Impact of Robotic kinematic variables on User Experience: Dataset on Performance, Physiological Response, and User Perception in Human-Robot Interaction during an Assembly Task
Description
This dataset presents comprehensive data derived from an experiment aimed at investigating the influence of robot kinematic variables on human-robot interaction (HRI) during assembly tasks in an industrial setting. The study sought to evaluate performance, physiological responses, and user perceptions associated with different robot kinematic configurations. Through a meticulously designed experimental procedure comprising pre-task execution, task execution, and post-task execution phases, participants engaged in an assembly task using a KUKA LBR iiwa 14 R820 collaborative robot. Two distinct robot behaviors, Slow Task (ST) and Fast Task (FT), were programmed to simulate different task conditions, allowing for a comprehensive assessment of the impact of robot kinematic variables on human factors. The dataset includes data collected from 20 volunteers (10 men and 10 women) evenly distributed across two procedures: Slow-Fast (SF) and Fast-Slow (FS). Participants' performance was evaluated based on key performance indicators, concretely, task execution time and errors. Physiological responses were measured using EEG and GSR/EDA devices, capturing variables such as Valence, Memorisation, Mental Workload, Engagement, Activation, and Impact. Perceptual indicators, including Pragmatic Quality, Hedonic Quality, Reliability, Controllability, and Perceived Usefulness, were assessed through UEQ-S and and additional self-generated questions. Key components of the dataset include: - Perceptual questionnaire (.pdf): This document contains the questionnaire provided to participants. - The Raw data (.xlsx) file consists of four tabs: The first tab contains sociodemographic information of participants, detailing gender, age, university role, robot experience, and educational background. It also presents task execution data collected during both the Slow Task (ST) and Fast Task (FT) for each participant, organized according to the experimental procedure. The second tab encompasses various T-test analyses, including comparisons between tasks, and procedures. The third tab reorganized the raw data for gender-based analysis, and the fourth tab shows additional T-test comparisons by gender across tasks, procedures, tasks within procedures. The collected data from industrial assembly tasks provides detailed perspectives on how robot kinematic variables, such as speed and acceleration, impact human performance, physiological responses, and user perceptions. This dataset can be utilized to optimize robot design, develop more intuitive user interfaces, study human factors in industrial settings, and validate human-robot interaction simulation models.
Files
Steps to reproduce
The dataset was gathered during an assembly task using a robotic system in a laboratory setting designed to emulate industrial conditions. Each of the 20 participants wore two devices: an EEG (Diadem by Bitbrain) and an EDA/GSR device (Ring by Bitbrain) on their non-dominant hand. The "Raw data.xlsx" file contains performance indicators, physiological responses, and perceptual data from each participant, enabling analysis of how robot kinematic variables influence human responses and user experience during human-robot interaction tasks. Throughout the experiment, participants went through three phases: pre-task execution, task execution, and post-task execution. Initially, participants signed a consent form detailing the test procedures and data collection methods. They then completed a socio-demographic questionnaire covering age, gender, university role, and previous robot experience, followed by instructions on how to interact with the robot using voice commands. Participants were equipped with EEG and EDA devices, and calibration was conducted using SennsLab software. The experiment comprised two procedures: Slow-Fast (SF) and Fast-Slow (FS). In SF, participants performed tasks initially in slow motion (62.5 mm/s speed, 1.75 m/s² acceleration), followed by fast motion (187.5 mm/s speed, 5.25 m/s² acceleration), with a perceptual questionnaire administered after each phase. Conversely, in FS, tasks began with fast motion, followed by slow motion, and corresponding perceptual questionnaires. Following the task, participants engaged in a free conversation to share their feelings and expectations.