P2P Cloud Manufacturing Based on a Customized Business Model: An Exploratory Study

Area of Science:

• Manufacturing Engineering
• Computer Science
• Materials Science

Background:

• Traditional manufacturing processes face challenges with long production cycles and high costs.
• Personalized custom business models require efficient methods for bespoke product creation.
• Integrating advanced technologies is crucial for modernizing manufacturing.

Purpose of the Study:

• To propose a novel platform-to-platform (P2P) cloud manufacturing method.
• To enable object-to-object fabrication from 2D images using deep learning and additive manufacturing.
• To reduce production cycle times and manufacturing costs for personalized products.

Main Methods:

• Development of an object detection extractor using the YOLOv4 algorithm.
• Construction of a 3D data generator leveraging DVR technology for 2D-to-3D conversion.
• Integration of deep learning and additive manufacturing (AM) within a P2P cloud framework.
• Case study involving online sofa and car photos for a 3D printing service scenario.

Main Results:

• Achieved object recognition rates of 59% for sofas and 100% for cars.
• Demonstrated a 2D-to-3D data conversion time of approximately 60 seconds.
• Successfully manufactured three unindividualized and one individualized 3D printed model, maintaining original shape.
• Validated the proposed method for personalized 3D model transformation and fabrication.

Conclusions:

• The P2P cloud manufacturing method effectively bridges the gap between digital concepts and physical products.
• The integration of YOLOv4 and DVR technology facilitates rapid 3D model generation from images.
• The approach supports personalized customization, offering a viable solution for efficient, cost-effective manufacturing.