E-Mobility Engineering 016 l Aurora Powertrains eSled dossier l In Conversation: Thomas de Lange l Automated manufacturing focus l Torque sensing insight l Battery Show Europe 2022 report l Sodium batteries insight l User interfaces focus

Focus | Automated manufacturing Acknowledgements The author would like to thank Stephen Bennington at Q5D and Benjamin Troskie at Cotes for their help with researching this article. line that consists of fewer than 50 steps, which is about 70% less than conventional assembly lines. It uses only one standard body frame, down from more than 80 sub-assemblies, with a wiring harness that is half the mass of that in average vehicles, and a fraction of the number of controllers, connectors and CPUs. Digital twins and automation software Creating a digital twin of the manufacturing process is becoming a key step in the automation of the assembly of the battery cell, pack and vehicle. Combining the physical models used to design the cells and packs with the automation tools allows different processes to be tested out virtually rather than in a pilot factory. Taking data from sensors across the factory floor builds up the digital twin, allowing more optimisation of the process. It can also highlight where the physical machines are falling behind the predicted performance, indicating that problems might be looming. This allows preventive maintenance to be scheduled in before an uncontrolled equipment failure. The digital twin works with automation software such as NeuroCAD, which analyses component properties with the help of machine learning methods and uses them to determine the extent to which a component is suitable for assembly automation. NeuroCAD also evaluates the gripping surfaces of the component, as well as how well it can be aligned. In addition, the neural network will calculate the degree of probability of the result being correct. Another modular software system, called PiTaSC, is used to program the robots to automate processes such as clipping, riveting or screwing, which were previously carried out manually. It used to be necessary to largely reprogram a robot system for each application but now, once tasks have been modelled, they can be quickly transferred to new product variants, products and even robots made by other manufacturers. The software is modular, enabling assembly processes to be programmed with prefabricated and reusable modules that can be individually assembled when setting up a robot system. This allows assembly applications such as snapping on or pushing components into each other to be automated. It can even extend to the specification of the vehicle itself, feeding back market data and customer orders to the automated assembly of the battery pack. In another machine learning development, a project in Germany has been working on a module called a configuration generator, which generates fully specified vehicle configurations based on sales, development and market data. A software prototype has been created that allows the generated planned orders to be optimally dispatched to production, and the corresponding material and capacity requirements to be determined. In the project’s final phase, an algorithm will be developed that assigns real customer and dealer orders to the planned orders. Conclusion Battery cells and packs are key new areas for the assembly of electric vehicles. While assembly processes have been optimised over many years for existing vehicles that use combustion engines, there are challenges for automating the assembly of battery cells, packs and wiring harnesses for the production of EVs. At the same time, manufacturers are taking advantage of more data to create digital twins to monitor and optimise the automated assembly line. A prototype AM automated wiring assembly system (Courtesy of Q5D) 42 Winter 2022 | E-Mobility Engineering