Bin Picking allows automation of previously manual and monotonous tasks in the picking process.
Automating picking tasks with bin picking
Bin Picking allows automation of previously manual and monotonous tasks in the picking process. In e-commerce, where staff have to deal with a high volume of orders, outsourcing these activities to robots can be a major relief. This helps keep staff from being tied to one specific station. In turn, employees can take on more demanding tasks, such as operating the application, monitoring multiple stations, and troubleshooting.
Advantages of bin picking with SIMATIC Robot Pick AI
Bin picking is versatile and especially useful in areas where order picking has seen little automation to date, or where it requires a large number of workers, such as intralogistics, warehousing, e-commerce and online grocery retail. The common challenge in these areas is the great variety of products for which no information, such as CAD models, is available. The SIMATIC Robot Pick AI bin picking application is equipped with artificial intelligence to handle the task of gripping and moving objects. The system allows the robot to grab objects from their bin and transfer them to a target bin for ordering.
Advantages of the application example
The application example shows how to set up a bin picking system based on SIMATIC Robot Pick AI. A 3D camera system captures the position of the objects in the box and delivers the required data to the SIMATIC Robot Pick AI software. After calculating the pick points, the 6-axis robot moves to the points and grips the objects with a vacuum gripper.
Users benefit from prefab programs for the PLC and the robot. The programs have already implemented the machine’s sequential control and the communication with the SIMATIC Robot Pick AI software and the robot. With a few adjustments to the mechanical configuration, the system is functional and can be used for testing purposes in a laboratory setup.
In order to capture the position of the objects being gripped, the scene is first recorded by a 3D camera system. The camera system takes a color image and a depth image. The camera is controlled via the USB connection by the SIMATIC Robot Pick AI application, which is run as a Docker container on the IPC127E. The captured image data passes through the neural network in the backend container which, once successfully processed, outputs a suitable pick point.
SIMATIC Robot Pick AI also provides a web interface that allows the user to view the current state of the system as well as the results of the calculation.
The command to calculate the pick point is transmitted from the PLC over the network to the Pick AI backend on the IPC. The PLC and the backend communicate via the local network using the block “LGetPickPoint_HTTP” from the LGetPickPoint library. The central sequential control of the system is located in the “MainSequence” PLC block. In addition, the PLC communicates with the UR robot controller via a PROFINET IO connection.
The “URRobotCommunication” block is used for easy control of the robot without additional software. The block provides the user with a uniform interface for accessing the functions programmed on the robot controller. With the help of the graphical user interface on the UR touch panel, the user can operate the robot during commissioning and teach the static positions necessary for the process. Then, the process can be started and stopped with the panel.