Robot Communication Via Motion: Closing the Underwater Human-Robot Interaction Loop

Abstract

In this paper, we propose a novel method for underwater robot-to-human communication using the motion of the robot as "body language". To evaluate this system, we develop simulated examples of the system's body language gestures, called kinemes, and compare them to a baseline system using flashing colored lights through a user study. Our work shows evidence that motion can be used as a successful communication vector which is accurate, easy to learn, and quick enough to be used, all without requiring any additional hardware to be added to our platform. We thus contribute to "closing the loop" for human-robot interaction underwater by proposing and testing this system, suggesting a library of possible body language gestures for underwater robots, and offering insight on the design of nonverbal robot-to-human communication methods.

Robot Communication Via Motion: A Study on Modalities for Robot-to-Human Communication in the Field

“…In the study, each participant provided their answer to the question "What is the robot trying to say in this video" as a free answer, not restricted to a number of words or a specific set of choices. To determine correctness, a rubric of five categories the answers could fall into was defined: An earlier version of this study's analysis presented in the work of Fulton et al [24] used a single expert rater following this scale. This creates a threat to internal validity from bias in the scoring of the participants' responses, so three additional raters with no other association with the study were asked to independently score each response.…”

“…2.1.1 Underwater. Motion as a vector for communication underwater was largely unexplored until the publication of our study that comprises Pilot I in this work [24]. Underwater robots have typically relied on digital displays, purpose-built interaction devices, and lights for robot-tohuman communication, although some types of motion (e.g., diver following [32,40,48]) can be thought of as implicit communication.…”

“…This was a Wizard-of-Oz study, in which the autonomous communication ability of the robots was mimicked by a human operator triggering responses. In this section, we discuss the construction and results of these pilot studies, the full versions of which may be seen in their relevant papers (Fulton et al [24] for Pilot I and Fulton et al [23] for Pilot II).…”

“…Our exploration of natural methods for robot-to-human communication underwater quickly brought us to a system of robot "body language gestures. " This system, referred to as RCVM, uses "motions with meaning, " called kinemes [24,44], to communicate interaction phrases (short, self-contained ideas or phrases) as seen in Figure 1. RCVM has a phrasebook consisting of approximately 14 phrases but could easily be expanded far beyond this.…”

Using Monocular Vision and Human Body Priors for AUVs to Autonomously Approach Divers

“…However, in recent decades, the advent of smaller AUVs [3][4][5] suitable for working closely with humans (termed co-AUVs) has enabled robots and humans to collaborate on tasks underwater. Over the last few decades, a variety of human-robot interaction (HRI) capabilities have been developed for co-AUVs to facilitate this work: from gestural programming [6][7][8] and various modes of robot communication [9,10] to diver following [11,12]. Unfortunately, underwater HRI systems typically assume that both divers and AUVs are in the ideal position for interaction, an assumption that does not always hold true.…”