Updating Robot Safety Representations Online from Natural Language Feedback

TL;DR

This paper presents a method for robots to update safety constraints in real-time using natural language feedback and vision language models, enabling safer operation in dynamic, human-centered environments.

Contribution

It introduces a novel approach combining vision language models with online safety updates, allowing robots to interpret language feedback and adapt safety constraints during deployment.

Findings

Robots successfully infer safety constraints from language and vision data.

The approach improves safety adherence in dynamic environments.

Hardware experiments validate real-world applicability.

Abstract

Robots must operate safely when deployed in novel and human-centered environments, like homes. Current safe control approaches typically assume that the safety constraints are known a priori, and thus, the robot can pre-compute a corresponding safety controller. While this may make sense for some safety constraints (e.g., avoiding collision with walls by analyzing a floor plan), other constraints are more complex (e.g., spills), inherently personal, context-dependent, and can only be identified at deployment time when the robot is interacting in a specific environment and with a specific person (e.g., fragile objects, expensive rugs). Here, language provides a flexible mechanism to communicate these evolving safety constraints to the robot. In this work, we use vision language models (VLMs) to interpret language feedback and the robot's image observations to continuously update the robot's representation of safety constraints. With these inferred constraints, we update a Hamilton-Jacobi reachability safety controller online via efficient warm-starting techniques. Through simulation and hardware experiments, we demonstrate the robot's ability to infer and respect language-based safety constraints with the proposed approach.