Stress and mental fatigue are recognised as significant challenges in industrial working environments, affecting workers’ well-being, cognitive performance, and operational safety. Prolonged exposure to high cognitive workloads may impair several physiological systems and increase the likelihood of human error. In this context, the widespread adoption of wearable technologies has enabled the development of monitoring systems based on the continuous, non-invasive acquisition of psychophysiological signals. However, the automatic assessment of stress and mental fatigue remains a complex challenge due to the inter-individual variability of physiological responses. This thesis proposes a multimodal system for monitoring stress and mental fatigue in warehouse operators, based on the integration of physiological and behavioural signals acquired through a sensorised wearable helmet.