Files in this item

FilesDescriptionFormat

application/pdf

application/pdfBUSHMAN-THESIS-2020.pdf (4MB)
(no description provided)PDF

Description

Title:Evaluation of human-like teleoperated robot motion through performance, perception, and preference-based studies
Author(s):Bushman, Allison
Advisor(s):LaViers, Amy
Department / Program:Mechanical Sci & Engineering
Discipline:Mechanical Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:M.S.
Genre:Thesis
Subject(s):teleoperation
telepresence
joint-space control
movement analysis
human-like motion
Laban/Bartenieff Movement System (LBMS)
Abstract:Teleoperation of an articulated robot in dynamic and human-facing environments may require the operator to produce fluid, expressive, and human-like motion. This work examines the performances and perceptions of two movement profiles generated by different methods of teleoperation via an Xbox One controller. The first method is a traditional method of control in which the angles of individual joints are prescribed in a sequential fashion until the desired configuration of the robot arm is achieved. The second method of teleoperation is a choreography-inspired method of control named Robot Choreography Center (RCC), which utilizes choreographic abstractions from the Laban/Bartenieff Movement System to index a database of poses, allowing for multiple joints in the robot arm to move simultaneously. The two methods of control are compared to one another using performance, perception, and preference metrics collected in two user studies: an in-lab user study and an observer-based perception study. Success rates indicated that both methods of control were over 80% successful for static tasks requiring a specific end configuration while the choreography-inspired (RCC) was an average of 11.85% more successful for dynamic tasks requiring a transfer of momentum to achieve a desired task. These performance-based studies showed that the choreography-inspired method facilitated improved control over the robot even in functional tasks. Further analysis showed that video game exposure was positively correlated with performance level. The preference-based results from the in-lab study described the traditional benchmark method as more precise, easier to use, safer, and more articulate while the choreography-inspired (RCC) method was identified as faster, more fluid, and more expressive. These results led to the development of a perception-based study of observers conducted on a new pool of participants who were asked to select descriptive labels for the movement profiles generated by both methods of teleoperation for static and dynamic tasks. The two methods of control were described similarly when completing static tasks; however, 45% of participants selected the word "human-like" to describe the movement profile generated using the choreography-inspired (RCC) method to complete dynamic tasks. Thus, these results provide initial ideas about how qualitative descriptors of movement, such as "fluid" and "human-like", may be quantified and produced in teleoperated motion through parameters such as number of joints moving simultaneously. Similarly, when comparing the knee joints of both humans and robots, it appears that the natural system has a greater number of points of simultaneous actuation. Future work could further these quantitative models of human-assigned adjectives to motion.
Issue Date:2020-05-15
Type:Thesis
URI:http://hdl.handle.net/2142/108069
Rights Information:Copyright 2020 by Allison Bushman. All rights reserved.
Date Available in IDEALS:2020-08-26
Date Deposited:2020-05


This item appears in the following Collection(s)

Item Statistics