Nuclear fusion holds major potential to supply the world with safe, clean and limitless energy. At the same time, a fusion device offers an entire range of opportunities in the area of data science. In this project, I intend to leverage our research group's expertise in integrated data analysis using Bayesian inference for real-time sensor fusion in a new generation of demonstration fusion power plants (DEMO). This comes with a number of challenges due to real-time requirements, uncertainty quantification at both the side of the diagnostics (sensors) and actuators, and integration into the control system of the device. Integrated data analysis will be crucial in DEMO due to the necessity of stable operation in the presence of limited quantities of sensor data. The project will break ground in bringing this expertise for the first time to the design activities for DEMO. At the same time, I intend to exploit our research group's unique position in this area to reach out to fusion labs and Flemish industry for transfer of know-how and to stimulate industry to contribute to the foundations of the Belgian fusion industry by means of advanced data science applications.