• eye-hand coordination
• gaze
• telemanipulation
• teleoperation
• virtual reality

Human gaze behaviour has been intensively studied and the importance of eye movements in eye-hand coordination is supported by many authors. Eye movements are related to the unfolding task and they play an anticipatory role, leading manipulative actions by an average time of about 500 ms. Thanks to the predictive nature of gaze signal, its use in the field of intention prediction has been studied. A particular application that requires intention estimation is the implementation of a shared autonomy module in robotic teleoperation. Since telemanipulation differs from natural manipulation, we argue that gaze strategy and eye-hand coordination may be affected. The goal of this thesis is to develop an experimental setup to investigate the similarities of eye behaviour between a teleoperated and a natural environment. To this aim, we adapted the protocol and the setup designed by Johansson R. et al. in the paper Eye-Hand Coordination in Object Manipulation, which represents a capstone in human behaviour.
Starting from an existing teleoperation platform comprising a robotic arm and an artificial hand controlled by an IMU-based suit, we integrated visual feedback by connecting an RGB-D camera to a head-mounted virtual reality viewer. The setup is composed of a manipulandum, different obstacles, a target to be reached by the participant, and sensors to allow automatic labelling of the different phases. The experimental setup was then validated through a pilot.