Vehicle automation has garnered increasing attention as a means of improving safety and efficiency. Advanced Driver Assistance Systems (ADAS) have gained popularity in the transport industry. However, establishing an appropriate level of trust in these systems is crucial for their successful implementation. This research explores the factors influencing driver trust calibration in different levels of automation within driver assistance systems for commercial mobility trucks to ensure drivers comprehend the limitations of these systems and uphold road safety. A qualitative approach involved eleven interviews and observations with drivers to explore their perceptions, experiences, and expectations regarding these systems. The study’s findings extend the Hoff and Bashir Trust model to include significant social factors in calibrating trust. These findings offer valuable insights into the various trust factors that impact driver trust calibration at different levels of automation in driver assistance systems for commercial mobility trucks. These insights contribute to academia in that they help understand how trust in automation is formed and calibrated in real-world settings. In the automotive industry, they can guide the design and implementation of these systems to enhance future drivers’ safety and overall experience.

Advanced Driver Assistance Systems (ADAS) and semi-autonomous driving systems are intended to enhance driver performance and improve transportation safety. The potential benefits of these technologies, such as reduction in number of crashes, enhancing driver comfort or convenience, decreasing environmental impact, etc., are well accepted and endorsed by transportation safety researchers and federal transportation agencies. Even though these systems afford safety advantages, they challenge the traditional role of drivers in operating vehicles. Driver acceptance, therefore, is essential for the implementation of ADAS and semi-autonomous driving systems into the transportation system. These technologies will not achieve their potential if drivers do not accept them and use them in a sustainable and appropriate manner. The potential benefits of these in-vehicle assistive systems presents a strong need for research. A comprehensive review of current literature on the definitions of acceptance, acceptance modelling approaches, and assessment techniques was carried out to explore and summarize the different approaches adopted by previous researchers. The review identified three major research needs: a comprehensive evaluation of general technology acceptance models in the context of ADAS, development of an acceptance model specifically for ADAS and similar technologies, and development of an acceptance assessment questionnaire. Two studies were conducted to address these needs. In the first study, data collection was done using two approaches: a driving simulator approach and an online survey approach. In both approaches, participants were exposed to an ADAS and, based on their experience, responded to several survey questions to indicate their attitude toward using the ADAS and their perception of its usefulness, usability, reliability, etc. The results of the first study showed the utility of the general technology acceptance theories to model driver acceptance. A Unified Model of Driver Acceptance (UMDA) and two versions (a long version with 21 items and a short version with 13 items) of an acceptance assessment questionnaire were also developed, based on the results of the first study. The second was conducted to validate the findings of first study. The results of the second study found statistical evidence validating UMDA and the two versions of the acceptance assessment questionnaire.

A comparison of performance between tradition support vector machine (SVM), single kernel, multiple kernel learning (MKL), and modern deep learning (DL) classifiers are observed in this thesis. The goal is to implement different machine-learning classification system for object detection of three dimensional (3D) Light Detection and Ranging (LiDAR) data. The linear SVM, non linear single kernel, and MKL requires hand crafted features for training and testing their algorithm. The DL approach learns the features itself and trains the algorithm. At the end of these studies, an assessment of all the different classification methods are shown.

Indiana University-Purdue University Indianapolis (IUPUI) / Traffic accidents cause millions of injuries and tens of thousands of fatalities per year worldwide. This thesis briefly reviews different types of active safety systems designed to reduce the number of accidents. Focusing on lane departure, a leading cause of crashes involving fatalities, we examine a lane-keeping system proposed by Minoiu Enache et al.They proposed a switched linear feedback (LMI) controller and provided two switching laws, which limit driver torque and displacement of the front wheels from the center of the lane. In this thesis, a state feedback (LQR) controller has been designed. Also, a new switching logic has been proposed which is based on driver's torque, lateral offset of the vehicle from the center of the lane and relative yaw angle. The controller activates assistance torque when the driver is deemed inattentive. It is deactivated when the driver regains control. Matlab/Simulink modeling and simulation environment is used to verify the results of the controller. In comparison to the earlier switching strategies, the maximum values of the state variables lie very close to the set of bounds for normal driving zone. Also, analysis of the controller’s root locus shows an improvement in the damping factor, implying better system response.

In this paper, the architecture and implementation of an embedded controller for a steering based semi-autonomous collision avoidance system on a 1/10th scale model is presented. In addition, the development of a 2D hardware-in-the-loop simulator with vehicle dynamics based on the bicycle model is described. The semi-autonomous collision avoidance software is fully contained onboard a single-board computer running embedded GNU/Linux. To eliminate any wired tethers that limit the system’s abilities, the driver operates the vehicle at a user-control-station through a wireless Bluetooth interface. The user-control-station is outfitted with a game-controller that provides standard steering wheel and pedal controls along with a television monitor equipped with a wireless video receiver in order to provide a real-time driver’s perspective video feed. The hardware-in-the-loop simulator was developed in order to aid in the evaluation and further development of the semi-autonomous collision avoidance algorithms. In addition, a post analysis tool was created to numerically and visually inspect the controller’s responses. The ultimate goal of this project was to create a wireless 1/10th scale collision avoidance research platform to facilitate human studies surrounding driver assistance and active safety systems in automobiles. This thesis is a continuation of work done by numerous Cal Poly undergraduate and graduate students.

More than 900 bicyclists died in motor vehicle crashes in the United States in 2020, which represents a 50% increase from 2010 and the highest number of bicyclist deaths in nearly 35 years [1]. Reversing this trend will require efforts on multiple fronts, including reducing vehicle speeds and improving roadways and vehicles to be more hospitable to cyclists. Automatic emergency braking (ABB) with cyclist detection is a vehicle countermeasure with potential to prevent bicycle-motor vehicle crashes. AEB systems, which typically warn drivers of an impending collision and brake if drivers do not respond, have been shown to reduce vehicle-to-vehicle rear-end crash rates by 50% [2] and pedestrian crash rates by 27% [3]. Little is known about the real-world effects of ABB with cyclist detection on bicycle crashes. Subaru's EyeSight system, which includes ABB, has been capable of detecting cyclists in parallel configurations beginning in model year (MY) 2013 in the United States. The ability to detect cyclists in perpendicular configurations was added to some models beginning in MY 2022. The goal of this study is to evaluate the effects of the early version of EyeSight on U.S. bicycle crashes. [from Introduction]

No description available.

Annually, road accidents cause more than 1.2 million deaths, 50 million injuries, and US$ 518 billion of economic cost globally. About 90% of the accidents occur due to human errors such as bad awareness, distraction, drowsiness, low training, fatigue etc. These human errors can be minimized by using advanced driver assistance system (ADAS) which actively monitors the driving environment and alerts a driver to the forthcoming danger, for example adaptive cruise control, blind spot detection, parking assistance, forward collision warning, lane departure warning, driver drowsiness detection, and traffic sign recognition etc. Unfortunately, these systems are provided only with modern luxury cars because they are very expensive due to numerous sensors employed. Therefore, camera-based ADAS are being seen as an alternative because a camera has much lower cost, higher availability, can be used for multiple applications and ability to integrate with other systems. Aiming at developing a camera-based ADAS, we have performed an ethnographic study of drivers in order to find what information about the surroundings could be helpful for drivers to avoid accidents. Our study shows that information on speed, distance, relative position, direction, and size & type of the nearby vehicles & other objects would be useful for drivers, and sufficient for implementing most of the ADAS functions. After considering available technologies such as radar, sonar, lidar, GPS, and video-based analysis, we conclude that video-based analysis is the fittest technology that provides all the essential support required for implementing ADAS functions at very low cost. Finally, we have proposed a Smart-Dashboard system that puts technologies – such as camera, digital image processor, and thin display – into a smart system to offer all advanced driver assistance functions. A basic prototype, demonstrating three functions only, is implemented in order to show that a full-fledged camera-based ADAS can be implemented using MATLAB. / Phone# 00966-56-00-56-471

M.Ing. (Electrical and Electronic Engineering) / Adaptive Cruise Control (ACC) is proposed to assist drivers tedious manual acceleration or braking of the vehicle, as well as with maintaining a safe headway time gap. This thesis proposes, simulates, and implements a vision-based ACC system which uses a single camera to obtain the clearance distance between the preceding vehicle and the ACC vehicle. A three-step vehicle detection framework is used to obtain the position of the lead vehicle in the image. The vehicle coordinates are used in conjunction with the lane width at that point to estimate the longitudinal clearance range. A Kalman filter filters this range signal and tracks the vehicle’s longitudinal position. Since image processing algorithms are computationally intensive, this document addresses how adaptive image cropping improves the update frequency of the vision-based range sensor. A basic model of a vehicle is then derived for which a Proportional-Integral (PI) controller with gain scheduling is used for ACC. A simulation of the system determines whether the ACC algorithm will work on an actual vehicle.

A segurança no tráfego de carros é um assunto em foco nos dias de hoje e, dentro dele, podem-se citar os sistemas de auxílio ao motorista que vêm sendo desenvolvidos com a finalidade de reduzir o grande número de fatalidades em acidentes de trânsito. Tais sistemas de auxílio buscam mitigar falhas humanas como falta de atenção e imprudência. Visto isso, o projeto SENA, desenvolvido pelo Laboratório de Robótica Móvel da Escola de Engenharia de São Carlos, busca contribuir com a evolução dessa assistência ao motorista. O presente trabalho realiza um estudo sobre uma técnica de inteligência artificial chamada de Redes Bayesianas. Essa técnica merece atenção em virtude de sua capacidade de tratar dados incertos em forma de probabilidades. A rede desenvolvida por esse trabalho utiliza, como dados de entrada, os classificadores em desenvolvimento no projeto SENA e tem como resposta um comportamento que o veículo deve executar, por um ser humano ou por um planejador de trajetórias. Em função da alta dimensionalidade do problema abordado, foram realizados dois experimentos em ambiente simulado de duas situações distintas. A primeira, um teste de frenagem próximo a um ponto de intersecção e a segunda, um cenário de entroncamento. Os testes feitos com a rede indicam que classificadores pouco discriminantes deixam o sistema mais propenso a erros e que erros na localização do ego-veículo afetam mais o sistema se comparado a erros na localização dos outros veículos. Os experimentos realizados mostram a necessidade de um sistema de tempo real e um hardware mais adequado para tratar as informações mais rapidamente / The safety of cars in traffic scenarios is being addressed on the past few years. One of its topics is the Advanced Driver-Assistance Systems which have been developed to reduce the fatality numbers of traffic accidents. These systems try to decrease human failures, such as imprudence and lack of attention while driving. For these reasons, the SENA project, in progress on the Mobile Robotics Laboratory at the Sao Carlos School of Engineering (EESC), aims to contribute for the evolution of these assistance systems. This work studies an artificial intelligence technique called Bayesian Networks. It deserves our attention due to its capability of handling uncertainties with probability distributions. The network developed in this Masters Thesis has, as input, the result of the classifiers used on SENA project and has, as output, a behavior which has to be performed by the vehicle with a driver or autonomously by the means of a path planner. Due to the high dimensionality of this issue, two different tests have been carried out. The first one was a braking experiment near a intersection point and the other one was a T-junction scenario. The tests made indicate that weak classifiers leaves the system more instable and error-prone and localization errors of the ego-vehicle have a stronger effect than just localization errors of other traffic participants. The experiments have shown that there is a necessity for a real-time system and a hardware more suitable to deal quickly with the information