Reducing jerk for bimanual control with virtual reality

Hill, SCM

Abstract

This work explores interaction in virtual environments where bimanual operations lacking kinematic constraints, such as a pair of tracked controllers held in free space, are mapped to a target object that obeys a constraint, such as a pair of grippers grasping a single object.

Visual and haptic feedback systems are developed and compared as solutions for minimising Jerk during motion. The outcome of the study is that visual feedback is shown to be superior to haptic feedback for use in free-air bimanual systems where maximising smoothness of motion is the objective.

The primary application is for improved teleoperation of a new generation of ordnance disposal robots which utilise dual grippers intended for bimanual operation. The work is undertaken in a generalised manner however so that it may be applicable for other uses such as medical systems or nuclear disassembly.

The systems developed herein initially utilise a custom framework designed for a fourteen screen, fully tracked immersive environment comparable to an enhanced CAVE, known locally as Octave. The final study subsequently utilises the Unreal Engine 4 framework for the HTC Vive HMD system. All software developed is provided as a core part of this work.

Citation

Hill, S. (in press). Reducing jerk for bimanual control with virtual reality. (Thesis). University of Salford

• Thesis Type: Thesis
• Acceptance Date: Apr 30, 2019
• Deposit Date: May 7, 2019
• Publicly Available Date: Jun 7, 2019
• Related Public URLs: http://www.sm-robotics.com/phd.html
• Award Date: Jan 1, 2019