CAREER: Data-Driven Extrusion-Based Robotic Three-Dimensional Printing of Reinforced Concrete

About This Grant

This Faculty Early Career Development (CAREER) grant supports research in data-driven robotic three-dimensional concrete printing to create materially efficient reinforced concrete building elements, enabling the rapid construction of multistory buildings and addressing urgent housing shortages. Current attempts at such printing suffer from a limitation of the printer’s depositing material only in layers parallel to the ground, restricting its application to simple forms and making it incompatible with the production of materially efficient parts that entail complex geometries. This technological limitation is exacerbated by the need for constant human monitoring and tuning of the process to eliminate defects. As a result, current applications are limited to single-family houses with simple geometries that use more concrete, not less, than if manufactured using traditional formwork methods, thus limiting widespread adoption by the industry. This project looks to develop new printing methods that reliably deposit material in complex geometries inherent to materially efficient parts. This research intends to transition its methods into education to combat the significant decline in youth interested in construction careers. The project strives to prepare future generations through courses in robotic three-dimensional printing, data modeling, and machine learning. It will engage youth through interactive puzzles and digital robotic workshops It positions the U.S. as a leader in robotic construction technologies by fostering patents and startups that drive economic growth and innovation. This CAREER project supports research that aims to develop scientific principles for waste-free, extrusion-based robotic three-dimensional concrete printing, to account for the geometric complexity of components, and the rapid time-dependent evolution of material rheology. A significant challenge lies in bridging the knowledge gap that connects three key system attributes: complex part geometries, part performance while being printed, and robotic printing process parameters. To advance the state of knowledge in robotic concrete printing, this project intends to create a novel data-driven, multi-objective optimization model incorporating machine learning that can generate and predict (1) optimal non-planar slicing and (2) robotic printing instructions for optimal part performance during printing. To drive this model, the research strives to create (i) a new slicing algorithm that is robust and generalizable for complex parts; (ii) a series of empirical models derived from real-time process data that describe process-part interactions; (iii) a framework linking empirical and established models to develop performance indexes (e.g., buildability index) as objectives optimization; and (iv) a new data-collection framework to obtain the data needed for the modeling and learning, as well as for evaluation and verification. If successful, the project will develop new frameworks and models integrating geometry, materials science, and robotics to lay the foundation for advanced data-driven, large-scale additive manufacturing. 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.