Digital-twin-assisted multi-stage machining of thin-wall structures using interchangeable robotic and human-assisted automation

• A smart, interconnected multi-stage machining ecosystem is realized through digital-twin-assisted information flow. • Versatility of the factory environment is achieved via interchangeable robotic automation and human-cyber-physical systems. • The lightweight digital twin allows for fast deformation predictions in different scenarios, namely in machining-based and clamping-based deformation estimation and control. • Robotic and human-cyber-physical clamping control showcase versatile intelligent technologies in smart machining ecosystems while relying on the same digital twin. • Digital-twin-assisted superposed CNC trajectory paths correct surface profile errors and improve part quality in different factory settings. Interconnected intelligent systems in multi-stage smart machining environments are an advancing area of research, demonstrating many real-life opportunities that can benefit from the development and integration of cyber-physical systems into machining habitats, while different automation levels in industrial manufacturing sites call for flexibility of core strategies towards smart machining ecosystems. This article introduces a versatile and smart multi-stage machining environment for the controlled clamping and machining of low-rigidity structures in an interconnected cyber-physical factory. This is exemplified by a deformation-prone thin-wall workpiece, which undergoes controlled clamping, enabled by interchangeable robotic automation and automation via human-cyber-physical systems, as well as digital-twin-assisted corrective machining enabled by the swift estimation of workpiece deformations and multi-stage communication between machining habitats. The underlying digital twin presents a fast, lightweight simulation approach, based on a mass-spring-lattice model, allowing information flow from and to systems, which is utilized by the CNC machine as well as the interchangeable robot- and human-in-the-loop clamping enablers. By employing this controlled clamping approach workpiece deformations are aimed to be minimized. At the same time, a desired total clamping force is achieved in order to perform subsequent digital-twin-assisted machining corrections to reduce deformation-caused flatness errors. Ultimately, this article presents an intelligent multi-stage machining scenario where digital-twin enabled information moves along with thin-wall structures and branches out for knowledge-based control and corrections to robots, humans and CNC machines respectively, showcasing a real-life example for versatile, information-driven smart machining ecosystems.