PhD Position: Physics-Informed Generative AI for Synthetic Energy Data

Radboud Universiteit; Nijmegen; On-site; EUR 36,000 - 45,000; Full time

Benefits offered by this job

Holiday allowance (8%)

End-of-year bonus (8.3%)

Pension plan

Flexible leave & work hours

Job summary

Radboud University Nijmegen invites applications for a PhD position focused on Physics-Informed Generative AI for Synthetic Energy Data. You will develop physics-constrained generative models for energy-system time series and publish in leading ML venues, using real data from Alliander and contributing to an open-source toolbox.

You will collaborate with privacy researchers, legal scholars and energy practitioners, and spend up to 10% of time on teaching activities within computing science

Qualifications

• MSc degree (or will obtain one before start) in CS/AI/Data Science/Applied Math/Physics/EE or related field.
• Strong background in machine learning; experience with deep generative models (VAEs, GANs, diffusion) or probabilistic modelling is a plus.
• Good programming skills in Python and experience with a deep learning framework such as PyTorch.
• Interdisciplinary work with privacy researchers, legal scholars and energy-sector practitioners.
• Good command of spoken and written English.

Responsibilities

• Design and compare deep generative approaches, UAEs, GANs, diffusion models/flow matching, and Gaussian processes for realistic load, generation and voltage time series.
• Embed physical constraints into generation: power-flow consistency, operational bounds, and network topology through graph neural networks.
• Build validated benchmark datasets and an evaluation framework covering fidelity, temporal and spatial structure, plausibility and downstream task performance.
• Collaborate with a fellow PhD candidate and postdoctoral researcher on integrating differential privacy into the generative pipeline.
• Contribute to an open-source synthetic data toolbox for use by DSOs, municipalities and researchers in the Netherlands.

Skills

Python programming

PyTorch

Machine learning background

Generative models

Interdisciplinary collaboration

English proficiency

Education

MSc degree in CS/AI/Data Science/Applied Math/Physics/EE

Tools

Python

Graph Neural Networks

Job description

In the PhD Position: Physics-Informed Generative AI for Synthetic Energy Data, you will develop physics-informed, domain-constrained generative models for energy-system data.

In the NWO-funded SHARE project, you will develop AI models that generate realistic, privacy-preserving synthetic energy data for grid planning and decision-making. Working with real-world data from Alliander, you will publish at leading machine learning venues while building tools with tangible impact on the Dutch energy sector.

Synthetic data offers a way out: realistic-but-artificial datasets that preserve the statistical, temporal and physical structure of real energy data without identifying real households or companies. Energy data consists of time series living on a physical network, governed by power-flow equations. Off-the-shelf generative models produce data that looks plausible but violates physics and is therefore of limited use for grid planning.

More concretely, your work will involve the following:

• You will design and compare deep generative approaches, UAEs, GANs, diffusion models/flow matching, and Gaussian processes for realistic load, generation and voltage time series.
• You will embed physical constraints into generation: power-flow consistency (Kirchhoff's laws) as soft or hard constraints, operational bounds (voltage limits, capacity, phase balance), and network topology through graph neural network architectures.
• You will build validated benchmark datasets and an evaluation framework covering statistical fidelity, temporal and spatial structure, physical plausibility and downstream task performance (e.g. train-synthetic-test-real forecasting).
• You will collaborate with a fellow PhD candidate and a postdoctoral researcher on integrating differential privacy into the generative pipeline, balancing privacy guarantees against data utility.
• You will contribute to an open-source synthetic data toolbox that DSOs, municipalities and researchers across the Netherlands will actually use.

This is research with a direct route to impact: you will work with real operational data from Alliander, with regular on-site visits and direct access to the practitioners who will use your models for congestion forecasting, spatial energy planning and flexibility assessment. You will publish at top machine learning venues while producing open datasets and tools with tangible societal impact.

You will be expected to spend a small part of your time (up to 10%) on teaching activities, such as assisting in courses of our computing science programmes.

Does this sound like you?

• You hold an MSc degree (or will obtain one before the starting date) in computer science, artificial intelligence, data science, applied mathematics, physics, electrical engineering, or a related field.
• You have a solid background in machine learning; experience with deep generative models (VAEs, GANs, diffusion models) or probabilistic modelling is a strong plus.
• You have good programming skills in Python and experience with a deep learning framework such as PyTorch.
• You enjoy interdisciplinary work: you will interact with privacy researchers, legal scholars and energy-sector practitioners.
• You have a good command of spoken and written English.

Prior knowledge of energy systems is not required, we and our consortium partners will provide the domain context.

What we offer you

• We will give you a temporary employment contract (1.0 FTE) of 1.5 years, after which your performance will be evaluated. If the evaluation is positive, your contract will be extended by 2.5 years (4-year contract).
• You will receive a starting salary of €3,204 gross per month based on a 38-hour working week, which will increase to €4,051 in the fourth year.
• You will receive an 8% holiday allowance and an 8,3% end-of-year bonus.
• You will receive extra days off. With full-time employment, you can choose between 30 or 41 days of annual leave instead of the statutory 20.

Additional employment conditions

You can make arrangements for the best possible work-life balance with flexible working hours, various leave arrangements and working from home. You are also able to compose part of your employment conditions yourself. For example, exchange income for extra leave days and receive a reimbursement for your sports membership. In addition, you receive a 34% discount on the sports and cultural activities at Radboud University as an employee. And, of course, we offer a good pension plan. We also give you plenty of room and responsibility to develop your talents and realise your ambitions. Therefore, we provide various training and development schemes.

Where you will be working

You will be embedded in the Data Science section of the Institute for Computing and Information Sciences (iCIS) at Radboud University in Nijmegen. iCIS conducts world-class research in machine learning, software science and digital security and consistently ranks among the top computer science institutes in the Netherlands. The atmosphere is informal, international and collaborative.

The SHARE consortium, funded by the NWO Knowledge and Innovation Covenant programme, brings together Radboud University, the Dutch Open University, DSO Alliander, national metrology institute VSL, Zenmo, Bronscode and the Municipality of Nijmegen, spanning AI, privacy engineering, energy systems, law and governance. You will be supervised by Dr Yuliya Shapovalova (probabilistic machine learning, time series) and Prof. Tom Heskes (machine learning and artificial intelligence).

Faculty of Science

The Faculty of Science (FNWI), part of Radboud University, engages in groundbreaking research and excellent education. We seek solutions to major societal challenges, such as cybercrime and climate change, and work on major scientific challenges, such as those in the quantum world. The faculty has a strong international character and provides an informal, accessible and welcoming environment, with attention and space for personal and professional development.

At Radboud University, I can fully focus on expanding my expertise while learning from my peers and mentors. Noemí Segura-Solé PhD candidate in Microbial Ecology

You will preferably start your employment on 1 January 2027.