Evaluating Efficiency and Engagement in Scripted and LLM-Enhanced Human-Robot Interactions

TL;DR

This study compares scripted and LLM-enhanced human-robot interactions, finding that while LLMs improve subjective engagement, scripted interactions perform similarly in efficiency and focus, with advantages in latency and energy use for simple tasks.

Contribution

It provides an empirical comparison of scripted versus LLM-enhanced interactions, highlighting their respective strengths and limitations in industrial robot tasks.

Findings

LLM-enhanced interactions increase subjective engagement.

Scripted interactions match LLMs in efficiency and focus for simple tasks.

Scripted responses have lower latency and energy consumption for repetitive interactions.

Abstract

To achieve natural and intuitive interaction with people, HRI frameworks combine a wide array of methods for human perception, intention communication, human-aware navigation and collaborative action. In practice, when encountering unpredictable behavior of people or unexpected states of the environment, these frameworks may lack the ability to dynamically recognize such states, adapt and recover to resume the interaction. Large Language Models (LLMs), owing to their advanced reasoning capabilities and context retention, present a promising solution for enhancing robot adaptability. This potential, however, may not directly translate to improved interaction metrics. This paper considers a representative interaction with an industrial robot involving approach, instruction, and object manipulation, implemented in two conditions: (1) fully scripted and (2) including LLM-enhanced responses. We use gaze tracking and questionnaires to measure the participants' task efficiency, engagement, and robot perception. The results indicate higher subjective ratings for the LLM condition, but objective metrics show that the scripted condition performs comparably, particularly in efficiency and focus during simple tasks. We also note that the scripted condition may have an edge over LLM-enhanced responses in terms of response latency and energy consumption, especially for trivial and repetitive interactions.