Design Simplification through AI-Enabled Part Clustering and Consolidation
Vinayak R Hebbar *
Department of Mechanical Engineering, RV College of Engineering, Bangalore, (An autonomous institution affiliated to Visvesvaraya Technological University, Belagavi, Karnataka – 59001), India.
Yashwanth R
Department of Mechanical Engineering, RV College of Engineering, Bangalore, (An autonomous institution affiliated to Visvesvaraya Technological University, Belagavi, Karnataka – 59001), India.
Prateek Ekbote
Department of Mechanical Engineering, RV College of Engineering, Bangalore, (An autonomous institution affiliated to Visvesvaraya Technological University, Belagavi, Karnataka – 59001), India.
Dhanush R T
Department of Mechanical Engineering, RV College of Engineering, Bangalore, (An autonomous institution affiliated to Visvesvaraya Technological University, Belagavi, Karnataka – 59001), India.
*Author to whom correspondence should be addressed.
Abstract
This paper presents a new approach to optimizing product design by using artificial intelligence techniques for part clustering and consolidation. Traditional design methods often lead to assemblies with many separate components. This results in more complex manufacturing, higher costs, and longer assembly times. To solve this, we propose a framework that uses unsupervised learning algorithms like K-means clustering to group similar components based on their function, material, and shape. We also explore the use of convolutional neural networks (CNNs) for recognizing patterns in 3D models to find opportunities for consolidation. A legacy mechanical component serves as a case study to demonstrate design simplification in a practical context. The findings show a significant reduction in part count, increased modularity, and improved manufacturability, confirming the effectiveness of AI-driven design simplification. This method offers a promising path for future engineering design practices, especially in situations where lightweighting, cost-effectiveness, and sustainability are crucial.
Keywords: Design for manufacturing and assembly (DFMA), part consolidation, K-means clustering, convolutional neural networks (CNN), artificial intelligence, product reengineering, modular design