In recent years, remote operation technology has been actively developed in the construction industry to address labor shortages and harsh working conditions. While such technology enables operators to control machinery from a distance and receive haptic feedback through seat-mounted actuators, the physical sensations transmitted to the operator, particularly at the waist, may inadvertently reproduce burdensome dynamics. To optimize feedback design and reduce operator strain, it is essential to identify which dynamic behaviors of the construction machinery significantly influence the operator’s perception. This study focuses on an excavator modeled as a planar 4-link system and assumes that the angular acceleration of the seat serves as the primary sensory input. We propose a method to extract key behavioral components by decomposing the joint torque at the base using the partial Lagrangian method. Through physical experiments and correlation analysis between decomposed torque terms and seat angular acceleration, we found that the inertial components of the partial torque strongly influence perceived motion. This finding provides a foundation for designing more effective teleoperation feedback systems.
Sumiyama et al. (Wed,) studied this question.