Ergonomic Optimization of a Force Feedback Joystick for Telerobotic Applications

摘要

In telerobotic systems, the human operator remotely controls a robotic system (e.g. exploration rovers or humanoids) with an input device. The performance and usability of such systems is highly dependent on haptic information displayed at the input device. Besides active force feedback (e.g. reflecting collisions of the remote system and its environment), passive force feedback (like inertia) should be provided to improve the operators sensorimotor performance. In the present study, we used a two-dimensional joystick providing high resolution force feedback for investigating the effects of passive force feedback (spring stiffness, damping and inertia) on human performance. 
\nThe N = 11 participants had to perform a series of target acquisition and pursuit tracking tasks in a simulation. While the target acquisition task encompassed 1) rapid movements to one of the eight target areas and 2) precise matching of these targets, the pursuit tracking required to match a target moving along four different axes with a sinusoidal speed profile. The mechanical properties were varied systematically, utilizing a 3 (no vs. moderate [0.22 Nm/rad] vs. high spring stiffness [0.44 Nm/rad]) x 3 (no vs. moderate [0.045 Nm*s/rad] vs. high damping [0.09 Nm*s/rad]) x 2 (no vs. moderate virtual inertia [2.2 g *m2]) within subjects design. 
\nThe results indicate that spring stiffness helps reducing overshoots during target acquisition, but also leads to significantly longer matching times. Dynamic tracking movements can be performed most accurately when being supported by moderate stiffness. While high damping negatively affects movement speed, moderate damping presents a good compromise between speed and precision with fewer overshoots, precise target matching and high tracking performance. Interestingly, we also found a positive effect of moderate inertia on target matching, indicating that inertia compensates tremor. In a next step, we will validate these findings in real telerobotic scenarios and in combination with active force feedback.