Day 4: As Far As The Eye Can See- Choosing The Right Vision System For Your Collaborative Robot
As many production managers are increasingly looking to automate mundane and repetitive tasks, what’s often overlooked is the human skill and dexterity involved in carrying out even the simplest of tasks. Picking up parts from a conveyor and placing them in a bin requires motor skills and visual processing, inherent human skills we tend to take for granted.
Automating these types of processes can be complex for a robot particularly when it involves variable part positioning. Depending on where you want the robot to pick parts from, your vision system will need to be able to determine the parts position and orientation in some or all six axes. In determining whether a 2D vision system or a 3D system is required will be dependent on multiple factors including part type, presentation, orientation, acquisition and space constraints etc. In applications where shape information is critical to performing a task, where a vision system will need to determine position and orientation on all 6 axes then a 3D system will be required. If you want to pick a part from a flat surface like a table or a conveyor belt and put it in a bin then the vision system will only need to determine part position and orientation along two of its axis as the z-axis is fixed so a 2D vision system is probably sufficient here. PickIt’s series of off-the-shelf guided vision systems offer a robust and versatile solution for pick and place applications handling parts of different shapes, materials and sizes. A great feature of the PickIt system is its ability to operate even in changing and poor light conditions. Unlike some 2D picking systems it does not need special lighting and is not affected by reflections. 2D vision systems are more commonly used than 3D vision systems primarily due to the fact they are more economical and less complicated to implement. Ultimately though, it will come down to the capability of the system in carrying out the required task and the type of parts involved. Robotiq’s wrist camera is a 2D vision system specifically designed to perform industrial applications with Universal Robots. One of the best features of this compact camera is it’s easy to use and set up and the cameras software is seamlessly embedded in the UR’s graphical user interface It can locate many parts in a single snapshot, comes with customisable colour validation and a simple teaching method to ensure repeatable picking. It also has barcode and 2D reading function allowing you to add traceability to your production line. Whatever guided vision system you choose, it’s success will be dependent on all parts of the system working together including the system itself and the data that provides in relation to backlighting, lens selection, filtering etc and the other variables of the cell including the gripper, product variation, orientation, lighting etc.