Core Concepts
Human demonstration using a virtual marker is faster, superior in quality, and imposes less workload than kinesthetic teaching.
Abstract
In the study comparing human demonstration with a virtual marker to kinesthetic teaching, it was found that the former method was significantly faster, of higher quality, and imposed less overall workload. The research focused on simplifying robot programming through Programming by Demonstration (PbD) methods. Participants demonstrated drawing tasks using both methods, and their experiences were evaluated through NASA raw Task Load Index (rTLX) and system usability scale (SUS). Results showed that the virtual marker induced less physical and mental demand while maintaining high performance rates compared to kinesthetic teaching. The study also highlighted potential biases in user ratings affecting study outcomes. Additionally, trajectory analysis revealed that demonstrations with the virtual marker were more consistent and accurate compared to those with the robot manipulator.
Stats
Human demonstration using a virtual marker is on average 8 times faster than kinesthetic teaching.
Human demonstration imposes 2 times less overall workload than kinesthetic teaching.