Effective methods for human-robot-environment interaction by means of haptic robotics

Publication Type:
Thesis
Issue Date:
2012
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail01Front.pdf10.85 MB
Adobe PDF
Thumbnail02Whole.pdf100.26 MB
Adobe PDF
Industrial robots have been widely used to perform well-defined repetitive tasks in carefully constructed simple environments such as manufacturing factories. The futuristic vision of industrial robots is to operate in complex, unstructured and unknown (or partially known) environments, to assist human workers in undertaking hazardous tasks such as sandblasting in steel bridge maintenance. Autonomous operation of industrial robots in such environments is ideal, but semi-autonomous or manual operation with human interaction is a practical solution because it utilises human intelligence and experience combined with the power and accuracy of an industrial robot. To achieve the human interaction operation, there are several challenges that need to be addressed: environmental awareness, effective robot-environment interaction and human-robot interaction. This thesis aims to develop methodologies that enable natural and efficient Human- Robot-Environment Interaction (HREI) and apply them in a steel bridge maintenance robotic system. Three research issues are addressed: Robot-Environment-Interaction (REI), haptic device and robot interface and intuitive human-robot interaction. To enable efficient robot-environment interaction, a potential field-based Virtual Force Field (VF2) approach has been investigated. The VF2 approach includes an Attractive Force (AF) method and a force control algorithm for robot motion control, and a 3D Virtual Force Field (3D-VF2) method for real-time collision avoidance. Results obtained from simulation, experiments in a laboratory setup and field test have verified and validated these methods. A haptic device-robot interface has been developed for providing intuitive human-robot interaction. Haptic devices are normally small compared to industrial robots. Thus, the workspace of a haptic device is much smaller than the workspace of a big industrial manipulator. A novel workspace mapping method, which includes drifting control, scaling control and edge motion control, has been investigated for mapping a small haptic workspace to the large workspace of manipulator with the aim of providing natural kinesthetic feedback to an operator and smooth control of robot operation. A haptic force control approach has also been studied for transferring the virtual contact force (between the robot and the environment) and the inertia of the manipulator to the operator's hand through a force feedback function. Human factors have significant effect on the performance of haptic-based human-robot interaction. An eXtended Hand Movement (XHM) model for eye-guided hand movement has been investigated in this thesis with the aim of providing natural and comfortable interaction between a human operator and a robot, and improving the operational performance. The model has been studied for increasing the speed of the manipulator while maintaining the control accuracy. This model is applied into a robotic system and it has been verified by various experiments. These theoretical methods and algorithms have been successfully implemented in a steel bridge maintenance robotic system, and tested in both laboratory and a bridge maintenance site located in Sydney.
Please use this identifier to cite or link to this item: