The ground-based remote sensing of clouds serves to investigate cloud formation and improve precipitation prediction, which is very important for the performance forecast of renewable energies. So far, Germany has a strong focus on the precipitation radars of the German Weather Service (DWD), which are used for the observation and short-term forecast of precipitation. Only a few years ago are networks of advanced remote sensing methods under construction that allow far more diverse atmospheric measurements (CLOUDNET). This results in two main objectives of the project: On the one hand, it will be analysed how artificial neural networks can be optimised for the performance forecast of photovoltaic and wind turbines when additional data from weather stations and ground-based remote sensing measurements are implemented. On the other hand, derivative algorithms are to be further developed for cloud observations in order to improve the determination of cloud properties. The focus here is on the development of application-specific artificial neural networks for the characterisation of liquid water in mixed-phase clouds.