Model-based automatic generation of digital twin models for the simulation of reconfigurable manufacturing systems for timber construction

Abstract

Reconfigurable manufacturing systems adapt to changing requirements and therefore offer a solution to the challenges associated with increasing product variety and shorter product life cycles. This is especially helpful in the prefabrication in timber construction, as components and requirements vary widely between projects. However, reconfigurations must be quick to keep downtimes short. The planning of reconfigurations is a complex problem where digital twins and simulation models of the manufacturing systems are necessary for simulation and validation. In our timber prefabrication use case, building components and requirements change from project to project, and the manufacturing system needs to reconfigure frequently. The building components and the configuration of the manufacturing system are designed simultaneously during a co-design process to ensure the integrity of the manufacturing process in a changing manufacturing environment. Hence, this may require many iterations of the planning process. Each iteration needs validation through a simulation. However, the manual generation of simulation models is time-consuming and error-prone and thus poses an obstacle for the digital planning phase. This paper addresses this issue and presents a model-based approach to generate simulation models for reconfigurable manufacturing systems. Our system and its components are modeled in automation markup language as this domain-specific language allows us to model not only kinematics and geometry but also behavior, skills, and interfaces to check the compatibility of components. We apply the method to our use case with the IntCDC wood prefabrication system. The proposed method builds a component graph for a configuration of the manufacturing system and generates the required packages for the simulation in Gazebo and ROS. To build simulation models automatically, we apply model-to-model and model-to-text transformation techniques. The proposed method is suitable for integration in the digital co-design workflow of IntCDC and allows fast iterations with continuous simulations of machine configurations throughout the planning process.