Thanks to advancements in automotive technology and the development of smart infras- tructures, the advent of autonomous vehicles (AVs) is expected in the coming decades. This thesis investigates the feasibility of introducing Level 4 Autonomous Vehicles (L4- AVs) on Italian roads, focusing particularly on the takeover maneuver that occurs when an L4 vehicle exits its Operational Design Domain (ODD) and the driver is required to regain control. This work is part of the HL4IT project (Interaction of Humans with Level 4 Automated Vehicles in an Italian Environment), a collaboration between the Mechan- ical Engineering Department at Politecnico di Milano, the Bio-Engineering Department, and San Raffaele University, which innovative considers both the psychological and phys- iological aspects of the driver to evaluate human reactions. To create a comprehensive database of information on human/AV interactions, experimental tests were conducted using the DriSMi dynamic driving simulator at Politecnico di Milano. Three typical Ital- ian scenarios were designed, each including an element outside the ODD: a city-center environment with a roundabout to be navigated in the opposite direction due to ongoing roadworks, a highway viaduct subject to strong lateral wind gusts at the exit of a tunnel and a suburban road with sudden dense fog. In the initial part of each scenario, the vehicle travels autonomously, but when encountering the element outside the ODD, a takeover request is issued, requiring the driver to manually control the vehicle. Participants were equipped with an electrocardiogram, electroencephalogram, eye-tracker glasses, and skin conductance sensors to evaluate their stress levels during the takeover event. Preliminary results indicate that drivers experienced increased cognitive load and stress levels during the takeover request but no critical situations arose.
Grazie ai progressi nella tecnologia automobilistica e allo sviluppo di infrastrutture intel- ligenti, ci si aspetta che l’avvento dei veicoli autonomi (AV) avvenga nei prossimi decenni. Questa tesi indaga sulla fattibilità di introdurre Veicoli Autonomi di Livello 4 (L4-AV) sulle strade italiane, concentrandosi in particolare sulla manovra di takeover che si verifica quando un veicolo L4 esce dal suo Operational Design Domain (ODD) e il conducente è tenuto a riprendere il controllo. Questo lavoro fa parte del progetto HL4IT (Interaction of Humans with Level 4 Automated Vehicles in an Italian Environment, Interazione degli Umani con Veicoli Autonomi di Livello 4 in Ambiente Italiano), una collaborazione tra il Dipartimento di Ingegneria Meccanica del Politecnico di Milano, il Dipartimento di Bio-ingegneria e l’Università San Raffaele, che considera in modo innovativo sia gli as- petti psicologici che fisiologici del conducente per valutare le reazioni umane. Per creare un database completo di informazioni sull’interazione umano/AV, sono stati condotti test sperimentali utilizzando il simulatore di guida dinamico DriSMi presso il Politecnico di Mi- lano. Sono stati progettati tre scenari tipici italiani, ognuno dei quali include un elemento al di fuori dell’ODD: un ambiente urbano con una rotatoria da percorrere in direzione op- posta a causa di lavori stradali in corso, un viadotto autostradale soggetto a forti raffiche di vento laterali all’uscita di un tunnel e una strada suburbana con improvvisa e densa nebbia. Nella parte iniziale di ciascuno scenario, il veicolo viaggia autonomamente, ma quando incontra l’elemento al di fuori dell’ODD, viene emessa una richiesta di takeover, che richiede al conducente di controllare manualmente il veicolo. I partecipanti sono stati dotati di elettrocardiogramma, elettroencefalogramma, occhiali eye-tracker e sensori di conduttanza cutanea per valutare i loro livelli di stress durante l’evento di takeover. I risultati preliminari indicano che i conducenti hanno sperimentato un aumento del carico cognitivo e dei livelli di stress durante la richiesta di takeover ma non sono sorte situazioni critiche.
Investigation of human response to the takeover request using a dynamic driving simulator
CATTANEO, PAOLO
2023/2024
Abstract
Thanks to advancements in automotive technology and the development of smart infras- tructures, the advent of autonomous vehicles (AVs) is expected in the coming decades. This thesis investigates the feasibility of introducing Level 4 Autonomous Vehicles (L4- AVs) on Italian roads, focusing particularly on the takeover maneuver that occurs when an L4 vehicle exits its Operational Design Domain (ODD) and the driver is required to regain control. This work is part of the HL4IT project (Interaction of Humans with Level 4 Automated Vehicles in an Italian Environment), a collaboration between the Mechan- ical Engineering Department at Politecnico di Milano, the Bio-Engineering Department, and San Raffaele University, which innovative considers both the psychological and phys- iological aspects of the driver to evaluate human reactions. To create a comprehensive database of information on human/AV interactions, experimental tests were conducted using the DriSMi dynamic driving simulator at Politecnico di Milano. Three typical Ital- ian scenarios were designed, each including an element outside the ODD: a city-center environment with a roundabout to be navigated in the opposite direction due to ongoing roadworks, a highway viaduct subject to strong lateral wind gusts at the exit of a tunnel and a suburban road with sudden dense fog. In the initial part of each scenario, the vehicle travels autonomously, but when encountering the element outside the ODD, a takeover request is issued, requiring the driver to manually control the vehicle. Participants were equipped with an electrocardiogram, electroencephalogram, eye-tracker glasses, and skin conductance sensors to evaluate their stress levels during the takeover event. Preliminary results indicate that drivers experienced increased cognitive load and stress levels during the takeover request but no critical situations arose.File | Dimensione | Formato | |
---|---|---|---|
2024_07_Cattaneo.pdf
non accessibile
Descrizione: Investigation of Human Response to the Takeover Request using a Dynamic Driving Simulator - Tesi di Laurea Magistrale in Mechanical Engineering - Ground Vehicles
Dimensione
7.14 MB
Formato
Adobe PDF
|
7.14 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/222626