Predictive cost comparison of manufacturing technologies through analyzing generic features in part screening

The selection and interaction of different manufacturing technologies are major challenges in product development and production processes. Costs are one criterion for selecting the most suitable manufacturing technology for a component. Therefore, they must be considered for an automated technology selection in part screening approaches. These costs arise during the component manufacture, during the transport of the components to the production site, and within the factory's assembly line. However, holistic cost data is not available at an early stage during product development. Hence, these information gaps must be filled to ensure the correct selection of a component's manufacturing technology. This paper first analyzes the cost data of past product developments regarding potential differences between manufacturing technologies. Abstracted geometric part information and production-related aspects are derived and predicted to serve as input features for the subsequent cost prediction. These AI-based cost prediction models are trained for relevant cost types along the part's lifecycle. The most capable prediction models are combined into a manufacturing technology-dependent cost analysis of the components. Based on that, the forecasted costs of different manufacturing scenarios represent one input for the overarching part screening methodology to select the most suitable manufacturing technology for each component.