Publications

M. Singhala,
Understanding Human Haptic Perception in Telerobotic Systems (2022),
[Doctoral Dissertation, Johns Hopkins University]. (Abstract,links)

Abstract

Human-in-the-Loop Telerobotic Systems (HiLTS) allow us to overcome the constraints of distance, speed, scale, and safety, when our natural limbs alone are insufficient. Using HiLTS, however, often carries with it the loss of sensory-rich haptic information that our natural limbs rely on when exploring our environment directly. This dissertation aims to improve our understanding of the perceptual tradeoffs of replacing direct manipulation with telerobotic manipulation. I systematically studied how sensory information from the environment is transformed into haptic percepts by the human operator as this information flows through the human-robot-environment system. I begin with a focus on the human operator. Using psychophysical methods, our team investigated how the operator’s perception of the remote/virtual environment is shaped by the physical manner in which they explored the environment through the telerobot. First, we focused on the dynamics of environment exploration, to see if changing the velocity at which the virtual/remote environment is explored affects percept development. We then investigated whether the hand used to control the telerobot affects perception of the virtual/remote environment (i.e., perceptual asymmetries). In both studies, we demonstrated that the manner in which the operator interfaces with the robot can have a significant effect on their perception of environment properties, such as stiffness. I then focus on the telerobot. Using a visuo-haptic object-tracking task, our team investigated how the impact of a telerobot’s transmission dynamics on task performance. Given the significant research focus on attenuating or circumventing these dynamics, I sought to understand the trade-off between telerobot transparency and telerobot performance. To accomplish this, I developed a novel telerobot testbed that allows for one-to-one comparisons between various mechanical and electromechanical leader-follower transmissions. We found that participants were able to accommodate substantial differences in telerobot transmission dynamics, and produced the compensatory strategies necessary to adequately perform the object tracking task. This dissertation ends with a focus on the environment and the stimuli associated with it. Using a novel haptic illusion paradigm, we investigated the assumption that underlies most modern telerobotic control schemes — that there is a bijection between sensation and perception. To accomplish this, I developed a novel experimental apparatus and demonstrated that different stimuli can lead to the same haptic percept, and that the same stimuli can lead to different haptic percepts. Through this study, we bring attention to the human intelligence in the loop of telerobotic systems, and the need to account for it in our control schemes.

Links

http://jhir.library.jhu.edu/handle/1774.2/67513