Robotic System for Depalletization of Pharmaceutical Products with 3D Camera

Several manual labor intensive activities are occurred in pharmaceutical products management in hospitals. Organizing pharmaceutical products is complicated and results  numerous serious problems, particularly medication error in patients. The management starts from checking products from suppliers, classifying, and storing them to proper conditions. Improving accuracy and quality of the process, robotics and automation are applied to the operations. The robotic system for depalletization of pharmaceutical products has been developed. This research robotic system is used to refill medicine boxes to specific magazines that will be used for further automated dispensing.  In that case, a SCARA robot and vacuum gripper is used to pick and place boxes. Convolutional Neural Network (CNN) has been applied to locate and classify the boxes. Previously, a 2D camera is used for the detection but the limitation is occurred with respect to the location accuracy. In this research, the system is developed with a 3D camera to improve the localization accuracy. The experiments were done to compare the performance by testing with boxes with various appearance.