Development of an Intelligent Robotic Manipulator

Abstract

The presence of hazards to human health in chemical process plant and nuclear waste stores leads to the use of robots and more specifically manipulators in unmanned spaces. Rapid and accurate performance of robotic arm movement and positioning, coupled with a reliable manipulator gripping mechanism for variable orientation and a range of deformable and/or geometric and coloured products, will lead to smarter/intelligent operation of high precision equipment. The aim of the research is to design a more effective robot arm manipulator for use in a glovebox environment utilising control kinematics together with image processing / object recognition algorithms and in particular the work is aimed at improving the movement of the robot arm in the case of unresolved kinematics, seeking improved speed and performance of object recognition along with improved sensitivity of the manipulator gripper mechanism
A virtual robot arm and associated workspace was designed within the LabView 2009 environment and prototype gripper arms were designed and analysed within the Solidworks 2009 environment. Visual information was acquired by barrel cameras. Field research determines the location of identically shaped objects, and the object recognition algorithms establish the difference between them. A touch/feel device installed within the gripper arm housing ensures that the applied force is adequate to securely grasp the object without damage, but also to adapt to any slippage whilst the manipulator moves within the robot workspace.
The research demonstrates that complex operations can be achieved without the expense of specialised parts/components; and that implementation of control algorithms can compensate for any ambiguous signals or fault conditions that occur through the operation of the manipulator. The results show that system performance is determined by the trade-off between speed and accuracy. The designed system can be further utilised for control of multi-functional robots connected within a production line. The Graphic User Interface illustrated within the thesis can be customised by the supervisor to suit operational needs.