Intelligent grasping is required by autonomous robots to lift objects with unknown weight. This requires the gripper robot to have enough grasping force so that the object does not slip out of the gripper, but not too much force, for otherwise the object may be damaged. The development of an optical tactile sensor can be an approach to intelligent grasping for a robotic gripper. Initial results collected showed that using an infrared sensor in a deforming material will provide slip detection for the gripper robot. Combining this tactile sensor with the gripper robot will create a feedback system that can detect slip and adjust gripping strength for intelligent grasping. The low cost and the small size of the tactile sensor offers a cost efficient and adaptable tactile sensor design for an autonomous small gripper robot used to pick and place packages.
The aim of this thesis is to improve an existing design and incorporate elements from other successful tactile sensor designs. Various issues are present with the proposed design which are solved. The infrared sensor will be affected by ambient lighting which can be controlled by placing the sensor in a cavity which no external light sources can affect. There is also the issue of choosing a deformable material which is not too soft otherwise it will always deform or too stiff otherwise it will not deform when the object is slipping. The system must also have a fast response time to detect slip in under one second as well as when slippage occurs at 1mm or less. This may be solved by using a mouse sensor which has a high response time in microseconds and can detect changes in position under 1 mm.