A common challenge when examining manufacturing or laboratory processes is successfully manipulating an object with a variable presentation, i.e. a different orientation, dimension, color, texture, etc., when an element of that process requires uniformity. For instance, consider a pick-and-place system that’s packing ceramic tiles. If the tiles (or any process with a comparatively “large” batch size) are exactly identical, the only variability in presentation may be their orientation on a conveyor. A simple mechanical feature could be used to align the tiles uniformly before presentation to whatever is being used to perform the pick-and-place, whether a SCARA robot, a traditional robot, or a cartesian system. If the tiles (or any process with a “small” or individual batch size) aren’t exactly identical, a machine vision system could determine the color, position offset, and rotation, then communicate that to a motion system which would compensate for the orientation during the pick to place the tile neatly with its same color. Along the same lines, the method used to mechanically accomplish the pick would need to accommodate the product variability; this could be done using anything from vacuum to force-monitored grippers depending on the target. In this article from MIT Technology Review, a research project to develop a more dexterous robot using machine learning is investigated. No matter the challenge with your process, Intellimech’s engineers have the knowledge and experience to help realize your goals and minimize risk. Connect with our team today to explore your particular objectives and how we can help create the application solution you’re seeking.
Automation. Control. Mechatronics.