Low-Burden LLM-Based Preference Learning: Personalizing Assistive Robots from Natural Language Feedback for Users with Paralysis
arXiv:2604.01463v2 Announce Type: replace-cross Abstract: Physically Assistive Robots require personalized behaviors to ensure user safety and comfort. However, traditional preference learning methods, like exhaustive pairwise comparisons, cause substantial physical and cognitive fatigue for users with severe motor impairments. To solve this, we propose a low-burden, offline framework that translates unstructured natural language feedback directly into deterministic robotic control policies. To safely bridge the gap between ambiguous human speech and robotic code, our pipeline uses Large Langu
Advances in large language models (LLMs) and robotic control are converging, making personalized human-robot interaction through natural language feasible for complex applications like assistive technology.
This development allows for a more intuitive and less burdensome method for individuals with severe motor impairments to personalize assistive robots, significantly improving quality of life and accessibility.
The paradigm for preference learning in assistive robotics shifts from laborious, repetitive physical tasks to natural language interaction, broadening the accessibility and usability of these devices.
- · Individuals with motor impairments
- · Assistive robot manufacturers
- · AI developers specializing in HRI
- · Healthcare providers
- · Traditional preference learning methods
Assistive robots become more widely adopted due to increased ease of personalization and use by a broader population.
This methodology could be generalized to other complex human-robot interaction scenarios, accelerating the development of more adaptable and user-friendly robotic systems across various sectors.
The success of natural language control for assistive robots might drive regulatory changes and ethical considerations regarding autonomous systems interacting directly with vulnerable populations.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI