Errors matter! The Influence of Error Frequency and Timing on Fluency and Workload in a Virtual Human-Cobot Collaboration Environment

ACM Symposium on Virtual Reality Software and Technology (VRST), pp. 1–8

Abstract

Cobots are widespread among manufacturing industries, where humans and robots work together in assembly tasks. However, cobots are not free from errors, which may affect collaboration in terms of human-robot workflow fluency. We present a study (N = 21) where we simulate a manufacturing workcell in VR to investigate how a cobot’s faulty behavior can affect the perception of human-robot collaboration during assembly, considering fluency (human-robot fluency, robot relative contribution, positive teammate traits, trust and human-robot working alliance goal) and workload. For that, participants assembled, together with a cobot, a complex piece inside electric vehicles (gearbox e-axle) five times. The assembly process consisted of six different steps to be performed sequentially, and we implemented a faulty behavior where the robot would follow the wrong sequence. After an error occurred, we allowed participants to choose freely between two error recovery behaviors: an automatic error recovery (the robot would undo the last step and continue the assembly process) or a training mode (the human would indicate the right sequence to the robot). Our results confirm that a cobot’s faulty behavior negatively affects the perceived fluency and workload, but does not affect the perceived cobot’s relative contribution to the task. The temporal sequencing of an error (error timing) can affect the goal of the human-cobot working alliance, assembly duration, temporal demand, and effort. Also, error frequency impacts differently the perceived physical demand. These findings can inform design choices of human-robot collaborative workflows in manufacturing, specifically, in error handling and mitigation strategies that preserve a positive perception of human-robot teams.

Authors 3

1. Hamd Mehfooz-Khan

   Technische Universität Ilmenau

   Affiliation as printed

   Audiovisual Technology Group, Technische Universität Ilmenau, Ilmenau, Germany

2. Stephanie Arévalo Arboleda Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen, Aachen, Germany

3. Alexander Raake Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen, Aachen, Germany