Virtual reality aided vehicle teleoperation

Walter, Bryan
Major Professor
Committee Member
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Mechanical Engineering
Organizational Unit
Journal Issue
Mechanical Engineering

This thesis describes a novel approach to vehicle teleoperation. Vehicle teleoperation is the human mediated control of a vehicle from a remote location. Typical methods for providing updates of the world around the vehicle use vehicle mounted video cameras. This methodology suffers from two problems: lag and limited field of view. Lag is the amount of time it takes for a signal to travel from the operator's location to the vehicle. This lag causes the images from the camera and commands from the operator to be delayed. This behavior is a serious problem when the vehicle is approaching an obstacle. If the delay is long enough, the vehicle might crash into an obstacle before the operator knows that it is there. To complicate matters, most cameras provide only a small arc of visibility around the vehicle that leaves a significant blind spot. Therefore, hazards close to the vehicle might not be visible to the operator, such as a rock behind and to the left of the vehicle. In that case, if the vehicle were maneuvered sharply to the left, it might impact the rock. Virtual reality has been used to attack these two problems. A simulation of the vehicle is used to predict its positional response to inputs. This response is then displayed in a virtual world that mimics the operational environment. A dynamics algorithm called the wagon tongue method is used by a computer at the remote site to correct for inaccuracies between the simulated vehicle position and the actual vehicle position. The wagon tongue method eliminates the effect of the average lag value. Synchronization code is used to ensure that the vehicle executes commands with the same amount of time between them as when the operator issued them. This system behavior eliminates the effects of lag variation. The problem of limited field of view is solved by using a virtual camera viewpoint behind the vehicle that displays the entire world around the vehicle. This thesis develops and compares a system using virtual reality aided teleoperation with direct control and vehicle mounted camera aided teleoperation.