Robotics has been moving into new areas outside the industry in recent years. One such area is robotic food preparation, which many researchers are working on. This thesis focuses on robotic food preparation, specifically a robotic cell for pancake making. The robotic cell consists of a manual baking device, a UR5e robot and a robotic gripper. The main objective is to robotise the manual rolling of pancakes. In doing so, we want to transfer the movements performed manually to the robotic manipulator using robot learning by demonstration approach. The manual movements of the spatula performed by a human while rolling the pancakes are first captured by an optical measurement system. The numerical data is then analysed in the Matlab programming environment. The trajectory of the spatula is segmented and the translational and rotational motions are observed as the arm moves. The result of the statistical analysis is the trajectory of the spatula represented in a form suitable for transfer to a robotic manipulator. Finally, we compare the trajectory executed by the robot with the movements executed by hand and observe the similarity between the executed movements.
|
