Physical Human–Robot Interaction
How robots sense, interpret, and respond to contact and force during shared manipulation tasks.
Research Interests
Imitation Learning
Learning manipulation skills from human demonstrations, including contact-rich and dual-arm tasks.
Telerobotics
Remote operation systems for dexterous tasks, with a focus on transparency and shared control.
Space Robotics
Robotic construction and manipulation in extraterrestrial environments — operating reliably under communication delay and harsh conditions.
Medical & Assistive Robotics
Robotic systems that augment clinicians and assist users in healthcare and lab-automation settings.
Publications
Computer-Implemented Method for Learning from Demonstrations
Patent application · Austrian Patent Office.
Learning Forward & Reverse Skills from a Single Unfinished Demonstration for Constrained Manipulation Tasks
arXiv preprint, 2026.
Learning both forward and reverse manipulation skills from a single incomplete demonstration by combining dynamic movement primitives for free motion with screw motion primitives for contact phases.
@article{hu2026learning,
author = {Hu, Y. and Zheng, H. and Heidersberger, J. and Lee, D.},
title = {Learning Forward \& Reverse Skills from a Single Unfinished Demonstration for Constrained Manipulation Tasks},
journal = {arXiv preprint arXiv:2607.13882},
year = {2026}
}
LOPAL: Local Performance-Aware Active Learning from Imperfect Demonstrations
IEEE Robotics and Automation Letters (RA-L), vol. 11, no. 7, pp. 9040–9047, 2026.
Learning robot skills from imperfect demonstrations by combining the locally best parts of each, and actively requesting user corrections only where good data is missing.
@article{heidersberger2026lopal,
author = {Heidersberger, J. and Jadav, S. and Lee, D.},
title = {{LOPAL}: Local Performance-Aware Active Learning from Imperfect Demonstrations},
journal = {IEEE Robotics and Automation Letters},
volume = {11},
number = {7},
pages = {9040--9047},
year = {2026}
}
Partner familiarity enhances performance in a manual precision task
Scientific Reports, vol. 15, 2025.
Human-subject study showing that repeated dyadic interaction produces partner-specific coordination strategies that are retained and reused.
@article{heidersberger2025partner,
author = {Heidersberger, J. and Kaiser, J. and Jadav, S. and Mihi{\'c} Zidar, L. and Curioni, A. and Johannsen, L. and Lee, D.},
title = {Partner familiarity enhances performance in a manual precision task},
journal = {Scientific Reports},
volume = {15},
year = {2025}
}
Computational models of the emergence of self-exploration in 2-month-old infants
IEEE Int. Conf. on Development and Learning (ICDL), 2025.
Computational models of how early sensorimotor self-exploration emerges in 2-month-old infants, bridging developmental robotics and cognitive science.
@inproceedings{spisak2025computational,
author = {Spisak, J. and Benad, J. and Heidersberger, J. and Verschoor, S. and Lanillos, P. and Lee, D. and Eppe, M. and Wermter, S. and Hoffmann, M. and Tcaci Popescu, S.},
title = {Computational models of the emergence of self-exploration in 2-month-old infants},
booktitle = {IEEE International Conference on Development and Learning (ICDL)},
year = {2025}
}
REASSEMBLE: A Multimodal Dataset for Contact-rich Robotic Assembly and Disassembly
Robotics: Science and Systems (RSS), 2025.
A multimodal dataset of 4,551 contact-rich assembly demonstrations with event cameras, force/torque sensors, audio, and multi-view RGB.
@inproceedings{sliwowski2025reassemble,
author = {Sliwowski, D. and Jadav, S. and Stanovcic, S. and Orbik, J. and Heidersberger, J. and Lee, D.},
title = {{REASSEMBLE}: A Multimodal Dataset for Contact-rich Robotic Assembly and Disassembly},
booktitle = {Robotics: Science and Systems (RSS)},
year = {2025}
}
Shared Autonomy via Variable Impedance Control and Virtual Potential Fields for Encoding Human Demonstrations
* Equal contribution
IEEE Int. Conf. on Robotics and Automation (ICRA), 2024, pp. 15151–15157.
Encoding human demonstrations with variable impedance control and virtual potential fields for robust shared-autonomy skill transfer.
@inproceedings{jadav2024shared,
author = {Jadav, S. and Heidersberger, J. and Ott, C. and Lee, D.},
title = {Shared Autonomy via Variable Impedance Control and Virtual Potential Fields for Encoding Human Demonstrations},
booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
pages = {15151--15157},
year = {2024}
}
Single Action Push to Tilt Mechanism for Cell Culture Media Change within Incubators for Screw Cap Flasks
IEEE Int. Conf. on Robotics and Biomimetics (ROBIO), 2022, pp. 2243–2248.
A compact mechanism enabling automated cell-culture media change inside incubators, developed for life-science lab automation.
@inproceedings{artmann2022single,
author = {Artmann, L. and Heidersberger, J. and Lueth, T. C.},
title = {Single Action Push to Tilt Mechanism for Cell Culture Media Change within Incubators for Screw Cap Flasks},
booktitle = {IEEE International Conference on Robotics and Biomimetics (ROBIO)},
pages = {2243--2248},
year = {2022}
}
Technical Expertise
Robotics & Methods
Software
Hardware & Platforms
About
I am a PhD candidate in the Research Unit of Autonomous Systems at TU Wien, advised by Prof. Dongheui Lee, investigating how humans and robots can physically collaborate.
I previously contributed to the SOLAR project on learning body representations during dyadic collaboration, and now work on Lunar Assembly — enabling remote robotic construction under the constraints of the lunar environment, including multi-second communication delays. I hold a B.Sc. degree in Mechanical Engineering and dual M.Sc. degrees in Mechatronics & Robotics and Mechanical Engineering from TU Munich .
Education
- 2022 — present Ph.D. candidate, Autonomous Systems Lab, TU Wien
- 2022 M.Sc. Mechatronics & Robotics, TU Munich
- 2022 M.Sc. Mechanical Engineering, TU Munich
- 2018 B.Sc. Mechanical Engineering, TU Munich
Contact
I'm happy to discuss collaborations and research questions. The fastest way to reach me is by email.