Predictive 3D Printing with IoT

Jing-Sheng Jeannette Song,Yue Zhang
DOI: https://doi.org/10.2139/ssrn.3895854
2021-01-01
SSRN Electronic Journal
Abstract:Industry 4.0 marks a new industrial revolution centered around technologies that connect digital and physical realms, revamping how information is used to manage work. Two core enablers are the Internet of Things (IoT) and 3D printing. The former monitors the real-time status of a complex system, and the latter responds to that information with agile manufacturing. Nevertheless, how exactly this can be done remains unclear. To gain insights, we consider the context of a 3D printer supplying a critical part installed in multiple machines embedded with sensors and interconnected via IoT. While it is tempting to perceive that the marriage of 3D printing and IoT would make on-demand printing a reality, our results indicate that this is not necessarily so. The true benefit of the marriage is to enable predictive printing. In particular, it is generally optimal for the 3D printer to predictively print-to-stock, triggered by a system-lifetime-status dependent threshold. Whether the optimal policy is to print-on-advance-demand that permits minimum inventory depends crucially on the printing speed. Moreover, we quantify IoT’s value in cost improvement by separately assessing the impact of advance information from embedded sensors and that of IoT’s real-time information fusion. Through a numerical study, we find that IoT brings significant value in general. However, the common belief that IoT’s value should scale up for larger systems can backfire if 3D printing capacity is not boosted comparably. Our framework can be leveraged to aid investment decisions regarding IoT/embedded sensors, and help develop scheduling tools for predictive 3D printing.
What problem does this paper attempt to address?