OmniCharger: CNN-based Hand Gesture Interface to Operate an Electric Car Charging Robot through Teleconference
Name
3659060.pdf
Size
21.99 MB
Format
Adobe PDF
Checksum (MD5)
e42a92255921185e4000d6d88b45e938
Author(s)
Altamirano Cabrera, Miguel
Rakhmatulin, Viktor
Fedoseev, Aleksey
Sautenkov, Oleg
Alyounes, Oussama
Puchkov, Andrei
Tsetserukou, Dzmitry
Date Issued
September 25, 2024
Journal
ACM Transactions on Human-Robot Interaction
Publisher
ACM
Citation
Altamirano Cabrera, Miguel, Rakhmatulin, Viktor, Fedoseev, Aleksey, Sautenkov, Oleg, Alyounes, Oussama et al. 2024. "OmniCharger: CNN-based Hand Gesture Interface to Operate an Electric Car Charging Robot through Teleconference." ACM Transactions on Human-Robot Interaction, 13 (3).
Version
Final published version
Abstract
The automation of the car charging process is motivated by the rapid development of technologies for self-driving cars and the increasing importance of ecological transportation units. However, it remains challenging to precisely position the charger plug autonomously due to the sensitivity of CV algorithms to lighting and weather conditions. We introduce a novel robotic teleoperation system based on hand gesture recognition through teleconferencing software. The users, connected by teleconference, use their hand gestures to teleoperate the electric plug located on the collaborative robot end-effector. We conducted a user study to evaluate the system performance and suitability using OmniCharger and two baseline interfaces (a UR10 Teach Pendant and a Logitech F710 Wireless Gamepad). Except for two trials, all the users were able to locate the plug inside of a 5 cm target using the interfaces. The distance to the target and the orientation error did not present statistically significant differences in the use of the three interfaces. The NASA TLX questionnaire results showed low values in all the sub-classes, the SUS results rated the usability of the proposed interface above average (68\%), and the UEQ showed excellent performance of the OmniCharger interface in the attractiveness, stimulation, and novelty attributes.
Terms of Use
Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1145/3659060
