Develop an autonomous product-based reconfigurable manufacturing system

Abstract

With the ever-emerging market including mass customization and product variety, reconfigurable manufacturing systems (RMS) have been presented as the solution. A manufacturing system that combines the benefits of the two classic manufacturing systems to increase responsiveness and reduce production time and costs. To cope with the lack of physical systems, an RMS system have been built at UiT Narvik. Today, both reconfiguration and deciding layout must be executed manually by a human. A task that is both incredibly time consuming and far from optimal. A method of automating the layout generation and thus the manufacturing system is presented in this thesis. To the author’s knowledge such experiment has not been performed previously. Layouts is generated with a NSGA-II algorithm in Python by minimizing objectives from a developed mathematical model. The results have been tested with a MiR-100 mobile robot placing five modules in two different layouts. The results have been compared with a digital visualization for validation. In addition to the visualization, videos of the physical system's automated layout generation are presented. The results concludes that the method both generates feasible layouts as well as enhancing the automation of the system.