Summary
Radboud University is offering a PhD position within the NWO-funded SHARE project to develop advanced AI models for generating realistic, privacy-preserving synthetic energy data. This crucial research aims to address the challenges of data availability for grid planning and decision-making in the energy transition, particularly given privacy regulations like GDPR. The successful candidate will work with real-world data from Alliander, focusing on physics-informed generative models, and contribute to an open-source synthetic data toolbox with significant societal impact.
Designation
PhD Candidate
Scholarship Details
| Employment | Gross Monthly Salary | Required Background | Organizational Unit |
|---|---|---|---|
| 1.0 FTE | €3,204 – €4,051 | Research University Degree | Faculty of Science |
Research Area
- Physics-Informed Generative AI
- Synthetic Energy Data Generation
- Deep Generative Models (VAEs, GANs, Diffusion Models, Flow Matching, Gaussian Processes)
- Power-Flow Consistency and Physical Constraints in AI Models
- Graph Neural Networks
- Differential Privacy Integration
- Energy Systems Data Analysis
Location
Nijmegen, Netherlands (Radboud University, Institute for Computing and Information Sciences – iCIS)
Eligibility/Qualification
Required Qualifications:
- MSc degree (or expected before start date) in Computer Science, Artificial Intelligence, Data Science, Applied Mathematics, Physics, Electrical Engineering, or a related field.
- Solid background in machine learning.
- Good programming skills in Python and experience with a deep learning framework (e.g., PyTorch).
- Enjoyment of interdisciplinary work.
- Good command of spoken and written English.
Strong Plus:
- Experience with deep generative models (VAEs, GANs, diffusion models) or probabilistic modelling.
Note:
- Prior knowledge of energy systems is not required; domain context will be provided.
Scholarship Description
The PhD candidate will be a core contributor to Work Package 3 of the SHARE project, focusing on developing physics-informed, domain-constrained generative models for energy-system data. Key responsibilities include:
- Designing and comparing deep generative approaches (VAEs, GANs, diffusion models, flow matching, Gaussian processes) for load, generation, and voltage time series.
- Embedding physical constraints into generation models, such as power-flow consistency (Kirchhoff’s laws), operational bounds, and network topology via graph neural networks.
- Building validated benchmark datasets and an evaluation framework for statistical fidelity, temporal/spatial structure, physical plausibility, and downstream task performance.
- Collaborating on integrating differential privacy into the generative pipeline.
- Contributing to an open-source synthetic data toolbox for DSOs, municipalities, and researchers.
- Working with real operational data from Alliander, with direct access to practitioners.
- Publishing research at top machine learning venues.
- Spending up to 10% of time on teaching activities within computing science programs.
What Radboud University Offers:
- A temporary employment contract (1.0 FTE) of 1.5 years, extendable by 2.5 years (total 4-year contract) upon positive evaluation.
- Starting salary of €3,204 gross per month, increasing to €4,051 in the fourth year.
- 8% holiday allowance and 8.3% end-of-year bonus.
- Extra annual leave days (30 or 41 days with full-time employment, instead of statutory 20).
- Flexible working hours and various leave arrangements.
- Opportunity to compose part of employment conditions (e.g., exchanging income for extra leave, sports membership reimbursement).
- 34% discount on sports and cultural activities.
- Good pension plan.
- Ample room for talent development and training schemes.
How to Apply
Applications must be submitted via the online application button on the original job posting page. Applicants should address their letter of application to Dr. Yuliya Shapovalova.
The application form will specify required documents. In your motivation letter, please address the following questions:
- Which generative modeling approach would you use as a starting point for energy time series on a physical network, why you would choose it, and what you expect the biggest challenge to be (focus on reasoning).
- Describe a software application or data analysis project you are proud of: what it does, its challenges, and your contribution (optional link).
- Provide a one-paragraph description of your MSc thesis, written for a non-expert audience.
The preferred start date for employment is 1 January 2027.
Last Date for Apply
25 October 2026
