Purpose and goal: There has been significant research and development for autonomous systems in recent years but many challenges remain. For example, the WASP NEST DISCOWER (https://discower.io) project is focused on autonomy for "weightless" systems both underwater and in space. Each of these domains has unique challenges including unmodeled disturbances and very high reliability requirements. In this context, VerDa explores both developing and verifying data driven methods for planning and control of weightless systems. One specific focus is methods for verifying neural network controlled systems also called neural feedback loops. Neural networks are highly complex and traditional verification methods are usually not applicable. Our current work involves computing reachable sets to ensure the safety of neural feedback loops. This requires solving very large optimization problems necessitating high performance computing resources. Expected results: We will have developed and tested algorithms that can be adapted to use both underwater and in space, and transferred insights between each of these two domains. One of the main results of VerDa in will be the development of verification algorithms for data-driven planning and control, with a focus on neural network controlled systems.