Recent advances in technology have led to the development of human-machine systems where the machine is able and expected to make and execute decisions on its own.  These decisions are based on its understanding of the situation, its system status and capabilities. The machine part of such human-machine systems can therefore act at least partially as autonomous agents. Examples of such systems are already existing advanced driver assistance and automation systems.  Even more sophisticated driving automation systems for highly and fully automated driving are currently in the focus of research activities in the automotive domain. Whereas such systems offer great possibilities for supporting the driver in terms of safety, efficiency and comfort, this potential can only be exploited if the machine acts as a supportive and cooperative partner of the human driver. In order to achieve these goals, several challenges have to be met in the driver-vehicle interaction design. One main challenge is the creation of observable and comprehensible behaviour of the assistance and automation system enabling the human and the machine to generate a shared situation representation. This talk will present some recent research results on aspects of human-machine cooperation based on examples from the automotive domain. These will include effects of levels of automation on drivers’ ability to regain control when the system reaches its system limits, the evaluation of a driver-vehicle interaction strategy for supporting the driver’s understanding of the system status in system limit situations and the evaluation of an interaction strategy to support the cooperation between driver-vehicle systems.