.

Synthetic Data: The Fastest Path to Access

Rather than challenging current data regulations and smart meter legislation, OpenSynth believes that generating synthetic data is the fastest way to achieve widespread, global access to detailed energy demand datasets. Our synthetic data will include important metadata such as property type, Energy Performance Certificate rating, and ownership of low carbon technologies (LCTs). This will enable a deeper understanding of behind-the-meter changes and support the development of future demand profiles for diverse demographics.

Recent Developments

Over the past few months, we've made significant progress:

  • Faraday Model Algorithm: CNZ’s Faraday model algorithm is now live in the OpenSynth Model Repository. For more details, refer to the paper presented at the “Tackling Climate Change with Machine Learning” Workshop at the 12th International Conference on Learning Representation.
  • Evaluation Framework: CNZ has released a paper titled “Defining ‘Good’: Evaluation Framework for Synthetic Smart Meter Data” in collaboration with Dr. Phil Grunewald (University of Oxford), Prof. Pascal Van Hentenryck (Georgia Tech), and Asst. Prof. Priya L. Donti (MIT). This paper proposes a common evaluation framework using the concepts of Fidelity, Utility, and Privacy.

What's Next?

In the coming months, the OpenSynth community plans to:

  • Release synthetic data generated with CNZ’s Faraday model, trained on Octopus Energy datasets, into the Data Repository. This data will be conditioned on property type, energy performance rating, and LCT ownership. Synthetic data will also be shared by community members. 
  • Implement the evaluation framework described in the ‘Defining Good’ paper.

Join Us at the LF Energy Summit

OpenSynth is excited to announce that we will be presenting a workshop at the upcoming LF Energy Summit in Brussels this September. The workshop will cover:

  • Training a generative model
  • Generating synthetic data
  • Evaluating data quality using our framework

We encourage attendees with basic familiarity with neural networks or machine learning to join us and learn how to generate and evaluate synthetic smart meter data. Additionally, we will hold an OpenSynth community meeting to discuss project contributions, community roles, and our product roadmap. 

OpenSynth is poised to revolutionize the way we access and utilize energy demand data. By building a community of data holders and innovators, we aim to democratize access to energy demand data and drive the global transition to a more sustainable energy future.

LF Energy | https://lfenergy.org/