An autonomous framework for interpretation of 3D objects geometric data using 2D images for application in additive...

Area of Science:

• Manufacturing Engineering
• Computer Science
• Artificial Intelligence

Background:

• Industry 4.0 integrates additive manufacturing, artificial intelligence, and cloud computing.
• Intelligent cloud-based additive manufacturing is key to Industry 4.0, yet integration with service-oriented paradigms is limited.
• Existing frameworks lack autonomous platforms for cloud-based additive manufacturing service composition.

Purpose of the Study:

• To propose a framework for automated shape interpretation in additive manufacturing.
• To address the lack of research in accurate shape interpretation for autonomous manufacturing platforms.
• To enable cloud-based service composition for additive manufacturing based on customer demands.

Main Methods:

• Developed a framework for automatic shape interpretation using decomposition into simpler shapes.
• Proposed two algorithms: a Recurrent Neural Network for decomposition and a 2D Convolutional Neural Network for recognition.
• Integrated these algorithms into a platform and demonstrated capabilities through case studies.

Main Results:

• The proposed framework enables automated interpretation of product shapes from 2D images.
• The decomposition algorithm effectively breaks down complex objects into simpler, categorizable shapes.
• The integrated system shows capability in handling complex object shapes through decomposition.

Conclusions:

• Accurate and automated shape interpretation is vital for integrating intelligent additive manufacturing with service-oriented paradigms.
• The proposed framework and algorithms provide a foundation for autonomous cloud-based additive manufacturing platforms.
• This research advances the potential for sophisticated shape interpretation in additive manufacturing applications.