The integration of Digital Twin technology represents a pivotal shift in the maritime industry, offering unprecedented opportunities for ship lifecycle management and operational optimization. The GAIA-TWIN project aims to develop a digital twin (DT) of the CNR R/V Gaia Blu through the use of advanced data-driven, data-lean, and equation-free methodologies. The activities focused on creating a predictive model for the nowcasting of ship's attitude and dynamics (heave, heave velocity, vertical acceleration, roll, pitch, and yaw angles and velocity) using Hankel dynamic mode decomposition (HDMD). Prediction reliability and decision support were enhanced by using a uncertainty aware formulation extension. The model is fed with real-time data and adapts to the vessel’s evolving physical state during real-world operations. The first two phases of the GAIA-TWIN campaign were carried out aboard the R/V Gaia Blu from June 7 to 10 and from October 24 to November 8, 2025, sailing through the Mediterranean Sea. Synchronized time-series data from the onboard systems were continuously collected during navigation and will be shared with the maritime research community to foster research and innovation. In particular, data were recorded for ship positioning, attitude, velocities and accelerations, anemometry, meteo, and thermosalinometry. The data collection implementation and the results obtained for the real-time nowcasting algorithm will be presented and discussed. Future efforts will be dedicated to acquiring the wave field surrounding the ship, extending the DT capabilities, and focusing on system identification and manoeuvrability, key to ensuring vessel safety under severe sea states. Acknowledgements: The authors wish to thank the CNR for the ship time granted through the Gaia Twin project. This research was funded by the Italian Ministry of University and Research through the National Recovery and Resilience Plan (PNRR), CN00000023--CUP B43C22000440001, “Sustainable Mobility Center” (CNMS), Spoke 3 “Waterways”. The authors are also grateful to the US Office of Naval Research for its support in the methodological development through NICOP Grants N62909-21-1-2042 and N62909-24-1-2102.