The article is published in the original.
The paper proposes flexible system that is based on Robot Operating System framework for integration of 3D computer vision and artificial intelligence algorithms with industrial robots for automation of industrial tasks. The system provides flexibility of 3D computer vision hardware and industrial robot components, allowing to test different hardware with small software changes. The experimental system consisting of Kinect V2 RGB+Depth camera and Universal Robots UR5 robot was set up. In experimental setup the pick and place task was implemented where randomly organized two types of objects (tubes and cans) where picked from the container and sorted in two separate containers. Average full cycle time for the task was measured to be 19.675 s.