Robots have transformed many areas of manufacturing, to the point that whole industries –from automotive to electronics and chemical – would now be unthinkable without them. But other sectors are far more resistant to automation, and yet could equally benefit from it.
The textile industry is a good example. Textile and clothing are essential pillar in the European economy, involving around 160,000 companies employing over 1.5 million workers, with a turnover of 162 billion euros. Here, manufacturing processes are mostly manual, and in an effort to contain costs, businesses often end up outsourcing them to low-cost labor countries. As a result, Europe exports € 61bn worth of textile products, but imports € 109 bn – a trend that automation could help reverse.
For the European clothing sector, innovating through automation could be a way to increase competitiveness and reduce outsourcing, but there are huge scientific and technical challenges to overcome for automating the manipulation of textile materials. That is the reason why MERGING, a European research project on soft objects robotic manipulation coordinated by CEA and involving 11 other partners from industry and academia, has chosen the textile industry as one of its key use cases. Within MERGING, researchers and engineers from several institutions are working with SELMARK, a leading manufacturer of lingerie based in Vigo (Spain), to develop a robotic system that can support human operators in the most repetitive and tiring tasks that happen before the stitching of the final product: in particular de-stacking textiles from storage, grasping, placing and unfolding them to avoid wrinkling, measuring parts for intermediate quality control before next steps of the process. These operations are impossible for traditional robotic manipulators designed for rigid objects, as they cannot easily adapt in real time to the unpredictable behavior of fabrics. Total or partial automation of these tasks can promote the allocation of personnel to tasks that have more added value, while reducing the global production costs.
The MERGING solution is to combine various building blocks. At the basis, there are two collaborative robot arms that can safely work close to humans. The arms are equipped with a dedicated gripper, enhanced with EPFL’s electroadhesive skin, a technology that uses electric fields to make objects stick to the jaws. Stereo cameras provide perception and high precision tracking of textiles developed by AIMEN, that -together with computer models of fabric’s behaviour– allow to adapt the robot’s movements in real time. The robot arms can be easily programmed by unexperienced users, thanks e.g. to robot programming by demonstration or other control features developed by CEA. Thus, the operator can show the robot how to perform the required tasks, rather than programming every movement on a computer. All data are combined into LMS’s “digital twin”, a digital representation of the working environment that allows to simulate process before transmitting orders to the combined industrial work cell supervision system.
When all these technologies are combined, the robotic grippers can grasp a ply from the textile materials stack, carefully detaching it from the ones below. The ply is then placed on the thermoforming device, and if any wrinkle is detected by the perception system, the robot will use corrective actions to undo it. The robot then feeds the fabric to presses that give the material the desired 3D shape (e.g. a cup, in the case of lingerie production), and places it on a stack of modified parts ready for the next step of the process.
MERGING started in late 2019 and is now halfway through its work programme, perfecting the system’s building blocks. In early 2022 it will start working on the integration phase before setting up the pilot demonstration at SELMARK premises, which will be overseen by AIMEN covered by IPC.
The textile industry is only one of the three use cases explored by this project, that will also apply similar robotic technologies to food packaging and to the manipulation of fabrics in the automotive industry.