Highly automated and autonomous agricultural machines have the potential to perform field work with a minimum of energy and manpower. For the application case of tillage, sowing and weed control, important tool-specific challenges are the precise and variable depth control of the tools, the identification of the tool conditions in the field and, in the application case of an automated tool change, systems for identifying the tools and the working equipment. To meet all these requirements, the machines need a smart tool control systems. These functionalities are integrated into the highly flexible and modular machine concept-Feldschwarm¹. The depth control is able to guide the tool at a defined working depth, thus guaranteeing the desired results and requiring a minimum of fuel consumption. The tool recognition works without contact. The tools have a memory device with a special set of parameters, for example the identification of the tool, the range of optimal working depth and, in case of an active tool, the nominal speed range. A camera system is used to identify the tool conditions. It detects wear, misalignment and loss of tools. The wear values are used to adjust the depth control and stored on the tool memory device.