Search Results

There has been rapid growth in the mobile-phone industry in terms of technology and growing number of users with migration into the car environment. There is also a significant demand for smart phones capable of accessing email, listening to music, organizing daily activities, linking to social networking sites, while the user is on the move. The automotive industry has been significantly impacted by such mobile-phone usage. Driving a car is a complicated and skillful task requiring attention and focus. However many people perceive driving to be easy - second-to-habit or an extension of their natural skills. This complacency encourages drivers to multitask while driving. While many drivers manage this multitasking comfortably, it becomes a distraction and contributes to increased risk while driving for some. Since the effect of multitasking is variable on different drivers, it is important to understand its impact on individual drivers.

The use of portable devices for in-vehicle environments has become a major cause for driver distraction which can be a contributing factor in crashes of varying intensity. Despite this fact, the number of drivers choosing to use using these devices while driving is increasing rapidly. On the positive side, smart portable devices are equipped with a variety of useful sensors such as cameras, microphones, accelerometer, gyroscope, etc. which could be leveraged to help reduce driver distraction. Careful utilization and delivery of information extracted from these sensors could potentially prove more useful to drivers rather than distracting them. As a proof of concept, using the sensor information available from an off-the-shelf smart portable device, an automatic system is proposed here for driving maneuver recognition and analysis. Driving maneuvers form the basic building blocks of the driver's intent in completing a route.

A number of studies have shown that driving an unfamiliar vehicle has the potential to introduce additional risks, especially for novice drivers. However, these studies have generally used statistical methods in analyzing crash and near-crash data from different driver groups, and therefore the evaluation might be subjective and limited. For a more objective perspective, we suggested that it would be worthwhile to consider the vehicle dynamic signals from the CAN-Bus. In this study, 20 drivers participated in our experiment, where a Gaussian model was used to model individual driver behavior, as well as using a dissimilarity score, which is measured by the squared Euclidean distance in the vehicle dynamical feature space, to evaluate driving performance. Results show that the variation of driving performance caused by driver experience and vehicle familiarity (i.e., driver experienced vs. non-experienced; familiar vs. unfamiliar with vehicle) was clearly observed.

In-vehicle signal processing plays an increasingly important role in driving behavior and traffic modeling. Maneuvers, influenced by the driver's choice and traffic/road conditions, are useful in understanding variations in driving performance and to help rebuild the intended route. Since different maneuvers are executed in varied lengths of time, having a fixed time window for analysis could either miss part of maneuver or include consecutive maneuvers in it evaluation. This results in reduced accuracies in maneuver analysis. Therefore, with access to continuous real-time in-vehicles signals, a suitable framing strategy should be adopted for maneuver recognition. In this paper, a non-uniform time window analysis is presented.

In the stage of automobile industry transition from SAE level “0,1” low autonomous through “2,3,4” human-in-the-loop and ultimately “5” fully autonomous driving, advanced driving monitor system is critical to understand the status, performance, and behavior of drivers for next-generation intelligent vehicles. By making necessary warnings or adjustments, they could operate collaboratively to ensure a safe and efficient traffic environment. The performance and behavior can be viewed as a reflection of the driver’s cognitive workload, which corresponds as well to the environment of their driving scenarios. In this study, image features extracted from driving scenarios, as well as additional environmental features were utilized to classify driving workload levels for different driving scenario video clips.

The use of smart portable devices in vehicles creates the possibility to record useful data and helps develop a better understanding of driving behavior. In the past few years the UTDrive mobile App (a.k.a MobileUTDrive) has been developed with the goal of improving driver/passenger safety, while simultaneously maintaining the ability to establish monitoring techniques that can be used on mobile devices on various vehicles. In this study, we extend the ability of MobileUTDrive to understand the impact on driver performance on public roads in the presence of distraction from speech/voice based tasks versus tactile/hands-on tasks. Drivers are asked to interact with the device in both voice-based and hands-on modalities and their reaction time and comfort level are logged. To evaluate the driving patterns while handling the device by speech/hand, the signals from device inertial sensors are retrieved and used to construct Gaussian Mixture Models (GMM).

Understanding driving behavior is crucial for enhancing traffic safety. While previous studies have primarily explored driving behavior using either statistical or machine learning methods, comprehensive assessments employing both methods under various driving mode are limited. In this study, we employ both machine learning and statistical approaches to model driving behavior. First, we design a comprehensive driver grading system to assess the behavior of drivers under different driving modes. Additionally, we present an extended isolation forest-based model to classify driving behavior using data without labels, saving time and effort. Results illustrate that safe driving is more consistent and stable, while aggressive driving exhibits more intensive changes. They also demonstrate that drivers can exhibit various behaviors under different modes, serving as a benchmark for further driver modeling.