Human-robot interaction requires robots to process language incrementally, adapting their actions in real-time based on evolving speech input. Our approach enables continuous adaptation to dynamic linguistic input, allowing robots to update motion plans without restarting execution. We evaluate our framework in real-world human-robot interaction scenarios, demonstrating online adaptions of goal poses, constraints, or task objectives.
@inproceedings{abramsetal25iros,
title={Incremental Language Understanding for Online Motion Planning of Robot Manipulators},
author={Mitchell Abrams and Thies Oelerich and Christin Hartl-Nesic and Andreas Kugi and Matthias Scheutz},
year={2025},
booktitle={Proceedings of IROS},
url={https://hrilab.tufts.edu/publications/abramsetal25iros.pdf}
}