April 2021

30 n MACHINE VISION April 2021 www.drivesncontrols.com Vision systems take on a PCB assembly challenge A ssembly of printed circuit boards (PCBs) is already largely automated, using technologies such as SMD placement, soldering and automated optical inspection (AOI). Until now, however, an exception has been the through-hole mounting of wired components such as capacitors, inductors and connectors. This is a complex process that cannot be easily automated and is still done mostly by hand. This was the challenge faced by engineers from the German machine-builder Glaub Automation & Engineering. “A client asked us whether we could find a solution for automating this step,” says CEO, Niko Glaub. The engineers set to work and came up with a design – which they call the GL- THTeasy robot cell – in which the component assembly is done by an ABB YuMi dual-arm collaborative robot (cobot). Because of its two arms, it can assemble PCBs twice as fast as a single-arm robot. The first step in the automated process is for a conveyor to supply the robot with blister packs containing capacitors, for example. The pack is identified by a data matrix code. The robot picks one capacitor after another from the pack and places it precisely on the PCB. Alternatively, it can pick the components from an ESD container or a vibratory conveyor. The next stage is to solder the components from below. Once empty, the blister packs are removed via a recirculation system and replaced automatically by full ones. The reason why this apparently simple process has been difficult to automate previously is that robots struggled with either the high variability in the component feeding, or slight inaccuracies in positioning of the components. Previous attempts to solve these problems were extremely complex, awkward to program, and not particularly reliable. Glaub’s new robot cell turns to smart cameras for the first time. The position of the components in the blister packs is detected using 3D surface sensors, which also allow targeted picking from a bin or vibratory conveyor. The capacitors and PCBs are then monitored using 2D cameras. Glaub's engineers worked with a German vision specialist M-Vis Solutions to specify the cameras. Several cameras capture the matrix codes on the blister packs and measure and locate each component. “With the 100% absolute measurement of components and PCBs, GL-THTeasy compensates for every inaccuracy in terms of components, gripping, workpiece carriers and conveyor belts,” explains M-Vis’ CEO, Vitali Burghardt. Components that do not fit precisely are rejected. As part of a feasibility study, M-Vis chose eight cameras – four to monitor each of the robot’s arms. A Cognex In-Sight 7802M vision system measures the parts and provides information to correct the position of the gripper. An In-Sight 9912M system measures the circuit board and if necessary helps to correct the gripper's movement as it places the component on the board. A 3D surface- scan camera (a Cognex 3D-A5060) with a patent-pending 3D technology and integrated image-processing software “sees” the position of parts in the feed line. “In each process step, the cameras capture the actual position of the component, the gripper, and the circuit board in relation to the electronic component,” Niko Glaub explains. “The ‘legroom’ of the components is aligned with the actual dimensions of the assembly positions. This allows the component to be found and removed automatically, and then enables through- hole mounting on the basis of actual position data.” Because the movements are controlled based on cameras, operators can generate new placement models using the images, without needing to do any programming. This simplifies and accelerates not only assembly, but also conversion. The YuMi's cobot’s arms allow round-the- clock operation at high speeds with cycle times of less than three seconds, depending on the components to be installed and the feed. The payback period works out at around 14 months. Glaub believes that its technology will persuade many electronics manufacturers to automate this aspect of PCB assembly. n A German machine-builder has developed a robot cell that automates the previously difficult task of through-hole mounting of PCB components. Several machine vision systems, combined with a dual- arm cobot, are key to the development. Barcode readers capture data matrix codes from trays of components Glaub’s machine uses an array of vision systems to automate the difficult process of inserting through- hole components into PCBs