34 n ROBOTICS AND AUTOMATION July/August 2024 www.drivesncontrols.com Reflecting on the challenges of protecting satellites The French robotics specialist Akeoplus has developed a system – for the aerospace manufacturer, Thales Alenia Space – that automates the placement and bonding of optical solar re ectors (OSRs) onto protective panels for satellites. The aim was to use robots, operating in cleanroom conditions, to place the OSRs onto panels measuring up to 5m x 3m, with an accuracy of better than 0.2mm. A major challenge was that the ratio of the positioning accuracy to the work area is 1:40,000. The panels are used to protect satellites from radiation. They act as outer walls tted with up to 7,000 re ectors, each measuring around 40mm x 40mm. Until recently, OSRs were mounted by hand by two or more people lying on mats to glue and position the OSRs using their outstretched arms. “The production of large panels has increased, and working such surfaces was not ideal for the employees for ergonomic reasons,” explains Maxime Motisi, Akeoplus’ COO, who managed the project. “The robotic cell has been designed to speed up the process and thus increase productivity. Our customer can now manufacture much faster, even though we have implemented additional steps for traceability and quality assurance.” The cell contains two robots as well as a positioning table on which the panels are placed. To ensure accuracy, a laser tracker is used with a six-axis robot arm, as well as three high-resolution 2D cameras. A small robot rst takes an OSR from a tray and places it onto a vacuum backlight under the rst camera, which takes a picture. MVTec’s Halcon machine vision software inspects this image for any breaks or scratches. A “shape matching” technique determines the exact position of the OSR and checks whether its size and shape correspond to the model in the design plan. It is important to determine the position of every re ector accurately because the robot grips each one slightly dierently. It moves it to the next production step with an accuracy close to 0.2mm. If the component is OK, it is placed under a glue dispenser nozzle. After the glue has been applied, the robot picks up the OSR again and places it where a second robot with a larger working envelope can grab it for further processing. This Kuka robot moves the part in front of a second image-processing station, where a camera captures images of the applied glue. These are analysed to check that the glue is evenly distributed and that it is in a precise pattern specied by Thales’ engineers. The OSR is then aligned precisely with the robot arm as a rst step to placing it in the right location on the satellite panel. A third high-res camera now takes images of the OSR, which is aligned using algorithms that determine a 3D transformation between the coordinates of the OSR corners and the coordinate system of the vacuum gripper that is holding the OSR. The corners of the OSR are determined by the intersection of its edges. From this, a coordinate system can be dened where the OSR’s corner is centred and aligned with its edges. The coordinate system of the robot eector is determined by hand-eye calibration of the robot. Determining the coordinate systems is important for two reasons. First, because the re ector may be tilted, its corners are compared with the eector, to act as reference point that can compensate for tiny tilts or shifts. And second, the corner of the OSR serves as a reference point for placing it on the panel. Because the 2.4m-long robot arm is not 100% accurate, Akeoplus has developed an iterative servoing process that allows the robot to mount the OSR precisely on the 18m2 panel. The cell is already in use. Motisi admits that the project was challenging – especially working on the glueing system. “We can now be very satised with the result, because the requirements in terms of quality, productivity and employee satisfaction were all achieved,” he reports. The cycle time for gluing and positioning OSRs is now less than 45s, with a OEE of better than 95%. A large panel can be covered with solar re ectors in ve days by one employee. Previously, it took around two people seven days to nish a panel. In addition, the work environment has been improved by eliminating the long hours of working lying down, and reducing contact with the glue. n A French robotics company has developed a robotic cell that places thousands of tiny re ectors with an accuracy of 0.2mm onto large panels that are used to protect satellites from solar radiation. The challenging project uses two robots, a laser tracker and three high-res cameras. One of the cell’s two robots places the tiny re ectors precisely on the large protective panel for satellites.
RkJQdWJsaXNoZXIy MjQ0NzM=