Effect of Variable Visual-Feedback Delay on Movement Time

Predictive Communication Quality Control in Haptic Teleoperation With Time Delay and Packet Loss

“…They were also told that avoiding collisions should have a higher priority compared to mere speed. To minimize possible learning effects that may be attributable to a specific path through the labyrinth, 12 different variations of the course, depicted in Figure 2, were presented: The original configuration as shown in Figure 2 (1), path rotated by 90 • , 180 • , 270 • (2-4), horizontally reflected original configuration (5), rotated by 90 • and reflected horizontally (6), and starting point and goal point interchanged (7)(8)(9)(10)(11)(12). All paths were counter-balanced and presented in a random order.…”

“…The range of loss rates investigated in the present study is close to the values found in wireless communication channels to space or underwater. Humans are, in general, capable of adapting their control strategy to a technical system to maintain a constant level of task performance [38]- [40], though this does not necessarily hold for time-delayed sensory feedback [10], [41]. The degraded task performance observed here may well be attributable to the change in environmental impedance discussed in Section IV-B1: The time delay-dependent inertia that must be moved by the human operator in addition to the inertia and damping from the position-based admittance controlled haptic interface requires more effort during acceleration and deceleration.…”

“…Instead, we must consider both packet loss rate and time delay affecting visual feedback. Time delay in visual feedback has been found to consistently deteriorate task performance [7], [10], [41]. Corruption of sensory feedback due to packet dropouts is manifested in a stagnation of visual (perceived) motion, followed by a position jump and a variation in haptic impedance during loss bursts.…”

“…The time delay was regulated at a rate of 25 Hz in 5 equal steps between 0 and 200 ms. We chose a wider range of time delays here since small values of T d did not produce a significant effect on human task performance in Experiment 1. The weighting factors λ 1 and λ 2 in the MPC algorithm resulting from equation (10) were set to weight collision probability higher than network costs. This means that the maximum latency was applied in noncritical situations, but the time delay was lowered in case of an imminent collision, where lower latencies reduce collision probability p col ([0, t p ]).…”

“…It has been shown that an increase in the packet loss rate reduces the accuracy of temporal order judgments between visual and haptic events [9]. Movement time can also be impacted by packet loss that is caused by time-varying time delays [10]. In this paper, we investigate the relation between task performance, communication delay, and information loss in teleoperation systems with haptic feedback in greater depth.…”