Julien

• Architected and developed a multi-agent simulation platform on ROS2, enabling scalable testing of autonomous agent behaviors for AI research initiatives (github.com/naval-group/lotusim). • Designed a tactical mission decision-support UI, integrating and enhancing existing mathematical models to improve clarity for end users. • Developed AI-driven prediction algorithms with custom UI design for Navy vessels, translating complex model outputs into actionable insights. • Coordinated multiple AI navigation research projects with NTU's CARTIN Lab, aligning research goals with engineering execution across cross-functional teams.

Developing a new hybrid method to couple physics-based model reductions and data-based Koopman method (with neural network dictionaries) to manage simulating complex multi-parametrized dynamical systems.

Research on fundations of thermodynamics and irreversibility using a Bayesian formalism. Study of fluctuation relations, with comparison with computational mechanics formalism and application on Langevin equations.