Internship at Siemens Digital Industries Software, Leuven, Belgium

Created on February 14, 2026

2026

As part of my doctoral research, I worked across industry and academia on virtual vibration testing, structural dynamics, digital twins, and reinforcement learning. The secondment provided an opportunity to connect high-fidelity experimental and numerical modelling with emerging intelligent test-design methods.

Siemens Digital Industries Software

At Siemens, I worked on virtual shaker testing frameworks for spacecraft qualification. My work included:

  • Scaling and enhancing virtual shaker testing methods using advanced frequency-based substructuring (FBS), enabling more accurate prediction of shaker-structure interactions and reducing physical test risk.
  • Designing and modelling complex aerospace-emulator test structures to validate hybrid FBS-based virtual shaker methodologies, integrating measured and simulated frequency response functions (FRFs) to build high-fidelity virtual systems.
  • Characterising the dynamic behaviour of a high-capacity slip-table shaker, including its active and driving-point FRFs, to support real-time virtual test execution.
  • Collaborating with a multidisciplinary team to integrate finite-element models, experimental data, and real-time execution workflows for scalable virtual vibration test environments.
  • Supporting the transition from emulation to hybrid simulation workflows, enabling earlier test planning, derisking, and greater confidence in spacecraft vibration qualification strategies.

KU Leuven

At KU Leuven, I developed and applied reinforcement-learning and unsupervised-environment-design methods for intelligent sensing and structural testing. This work included:

  • Developing a manifold-constrained adversarially compounding complexity by editing levels (mc-ACCEL) framework for open-ended reinforcement learning in digital-twin-based structural health monitoring.
  • Applying unsupervised environment design to the optimisation of modal tests at the Labo Voertuigtechnologie en Lichtgewicht-constructies (LVL), combining reinforcement learning with physics-based structural-dynamics constraints to improve sensor placement, test efficiency, and structural insight.
  • Designing a vibration test jig for the CubeSpec CubeSat spectrometer, supporting pedagogical qualification testing on a G&W DSA4-8k shaker.

This secondment strengthened the connection between simulation, experimentation, and intelligent decision-making in structural testing, while providing valuable experience across both industrial and academic research environments. Many thanks to my supervisors Prof. David Wagg, Dr. Timothy Rogers and Dr. Mattia Dal Borgo.