Towards Personalizing Shared Autonomy with Human-in-the-Loop Robot Programs

Abstract

We introduce a toolkit enabling users to specify intervention points in robot programs for customized task execution. Our user study with eight participants of varying expertise examined how people choose to insert themselves into control loops during complex manipulation tasks. Key findings indicate that integrating users into the loop in shared autonomy improves task success rates, perceived sense of control, and user experience, while generating intervention data that supports robot adaptation across different user preferences.

BibTeX Entry

@inproceedings{chen2025towards,
  author = {Chen, Daphne and Vuddagiri, Sahitha and Yang, Claire and Walker, Nick and Cakmak, Maya},
  title = {Towards Personalizing Shared Autonomy with Human-in-the-Loop Robot Programs},
  year = {2025},
  month = sep,
  booktitle = {CoRL 2025 Workshop on Human-to-Robot},
}