In the final test campaign of the Prepare Ships project the predictor system will be tested onboard Stena Germanica. Initially, the predictor system will predict the future motions of the Stena Germanica based on a generic dynamic model of the ship. Afterwards, once the predictor system has had time to capture some data, it will improve itself through “training”. Will the training succeed and improve the predictions? If so, by how much?
We wish work package 3 good luck!