Abstract:
Additive Manufacturing (AM) has experienced rapid development in recent years. As the files 3D Printing requires are in digital formats (STL, CLI, G-code, etc.), location-independent printing services have emerged on the market and become increasingly popular. As a simplified form of cloud manufacturing, functionalities of these existing systems are limited to service retrieving and ordering. To explore the real potential of integration between Cloud Manufacturing and AM, a Cloud-based AM Platform is proposed to support customers throughout their product development processes. In this research, three modules in the platform are the major focuses – cyber-physical 3D printer, knowledge management system and service selection decision support system. Some original contributions have been made. At a high level, a new architecture of a cloud plat-form for AM is designed and a new definition of AM services is created. Several sub-systems are integrated to provide customers with support from design to manufacturing. Firstly, the cyber-physical 3D printer, enabled by Internet of Things, is the key to connecting 3D printers to the cloud, and realizing remote control, monitoring and self-diagnosis. An information model based on OPC-UA is developed and an Artificial Neural Network (ANN) is proposed to detect surface defect. Secondly, a Bayesian Network is proposed to model the domain knowledge under uncertainty. A systematic approach is developed from parameter learning, conflict detection, to bi-directional inference. The structure of knowledge model is proposed and two sub-models have been developed. Thirdly, a hybrid Multi-Criteria Decision Making approach is presented to help users select the most suitable printing strategies. There are two important modules - option navigation and ranking. The Design by Shop-ping approach and Analytic Hierarchy Process and Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) method are integrated to provide an informative decision support system. Due to the special requirements of the dynamic option space, the TOPSIS is modified to improve its stability and accuracy in the proposed system. Finally, the implementation of a prototype system and case study is given to show the functionalities of the platform at different product development stages.