Optometrist's Algorithm for Personalizing Robot-Human Handovers

TL;DR

This paper introduces the use of the Optometrist's Algorithm to personalize robot-human handovers, improving fluency and user satisfaction by tuning key interaction parameters based on user preferences.

Contribution

It presents a novel application of the Optometrist's Algorithm for interactive personalization of robot behavior in handovers, tailored to individual user preferences.

Findings

Personalization increases handover fluency.

Users find the tuning process engaging and easy.

Participants develop unique personalized handover settings.

Abstract

With an increasing interest in human-robot collaboration, there is a need to develop robot behavior while keeping the human user's preferences in mind. Highly skilled human users doing delicate tasks require their robot partners to behave according to their work habits and task constraints. To achieve this, we present the use of the Optometrist's Algorithm (OA) to interactively and intuitively personalize robot-human handovers. Using this algorithm, we tune controller parameters for speed, location, and effort. We study the differences in the fluency of the handovers before and after tuning and the subjective perception of this process in a study of $N=30$ non-expert users of mixed background -- evaluating the OA. The users evaluate the interaction on trust, safety, and workload scales, amongst other measures. They assess our tuning process to be engaging and easy to use. Personalization leads to an increase in the fluency of the interaction. Our participants utilize the wide range of parameters ending up with their unique personalized handover.