LAUSR

Abstract Educational and research manufacturing systems, such as learning factories, provide an environment to learn, test and implement new product and business solutions, research ideas and system paradigms. When learning factories become physically changeable, they are called Changeable Learning Factories (CLF), and can be used for investigating and teaching the effects of change of product design and production planning on manufacturing system layout and control. However, changeability requires a high level of system granularity and complexity, accompanied to a tendency to prevent students and trainees from developing deeper understanding of the underlying technology, or being able to change the physical system components on the machine level, especially for turnkey solutions. This paper introduces a model that selects the best system design from a pool of learning factory configuration alternatives, such as different types of material handling systems, individualized vs. clustered components, number of material routes and decision-making nodes, etc. This paper uses a selection model based on system structure complexity that changes with system granularity level. Results show that highly granular modular learning factories, and their complete opposite, low granular integrated learning factories have higher complexity than middle level granular learning factories, that are operationally changeable while being simple to understand and physically being able to change on the machine level.