Collaborative robots may be required to do complex assembly operations during production and programming the Robot to perform such operations can be a challenging task. If the motions of the human palm could be mimicked on the Robot, it would mitigate the com-plexities involved in teaching such skills to the Robot. This mimicking action could also be used in other applications like robot based painting where it is necessary to capture motion generated by a skilled human hand and reproduce it on a robot.

To achieve this there may be many methods. A point cloud camera could be used to capture and store point cloud data of an object at different positions and orientations inside the Robot work volume. Corresponding Tool centre point positions and orientations read from the Robot controller serve as labelled data necessary for machine learning.

The task involves generation of Robot programs to synchronize point cloud capture and robot pose capture resulting in generation of data necessary for machine learning. The point cloud data is generated using Kinect point cloud camera by establishing the communication between the robot and the Kinect V2.

The objective in this paper is concluded by gathering the point cloud data for limited positions using Kinect V2 which is necessary for machine learning. Since machine learning techniques can be applied for further process huge amount of data is required hence this data is generated.