Machine learning can improve the important digital tools that are central to virtual analysis of software prior to deployment. Such software finds usage in everything from traffic control, over launch of nanosatellites to the healthcare sector. The tools are essential to computer modelling and can detect software errors in due time and hence reduce failure – ultimately save lives. In this research two tools (Uppaal and Tapaal) are optimized so they today stand more precise and faster than ever before due to a combination of machine learning and theory from computer games, altogether providing the software a better understanding of the surrounding context.
