Search

GEMSE4 Smart Grids

About

AI-powered real-time grid state estimation

GEMSE addresses the challenge of maintaining power grid stability amidst the increasing variability introduced by renewable energy sources. Traditional state estimation methods lack the speed and accuracy needed for real-time applications. By utilising an AI-driven, graph-based model, this solution integrates sensor data across subsystems, improving accuracy and reducing computational time to milliseconds. Successfully tested in a Swiss hydro-power plant, this approach enhances grid reliability, scalability, and promotes sustainability in renewable energy management

Trends

# Energy

Disciplines

# Machine Learning

University

Team

  • Raffael Theiler

    Raffael Theiler

    Raffael Theiler is a dedicated PhD student at the IMOS lab at EPFL. He holds a Master’s degree in "Neural Systems and Computation" from ETH and has a strong background in Neuroinformatics. Raffael’s research centers on the development of advanced AI-driven predictive algorithms, with a focus on applications in power generation and railway systems. His current work involves Transformer models for load forecasting and Graph Neural Networks to enhance short-term forecasting of pumped-storage hydroelectricity. Raffael is committed to translating his research into practical solutions, with the aim of founding a startup focused on state estimation technology. more

  • Mengjie Zhao

    Mengjie Zhao

    Mengjie Zhao is a passionate Ph.D. researcher at the IMOS lab, EPFL, specializing in developing graph-based methods to model and understand the complex dynamics of industrial systems. With a solid foundation in computational mechanics, she is driven by a commitment to solving real-world challenges that remain at the frontier of research and application. Beyond academia, Mengjie is an active contributor to the Swiss startup ecosystem. Her entrepreneurial spirit fuels her ambition to transform cutting-edge research into innovative products that address critical industry needs.more

    LinkedIn

Discover
the
Project


Similiar Projects

3D-Printed CSP Receiver

3D-Printed CSP Receiver

Temperature control

3D-printed solar receiver for industrial heat

ATOM H2

ATOM H2

Energy consumption

ELISAVA

Efficient, solid-state hydrogen storage for clean energy

BioPower Cells

BioPower Cells

Energy alternatives

Northumbria University

Biodegradable batteries from organic waste

E2Pilot

E2Pilot

Algae-based bio-fuels

Fuel-efficient routing for heavy-duty trucks