کانال دانشکده مهندسی صنایع

#وبینار_علمی_امور_بین_الملل

🔷 عنوان سخنرانی:
Driving in the Era of Autonomous Vehicles

🔷 سخنران:
Dr. Sara Masoud (Assistant professor of industrial and systems engineering at Wayne State University)

🔷خلاصه سخنرانی:
As automated driving systems (ADS) evolve, understanding human drivers' behavior in mixed traffic environments and during control transitions is critical to ensuring safety and efficiency. This study combines insights from real-world driving data and simulated environments to explore two key questions:

1. Do human drivers change their behavior when interacting with autonomous vehicles (AVs)?
2. Can driving simulators help identify driver types for easier takeovers?
Using naturalistic driving data from the NGSIM and Lyft Level 5 datasets, we investigate car-following behavior and find that human drivers exhibit statistically significant changes when following AVs, characterized by reduced mean and variance in headway times. These behavioral shifts are influenced by factors such as road type, lane position, and time of day, highlighting the nuanced impacts of AVs on traffic dynamics.

Complementing this, data from the CARLA simulator was analyzed to identify and classify driver types based on behavioral features like gaze patterns, velocity changes, and gaze shift rates. K-means clustering revealed distinct driver typologies—normal, cautious, aggressive, and erratic—while deep learning models achieved high accuracy in classifying these patterns.

These findings support the development of adaptive automated systems that can anticipate and respond to driver behaviors, particularly during control transitions. Together, the results bridge the gap between real-world and simulated environments, providing a holistic understanding of human-AV interactions. By quantifying behavior changes and enabling tailored interventions, this work offers meaningful insights into the coexistence of human drivers and AVs, fostering safer and more efficient roadway systems.

📅 زمان:
دوشنبه ۲۶ آذر ماه ۱۴۰۳
⏱️ ساعت ۱۶:۳۰

💻 لینک برگزاری وبینار:
vc.sharif.edu/ie-seminar

#دانشکده_مهندسی_صنایع
#دانشگاه_صنعتی_شریف

🆔 @IE_SUT

www.tgoop.com/IE_SUT/1213

2.2K viewsedited  Dec 14 at 17:44

#وبینار_علمی_امور_بین_الملل

🔷 عنوان سخنرانی:
Driving in the Era of Autonomous Vehicles

🔷 سخنران:
Dr. Sara Masoud (Assistant professor of industrial and systems engineering at Wayne State University)

🔷خلاصه سخنرانی:
As automated driving systems (ADS) evolve, understanding human drivers' behavior in mixed traffic environments and during control transitions is critical to ensuring safety and efficiency. This study combines insights from real-world driving data and simulated environments to explore two key questions:

1. Do human drivers change their behavior when interacting with autonomous vehicles (AVs)?
2. Can driving simulators help identify driver types for easier takeovers?
Using naturalistic driving data from the NGSIM and Lyft Level 5 datasets, we investigate car-following behavior and find that human drivers exhibit statistically significant changes when following AVs, characterized by reduced mean and variance in headway times. These behavioral shifts are influenced by factors such as road type, lane position, and time of day, highlighting the nuanced impacts of AVs on traffic dynamics.

Complementing this, data from the CARLA simulator was analyzed to identify and classify driver types based on behavioral features like gaze patterns, velocity changes, and gaze shift rates. K-means clustering revealed distinct driver typologies—normal, cautious, aggressive, and erratic—while deep learning models achieved high accuracy in classifying these patterns.

These findings support the development of adaptive automated systems that can anticipate and respond to driver behaviors, particularly during control transitions. Together, the results bridge the gap between real-world and simulated environments, providing a holistic understanding of human-AV interactions. By quantifying behavior changes and enabling tailored interventions, this work offers meaningful insights into the coexistence of human drivers and AVs, fostering safer and more efficient roadway systems.

📅 زمان:
دوشنبه ۲۶ آذر ماه ۱۴۰۳
⏱️ ساعت ۱۶:۳۰

💻 لینک برگزاری وبینار:
vc.sharif.edu/ie-seminar

#دانشکده_مهندسی_صنایع
#دانشگاه_صنعتی_شریف

🆔 @IE_SUT

BY کانال دانشکده مهندسی صنایع