ERI: An Artificial Intelligence-based Computer Aided Manufacturing Framework for Hybrid Manufacturing

This Engineering Research Initiation (ERI) grant supports research that contributes new knowledge in manufacturing process planning automation and promotes the progress of fundamental science in the fields of advanced manufacturing, computer science, mathematical modeling, and geometric reasoning. Hybrid manufacturing integrates different manufacturing processes in one system, enabling the creation of a ready-to-use functional part directly from raw or stock material.

Hybridizing two advanced manufacturing processes, additive and subtractive manufacturing, potentially unleashes nearly full manufacturing capability by providing the freedom of adding and removing material in three-dimensional space. This permits the realization of part designs of complex shapes and functionality for a variety of applications. However, this extraordinary manufacturing capability also introduces unprecedented challenges in toolpath planning and motion control, impeding the broader application of hybrid manufacturing.

This award supports fundamental research to explore and develop artificial intelligence (AI)-based methods to facilitate smarter and better computer aided manufacturing (CAM) tools for hybrid manufacturing processes. The project advances the understanding of automated manufacturing toolpath planning and control and enables goal-oriented autonomous fabrication of parts of any geometry.

This research advances digital manufacturing, enhances sustainability, and trains the future skilled workforce, which benefits the U.S. economy and society. The project benefits several industries such as aerospace, defense, healthcare, energy, agriculture, and others.

This research lays out a new fully automated computer-aided manufacturing (CAM) framework for advanced High-Degree-of-Freedom (i.e., 5 or more axes operation) hybrid manufacturing processes. This framework leverages state-of-the-art artificial intelligence (AI) algorithms for computer-aided design (CAD) geometry analysis and CAM toolpath planning and control.

A generalized model for various manufacturing processes and the AI approach that provides the best solution is the thrust of this research. The data format of the model allows inherent support for AI methodology. New AI algorithms that are built on neural networks, evolutionary algorithms, and reinforcement learning are investigated for automated toolpath planning.

This work advances the knowledge base in advanced manufacturing by filling the knowledge gap on how human knowledge and production data can be harnessed and extended to realize new manufacturing capabilities. The research team plans to hybridize a 5-axis milling subtractive process and a 5-axis material extrusion/directed energy deposition type additive manufacturing process, explore the AI-CAM framework’s capability to expand from one to two and then to multiple processes, and establish a standard training and testing methodology for AI-CAM for further expansion and generalization of the framework.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.