Dottorato in MORFOGENESI E INGEGNERIA TISSUTALE

Titolo della tesi: Neurophysiological models for the training of Artificial Intelligence in automotive safety systems

The objective of this Ph.D. research was to conduct a multimodal assessment of mental fatigue onset in professional drivers, adopting an Electroencephalography-based index as an objective benchmark to recognize early signs of fatigue. My research addressed a key gap in the existing literature by comparing mental fatigue in both simulated and realistic driving environments. While much of the current research on driving fatigue focuses on severe or prolonged states of fatigue, this work concentrated on the onset of fatigue, a crucial yet understudied area. Understanding fatigue onset is essential for timely interventions, which can prevent the negative consequences associated with advanced fatigue, especially in high-risk tasks like driving. Data was collected through both simulated and real driving tasks, designed to be identical in terms of protocol and driving path. The results indicate that these tasks induced comparable levels of mental fatigue in both environments, allowing for a meaningful comparison. Regarding neurophysiological responses to fatigue, my research found that heart rate variability-related parameters, low and high frequency, were consistently influenced by fatigue across both simulated and realistic scenarios. In contrast, ocular-related parameters—such as blink rate, blink duration, and blink amplitude—showed notable differences between the two settings. This divergence in responses suggests that the two environments may engage drivers differently at the onset of fatigue, highlighting the need for further investigation to understand these differences. Additionally, my research provides valuable insights into the early detection of fatigue, which can be applied to improve driver monitoring systems. By focusing on fatigue onset, the findings from this study go beyond the state of art, and they can contribute to the development of more responsive systems that can intervene earlier and potentially reduce the risks associated with driver fatigue. Overall, this research not only fills critical gaps in the understanding of fatigue in both simulated and real driving conditions but also offers practical implications for enhancing safety in the automotive field through improved monitoring technologies.