Machine Learning in Metals Additive Manufacturing

Abstract: Recent evidence of distributed additive manufacturing networks revolutionizing supply chains to meet drastic, sudden shifts in demand for life-saving medical equipment in times of a global pandemic makes evident that additive manufacturing technologies are at a tipping point of enabling local, on-demand, fit-for-purpose manufacturing that transforms societies. Machine learning is also transforming societies, but it is yet to be widely adopted in metals additive manufacturing technologies. This seminar will introduce pitfalls, successes, and open challenges in application of machine learning to innovate metals additive manufacturing technologies using case studies from my career. Pitfalls will be discussed through an example of my first attempts to use machine learning to create a statistical qualification means for companies. Success will be demonstrated through more recent results using machine learning together with pre-existing data to statistically inform the onboarding of a new laser powder bed fusion machine with a more powerful laser than the previous state of the art. These case studies will motivate a summary of current recommended best practices for deciding when, how, and why to use machine learning in metals additive manufacturing, or more broadly any materials or manufacturing problem, including how to verify that the answer is reliable. In conclusion, open challenges will be framed through the lens of a vision for establishing the Advanced Manufacturing Pilot Facility at Georgia Tech as an epicenter for future decades of research and innovation of artificially intelligent metals manufacturing systems. 

Bio: Prof. Stebner works at the intersection of manufacturing, machine learning, materials, and mechanics. Prof. Stebner joined the Georgia Tech faculty as an Associate Professor of Mechanical Engineering and Materials Science and Engineering in 2020. Previously, he was the Rowlinson Associate Professor of Mechanical Engineering and Materials Science at the Colorado School of Mines (2013 – 2020), a postdoctoral scholar at the Graduate Aerospace Laboratories of the California Institute of Technology (2012 – 2013), a Lecturer in the Segal Design Institute at Northwestern University (2009 – 2012), a Research Scientist at Telezygology Inc. establishing manufacturing and “internet of things” technologies for shape memory alloy-secured latching devices (2008-2009), a Research Fellow at the NASA Glenn Research Center developing smart materials technologies for morphing aircraft structures (2006 – 2008), and a Mechanical Engineer at the Electric Device Corporation in Canfield, OH developing manufacturing and automation technologies for the circuit breaker industry (1995 – 2000). 

He has won numerous awards, including an NSF-Career award (2014), the Colorado School of Mines Researcher of the Year Award (2017), and a Visiting Professor Fellowship from the Japan Society for the Preservation of Science (JSPS, 2019). 

Stebner serves as a board member of the ASM International Organization on Shape Memory and Superelastic Technologies (SMST) and an international advisory committee member of the International Conference on Martensitic Transformations (ICOMAT). Stebner is an Associate Editor for the journal Additive Manufacturing.