China
Africa
Explore chapters and articles related to this topic
Driver State Monitoring for Decreased Fitness to Drive
Published in Donald L. Fisher, William J. Horrey, John D. Lee, Michael A. Regan, Handbook of Human Factors for Automated, Connected, and Intelligent Vehicles, 2020
Michael G. Lenné, Trey Roady, Jonny Kuo
Related to PERCLOS, blink duration is another metric that has received considerable attention in drowsy driving research. In a study of real-world driving, Hallvig, Anund, Fors, Kecklund, and Akerstedt (2014) reported intra-individual mean blink duration to be a significant predictor of unintentional lane departures. The importance of warning latency cannot be overlooked in an operational real-world system and, in comparison to existing implementations of PERCLOS, measures of blink duration are generally able to perform closer to real time. An extension to this concept is the idea of a pre-drive assessment of fitness to drive. Using a camera-based driver monitoring system, Mulhall et al. (2018) demonstrated the significant predictive ability of mean blink duration measured before a drive on subsequent, real-world lane departure events.
Driver response and recovery following automation initiated disengagement in real-world hands-free driving
Published in Traffic Injury Prevention, 2023
Pnina Gershon, Bruce Mehler, Bryan Reimer
Concerns over driver inattention, coupled with known limitations of partial automation, has led research to focus on time-critical, system-initiated disengagements where drivers are required to take over either the lateral (steering) or both lateral and longitudinal (steering and speed) vehicle control (Louw et al. 2017; Gaspar and Carney 2019). Essentially, drivers may be unprepared to regain control when automation reaches its operational limits (Lin et al. 2018; Parasuraman and Riley 1997; Reagan et al. 2020). Furthermore, in these situations, it is also unknown how the use of systems like SC, that allows hands-free driving, will impact a driver’s ability to regain control in a timely manner. Driver attention monitoring systems are one mechanism designed to mitigate lapses in driver engagement by providing feedback to the driver or adapting the automation functionality in real-time (Donmez et al. 2009; Lee et al. 2013; Reimer 2020). Current driver monitoring systems use steering wheel torque-based sensors and/or driver facing cameras that track gaze and/or head position to infer driver state and intervene when a threshold for apparent inattention is exceeded. SC, for example, has a camera-based driver monitoring system that employs multimodal cues (visual, auditory, and haptics) to support driver attention on a moment-to-moment basis. While partial automation systems are increasingly available, the literature to date is limited by the lack of objective data on the extent to which drivers use partial automation, the context and frequencies in which the automation initiates disengagements, and how drivers respond to and recover from such events.