Master’s Thesis “Graph Neural Networks for Predictive Maintenance and Environmental Monitoring in Swarms of Micro-Hydropower Plants”
In recent years, the landscape of energy distribution systems has changed significantly. The modernization and diversification of energy grids have brought major challenges and opportunities, and the emergence of smart grids has ushered in an era of massive data generation and exchange in this field. Data from multiple measurement devices distributed across the grid is now available, which, combined with the capabilities of modern machine learning methods, opens up opportunities for a more resilient, efficient, and sustainable energy infrastructure.
The goal of the project is to develop an automated process for robust and accurate predictive maintenance, diagnostics, and environmental monitoring in a new form of renewable energy consisting of a large number of run-of-river hydropower plants. To this end, learning methods must be developed that incorporate graph neural networks, taking into account topological information as well as external knowledge sources. Our research is based on the integration of data from heterogeneous measurement devices at various locations and builds upon elements from previous research on fault localization in medium-voltage grids.
Your tasks:
- Research on the current literature regarding graph neural networks
- Extending methods from previous research on graph neural networks for fault location in low- and medium-voltage grids
- Scaling and parallelizing simulations and training, e.g., using cloud computing, containerization, and virtualization techniques (Cloud/Docker/Kubernetes/VMs)
- Development of scripting solutions to optimize data pipelines (Python/Bash)
- Integration of data from heterogeneous sensors in a 5G network
- Development and improvement of graph neural network models for forecasting, diagnostics, and predictive maintenance (Python/PyTorch/MATLAB)
- Integration of river topology and environmental measurements into dispersion models
- Code version control and issue management (Git)
- Training and co-development of new methods and ML models
- Possible co-authorship of research papers and assistance with dissemination.
Your profile:
- A student in the final semesters of an M.Sc. program in computer science,
- Electrical Engineering, or related fields
- Experience in software development
- (Preferred) experience with Linux
- Knowledge of neural networks, machine learning, and ideally graph neural networks
- Self-motivated and organized work style
- Good communication skills in English
- (Preferred) communication skills in German
Our offer:
- An international and dynamic work environment
- Flexible working hours and a location with excellent transportation access
- The opportunity to conduct research on challenging and exciting topics in the fields of machine learning and new technologies for renewable energy
- The opportunity to contribute to the creation, publication, and dissemination of research articles
Have we piqued your interest?
If so, please submit your application, including a cover letter, a detailed resume, and a recent transcript.
Job ID: ASCI-MA-01-2026
Contact: Lana Amaya personal@fortiss.org