Nagoya University · Engineering
Professor Yuanfang Zhu's research lab specializes in intelligent transportation systems, with a focus on driver behavior analysis and driving style assessment using advanced data analytics. The lab leverages naturalistic driving data, GPS trajectories, and G-G diagram-based methods to classify driving styles—particularly distinguishing aggressive from non-aggressive behaviors—through unsupervised and statistical learning techniques. Research also addresses the unique challenges of elderly drivers, emphasizing safe driving support through behavioral modeling and risk detection. The lab integrates signal processing, machine learning, and transportation informatics to support applications in usage-based insurance, driver feedback systems, and aging driver safety.
Figures are computed from collected data and may differ slightly.
Driving style assessment plays an important role in intelligent transportation system (ITS) applications, such as driving feedback provision and usage-based insurance. Many previous studies used supervised algorithms to profile drivers. However, this cannot be applied to large-scale unlabeled driving data, which are increasingly prevalent in the ITS context. This paper proposes a framework that combines lateral and longitudinal accelerations to assess a driver’s driving style using an unsupervis
With the aging of the population, the number of elderly drivers is increasing. Elderly drivers tend to overestimate their driving abilities, despite their maneuvering skills and cognitive function worsening with aging. Therefore, understanding the driving behavior of older drivers is becoming increasingly important for assisting their safe driving. This research analyzes the characteristics of old drivers' driving behaviors by utilizing the GPS trajectory data collected by the driving recorders
This paper proposes a novel method to classify drivers' driving styles based on G-G diagrams that display the lateral and longitudinal accelerations on the x- and y-axes. A data-based safe driving area was defined to distinguish risky acceleration points from safe acceleration points. The Jensen–Shannon divergence was used to measure the similarity between the distributions of data points outside the safe driving area. A hierarchical clustering algorithm was used to classify the drivers into two
Open papers in the app to read, cite, and organize with AI.
