Mobile heavy machinery such as mobile cranes, excavators, and wheel loaders often operates in hazardous environments and under harsh conditions. This situation put the well-being of the operator and the machine's condition at risk. At the same time, assisted systems and teleoperation possibilities are currently underdeveloped. In our research, we want to leverage state-of-the-art machine learning methods to automate everyday machinery tasks. The focus is on delivering simulation tools and benchmark environments for open research. Moreover, reinforcement learning and virtual sensing techniques are examined in demanding earth-moving operations. The big goal is to provide technology for the remote operation of a fleet of autonomous machinery.