Part representation isn't always the easiest automation problem to solve, but it can be one of the most important when determining the level of productivity you can add to a production line. An efficient parts presentation platform can provide enough inventory to run an entire night shift without operator intervention and can eliminate unnecessary process building steps that slow down cycle times. Not all parts and machining operations are good candidates for manufacturing continuity, so understanding the hurdles and design opportunities your challenge presents will help you develop the best method to meet your robotic manufacturing needs.
Presentation of parts for robotic systems
For some machining operations, automatic feed lines do all the work. Rolled steel can be flattened for stamping presses and long bars can be advanced through the back of a lathe chuck. Other operations require each part to be represented and oriented individually, such as blanks for a press brake or a dovetail blank for a CNC machine.
In the last two cases, the process consists of two steps: presentation and alignment. Presentation is the step at which parts are made available for capture by the system. Alignment is the step at which inconsistencies in the presentation method are eliminated. A well-designed fixture performs both steps at the same time, if the geometry of the parts allows it.
Automated detail presentation also presents a number of challenges. How to make feed continuous? How does the automated system find the next part? What if the feed system fails and the task continues without parts? Let's take a look at them below.
Continuous feed
The Continuous Feed Part Representation Method is a method where the next workpiece is automatically positioned at a predetermined position to be picked up by an industrial robot arm. The best way to think of this fixture is as a slide where rods of a given length can be rolled to the feed tray. The operator loads the rods and gravity feeds them to the bottom of the rail where a stop holds them in place. The robot arm only reaches for the part that is adjacent to the fence, and under the influence of gravity, the next part moves to its place.
In this scenario, the next part to be gripped is always present in the same place, and the fixture can be continuously replenished from above or designed to stay as long as needed for the desired production time.
Part Location
True container picking solutions are prohibitively expensive and can be difficult to program. If it is possible to represent the array of parts of your system in a grid or stack, you can simplify programming the robot and adapting to different parts of the feeder.
Both grid mode and stack mode have their advantages. Stacking uses force to find the top of the stack and pulls out items/blanks until there are none left in the stack and the arm reaches the bottom. Many programmable instruments can use the Save Position block to program the base system to remember where the top of the stack is when it retrieves pieces, reducing the time it takes to find each piece. Stacking is best for flat parts. Feed/select on grid works well when parts cannot be stacked exactly on top of each other or are too large for the stack. Grid selection software tools typically use a Grid block to create an array of evenly spaced waypoints. Each time this execution block is executed, the system will move the robot's "arm" to the next part of the grid.
Align workpieces
When the operator loads the machine, he can check that the part is oriented correctly and that it fits snugly against the stops or chuck. For example, in the modern Task Canvas programming interface for any automated lines that runs under Forge/OS, the feed system can use force motions to check the correct placement of parts, but this is provided that they have been correctly selected. If some of the elements in the presentation are slightly misaligned, this discrepancy may be reflected in the placement step.
Gravity vibrating chutes are an effective way to ensure that all parts are captured in succession.
