Development of an assistant system for regulating the mental strain state of agricultural machinery operators

The increasing automation of agricultural machinery leads to fluctuating phases of mental underload and overload for operators, impairing performance, safety, and well-being. This dissertation evaluates an adaptive assistance system that monitors and regulates mental strain in real time. Using a combine harvester, multimodal sensor data trained a model to classify mental states. A virtual assistant provides task recommendations. Studies show improved strain regulation and practical applicability.