Richard Scalzo's research concerns statistical inference over probabilistic models with deterministic constraints, for example in physical systems.
His signature achievement at ANU was the invention of a method to accurately infer the mass of ejecta in a type Ia supernova from multi-band photometry, and its use to convincingly demonstrate a range of masses among the progenitors of type Ia supernovae.
Since 2022 Dr Scalzo has been a member of Data61's Computational Modelling Group at CSIRO, and is presently a co-leader for the AI4Design porfolio of projects. AI4Design is an integrated template for embedding artificial intelligence into the design process for the geometry of 3-D printable industrial components, seeking to dramatically lower costs to test and scale up new designs and to accelerate the transition to a sustainable economy. His current research interests include physics-informed machine learning and hierarchical Bayesian methods for calibration, model selection, and model mis-specification in physical systems.