Seminar: Finding New Ways to Team: Towards Smarter, More Collaborative Robotic Manufacturing
Seminar by Andrew Gillman
Research Materials Engineer
Air Force Research Laboratory
Many manufacturing and maintenance ecosystems have seen automation introduced to augment human tasks. However, many implementations are rigidly programmed to a particular task and thus lack functional agility and the ability to respond to large deviations from the planned objective. Recent advances in machine learning and artificial intelligence have shown promise in providing this agility, and our team is exploring how these algorithms can augment physics-based models and insight towards enabling more autonomy in manufacturing. In particular, additive manufacturing (AM) is an agile robotic manufacturing platform and an excellent representative of a high mix/low volume process that would benefit from increased adaptability and autonomy to meet a constantly changing set of part specifications. In our team’s recent exploration of Direct-Ink-Writing (DIW), an extrusion-based AM technique, we have explored dimensionality reduction and classification techniques for understanding the variations that appear in AM processes, and recent efforts have integrated these approaches into feedback control for more complex printing. Beyond our AM work, our recently established Digital Manufacturing Research Team is working more broadly toward “Teaching Tools to Be Teammates” through making a variety of robotic manufacturing processes more agile and autonomous while actively collaborating with human subject level experts.
Beyond a look into the background and vision of our newly formed research team, opportunities for collaboration will be discussed. Our team is working with multiple universities to partner with experts from a variety of relevant fields (signal processing, computer vision, robotics, machine learning, ontologies, data fusion, material/manufacturing science) as we solve these multi-disciplinary problems. At Ohio State, we have initiated collaboration with the Artificially Intelligent Manufacturing Systems Lab at the Center for Design and Manufacturing Excellence (CDME) and are looking to expand our collaborations to working with faculty members across the university and partnering directly with our team members located at OSU and CDME working on problems of shared interest.
Dr. Andrew Gillman is a Research Materials Engineer in the Air Force Research Laboratory’s (AFRL) Manufacturing and Industrial Technologies Division as a member of the Digital Manufacturing Research Team, where his research projects leverage physics-based modeling, data analytics, and machine learning for enhancing understanding of complex manufacturing processes and require development of frameworks for automated learning and collaboration between humans and manufacturing robotic systems. Prior to joining AFRL, Dr. Gillman was a postdoctoral Research Scientist in the Functional Materials Division, where he focused on integration of topology optimization and nonlinear origami mechanics for fold pattern discovery in multi-physics applications. Dr. Gillman received his PhD in Aerospace and Mechanical Engineering from the University of Notre Dame after receiving a BS in Mechanical Engineering from Purdue University, and his doctorate work included the development of image-based modeling strategies for prediction of effective thermos-mechanical properties of heterogeneous microstructures.