UNIVERSITY OF SOUTHERN CALIFORNIA (USC), GE RESEARCH (GER), UNIVERSITY OF PITTSBURGH (PITT), AND UNIVERSITY OF VIRGINIA (UVA) WILL COLLABORATE TO DEVELOP AND DEMONSTRATE ENHANCEMENT OF THERMAL BARRIER COATING (TBC) DURABILITY AND SPRAY PROCESS FOR HYDROGEN (H2) GAS TURBINE (GT) COMBUSTION WITH ADVANCED PROBABILISTIC MACHINE LEARNING (ML) METHODS - H2THERMAL. THE GOAL IS TO IMPROVE COMPONENT LIFE DUE TO SPALLATION AND H2 IMPACTS ON MATERIALS UNDER HARSH SERVICE CONDITIONS ON LARGE LAND-BASED GT COMPONENTS AT VIABLE COST. THE PROPOSED APPROACH UTILIZES PROBABILISTIC ML METHODS TO ACCELERATE TBC PROCESS OPTIMIZATION, DISCOVER NOVEL MATERIAL RESPONSE, AND REDUCE PRODUCTION COST. THE PROPOSED TBC APPLICATION PROCESS WILL BE DEMONSTRATED ON A REPRESENTATIVE COMBUSTOR LINER. THE ROBUST AND ADAPTABLE PROCESS WILL BE CAPABLE OF CREATING TAILORED COLUMNAR MICROSTRUCTURES. APPLICATIONS INCLUDE LAND-BASED GTS, ADVANCED NUCLEAR REACTORS, MOLTEN SALT HEAT EXCHANGERS, AND OTHERS.