The availability of a wide variety of labelled data (e.g. safety related) is major bottleneck impeding the accuracy of AI models developed for ATM. One possible solution is to generate synthetic data from original data and a model that is trained to reproduce the characteristics and structure of the original data. The project explored and defined AI-based methods to generate synthetic data. These methods are attractive since they require less user knowledge expertise and better generalisation capabilities. The project took advantage of advances in computer vision and language technology to develop a universal time series generator (UTG). The generator was able to train on a certain set of data obtained from a small number of airports in order to generate synthetic data about a new airport.