Sound Matters: Auditory Detectability of Mobile Robots

TL;DR

A user study measuring auditory detection distances for a wheeled and quadruped robot, which emit different consequential sounds when moving showed that the quadruped robot sound was detected significantly better than the wheeled one, which demonstrates that the movement mechanism has a meaningful impact on the auditory detectability.

Abstract

Mobile robots are increasingly being used in noisy environments for social purposes, e.g. to provide support in healthcare or public spaces. Since these robots also operate beyond human sight, the question arises as to how different robot types, ambient noise or cognitive engagement impacts the detection of the robots by their sound. To address this research gap, we conducted a user study measuring auditory detection distances for a wheeled (Turtlebot 2i) and quadruped robot (Unitree Go 1), which emit different consequential sounds when moving. Additionally, we also manipulated background noise levels and participants' engagement in a secondary task during the study. Our results showed that the quadruped robot sound was detected significantly better (i.e., at a larger distance) than the wheeled one, which demonstrates that the movement mechanism has a meaningful impact on the auditory detectability. The detectability for both robots diminished significantly as background noise increased. But even in high background noise, participants detected the quadruped robot at a significantly larger distance. The engagement in a secondary task had hardly any impact. In essence, these findings highlight the critical role of distinguishing auditory characteristics of different robots to improve the smooth human-centered navigation of mobile robots in noisy environments.

Figures (5)

• Figure 1: This study explores the auditory detectability of two mobile robots in low vs. high background noise, and with vs. without cognitive engagement of the human observers. Our findings serve as a valuable foundation for the intelligent design of HRI tailored to specific human needs.
• Figure 2: Schematic of our user study design showing a) a photo of a participant conducting the experiment. b) Variables: We investigate the audible distance $d_{HR}$ of robot detection beyond human sight based upon the independent variables IV1 consequential robot sound type, IV2 environment noise level and IV3 engagement of the participant in a secondary (one-back) task. The auditory detection distance of a robot for three independent variables is measured via the transformed up-down adaptive procedure. c) Characteristic of each trial: In one trial, the participant is first presented with an auditory stimulus followed by a question. Each auditory stimulus, consists of two intervals A and B of 1.5s each, each preceded by a 1.0s silence, respectively. Note that the robot sound is only presented in one of the intervals. The user has to tell apart the audibility of the robot from the environment noise by choosing the interval that presumably contained the robot sound. d) IV levels: All 8 combinations of the 3 independent variables are presented in an interleaved, random manner to the participant. They are, however, grouped and presented in two different blocks - with and without secondary task to make it simpler to answer workload related question for the participants.
• Figure 3: Spectrogram of the sound profiles of a) the quadruped robot, showing interleaved high energy moments observed in the bands shown, b) the wheeled robot, which is similar in terms of energy spread over time as environment sound depicted in c).
• Figure 4: Histograms of the individual gains in dB (light blue bars) for the quadruped robot to match the loudness of the wheeled robot at a distance of 1 m (bin width = 0.5 dB). The dark blue horizontal box indicates the 95% confidence interval, the vertical line within the confidence interval represents the mean. Note that the mean was not significantly different from 0, indicating that both sounds were perceived to be rather similarly loud.
• Figure 5: Mean detection distances for the two robot sounds as a function of background noise (x-axis) and secondary task (left: present, right: absent). Red triangle: Quadruped robot. Blue circle: Wheeled robot. Error bars indicate $\pm1$ SE of the individual means.