This research is focused on driver behavior in traffic, especially during car-following situations and safety critical events. Driving behavior is considered as a human decision process in this research which provides opportunities for an artificial driver agent simulator to learn according to naturalistic driving data. This thesis presents two mechine learning methodologies that can be applied to simulate driver naturalistic driving behavior including risk-taking behavior during an incident and lateral evasive behavior which have not yet been captured in existing literature. Two special machine learning approaches Backpropagation (BP) neural network and Neuro-Fuzzy Actor Critic Reinforcement Learning (NFACRL) are proposed to model driver behavior during car-following situation and safety critical events separately. In addition to that, as part of the research, state-of-the-art car-following models are also analyzed and compared to BP neural network approach. Also, driver heterogeneity analyzed by NFACRL method is discussed. Finally, it presents the findings and limitations drawn from each of the specific issues, along with recommendations for further research. / Master of Science

In 2010, there were 6,714,288 total registered drivers in New Jersey. Approximately 4% (267,485) of these drivers had a suspended driver's license. The intent of suspending a driver's license is to keep hazardous drivers off of the roads, in hopes of having a safer driving environment for others on the road. Drivers in New Jersey can have their driver's license suspended for a number of reasons. These include dangerous driving behaviors such as reckless driving and driving under the influence of alcohol or drugs. However, there are also reasons for suspension that have little or nothing to do with driver behavior, such as failure to pay child support, failure to pay MVC insurance surcharge, or failure to appear in court. While these offenses are all due of consequence, they have little or nothing to do with driver behavior. This research program will conduct an analysis of the issues and implications of implementing a restricted-use license program for suspended New Jersey drivers, detailing key issues associated with restricted-use license programs. It was found that over two-thirds of suspended drivers in New Jersey receive driver's license suspensions for both driving and non-driving-related offenses, whereas only about four-percent of suspended drivers in New Jersey receive a driver's license suspension for driving-related reasons only. It was also found that drivers suspended for non-driving related reasons have different driver behavior than drivers suspended for driving related reasons. Surveying both New Jersey police chiefs, as well as U.S. state motor vehicle agencies, it was found that there is a generally positive perception of restricted driver's license programs. Overall, it is recommended that the New Jersey Motor Vehicle Commission implement a restricted driver's license program in New Jersey. / Master of Science

The combination of airbag and seatbelt is considered to be the most effective vehicle safety system. However, despite the widespread availability of airbags and a belt use rate of over 85% U.S. drivers involved in crashes continue to be at risk of serious thoracic injury. One hypothesis is that this risk may be due to the lack of airbag deployment or the airbag \'bottoming-out\' in some cases, causing drivers to make contact with the steering. The objective of this study is to determine the influence of various advanced airbags on occupant injury risk in frontal automobile crash. The analysis is based upon cases extracted from the National Automotive Sampling System Crashworthiness Data System (NASS/CDS) database for case years 1993-2011. The approach was to compare the frontal crash performance of advanced airbags against depowered airbags, first generation airbags, and vehicles with no airbag equipped. NASS/CDS steering wheel deformation measurements were used to identify cases in which thoracic injuries may have been caused due to steering wheel impact and deformation. The distributions of injuries for all cases were determined by body region and injury severity. These distributions were used to compare and contrast injury outcomes for cases with frontal airbag deployment for both belted and unbelted drivers. Among frontal crash cases with belted drivers, observable steering wheel deformation occurred in less than 4% of all cases, but accounted for 29% of all serious-to-fatally injured belted drivers and 28% of belted drivers with serious thoracic injuries (AIS3+). Similarly, observable steering wheel deformation occurred in approximately 13% of all cases with unbelted drivers involved in frontal crashes, but accounted for 58% of serious-to-fatally injured unbelted drivers and 66% of unbelted drivers with serious thoracic injuries. In a frontal crash, the factors which were statistically significant in the probability of steering wheel deformation were: longitudinal delta-V, driver weight, and driver belt status. Seatbelt pre-tensioner and load limiters were not significant factors in influencing steering wheel deformation. Furthermore, belted drivers in vehicles with no airbag equipped were found to have 3 times higher odds of deforming the steering wheel, as compared to driver in similar crash scenario. Similarly, unbelted drivers were found to have 2 times greater odds of deforming the steering wheel in vehicles with no airbags equipped as compared to vehicles with advanced airbag. The result also showed no statistically significant difference in the odds of deforming the steering wheel between depowered and advanced airbag. After controlling for crash severity, and driver weight, the study showed that crashes with steering wheel deformation results in greater odds of injury in almost all body regions for both belted and unbelted drivers. Moreover, steering wheel deformation is more likely to occur in unbelted drivers than belted drivers, as well as higher severity crashes and with heavier drivers. Another potential factor in influencing driver crash injury is the knee airbag. After comparing the odds of injury between vehicles with and without knee airbags equipped, belted drivers in vehicles equipped with knee airbag were found to have statistically smaller odds of injury in the thorax, abdomen, and upper extremity. Similarly, the findings showed that unbelted drivers benefited from knee airbag through statistically significant lower odds of chest and lower extremity injuries. However, the results should be considered with caution as the study is limited by its small sample of vehicles with knee airbags. / Master of Science

This study was performed with the goal of delineating how drivers' interactions with an Automated Driving System were affected by a Driver Monitoring System (DMS), which provided alerts to the driver when he or she became inattentive to the driving environment. There were two specific research questions. The first was centered on addressing how drivers' trust and satisfaction with an Automated Driving System was affected by a DMS. The second was centered on addressing how drivers' abilities to detect changes in the driving environment that required intervention were affected by the presence of a DMS. Data were collected from fifty-six drivers during a test-track experiment with an Automated Driving System prototype that was equipped with a DMS. DMS attention prompt conditions were treated as the independent variable and trust, satisfaction, and driver performance during the experimenter triggered lane drifts were treated as dependent variables. The findings of this investigation suggested that drivers who receive attention prompts from a DMS have lower levels of trust and satisfaction with the Automated Driving System compared to drivers who do not receive attention prompts from a DMS. While the DMS may result in lower levels of trust and satisfaction, the DMS may help drivers detect changes in the driving environment that require attention. Specifically, drivers who received attention prompts after 7 consecutive seconds of inattention were 5 times more likely to react to a lane drift with no alert compared to drivers who did not receive attention prompts at all. / Master of Science

Motivated by the desire to explore future traffic flows that will consist of a mixture of classical vehicles and vehicles equipped with advanced driver assistance systems, new mathematical theories and models are developed. The basis for this theory was borrowed from the kinetic description of gas flows, where we replaced the behaviour of the molecules by typical human driving behaviour. From a methodological point of view, this 'human-kinetic' traffic flow theory provides two major improvements with respect to existing theory. Firstly, the model builds exclusively on a mathematical description of individual driver behaviour, whereas traditionally field measurements of traffic flow variables like flow rate and average speed of the flow are needed. This is of major importance for the exploration of future traffic flows with vehicles and equipment that are not yet on the market, and for which at best individual test results from driving simulator experiments or small scale field trials are available. Secondly, the model accounts for the more refined aspects of individual driver behaviour by considering the 'internal' state of the driver (active/passive, aware/unaware,...) and the variations of driving strategy that occur during driving. This is important when the ambition is to capture refined congestion patterns like the occurrence of stop-and-go waves, oscillating congestion and long jams, where the driving strategy may depend for instance on the motivation of the driver to follow closely. This new theory links together the worlds of traffic engineers and behavioural scientists. As such, it is a promising tool that increases the insight in the human behaviour as a basis of various dynamic congestion patterns, and it facilitates the design and evaluation of electronic systems in the vehicle that assist the driver to behave safer, more comfortable and more efficient in busy traffic flows. Herewith, the results of this research are relevant, both for the theoretical interest of the TRAIL research school, and for the more practically oriented work of TNO, who provided financing for this research in the joint T3 research program.

This research examined the relationship between experience and driving performance with secondary tasks. Data were collected from 42 teenage drivers and their parents using an instrumented vehicle for two one hour test track sessions spaced 12 months apart. For part of the sessions, participants followed a lead vehicle which allowed for range data to be collected. Teenage and experienced drivers' driving were compared for cell phone and odometer tasks. Variables such as Speed, Range to Forward Vehicle, and Driving-Related Eyeglance percentages were all analyzed utilizing ANOVA. Post-hoc analysis on continuous data was performed using a Tukey HSD test. Lane Deviations were examined using Chi-Square analyses. Experienced drivers drove faster overall than teenage drivers. Teenage drivers drove faster in the 12 month session than the first session. No significant effects were found for Speed Variance, Range Variance, or Lane Deviations. Experienced drivers had a higher percentage of driving-related glances than teenage drivers. For the odometer task, teenage drivers were found to follow further behind a lead vehicle than adults. Driving experience was believed to have an effect on driver eyeglance patterns due to increased development of attentional control resulting in better switching between the task and the driving environment. Experienced drivers likely drove faster due to increased confidence in their driving ability. This research supports current GDL cell phone restrictions. A drivers' education lesson plan framework was developed to address these differences. Future research should focus on further refining GDL legislation to address the cognitive differences between teenage and experienced drivers. / Master of Science

While haptics, tactile displays, and other topics relating to non-visual user interfaces have been the subject of a variety of research initiatives, little has been done specifically related to those for blind driving. Many automation technologies have been developed for the purpose of assisting and improving the safety of sighted drivers, but to enable a true driving experience without any sense of sight has been an essentially overlooked area of study. Since 2005, the Robotics & Mechanisms Laboratory at Virginia Tech has assumed the task of developing non-visual interfaces for driving through the Blind Driver Challenge®, a project funded by the National Federation of the Blind. The objective here is not to develop a vehicle that will autonomously mobilize blind people, but to develop a vehicle that a blind person can actively and independently operate based on information communicated by non-visual interfaces. This thesis proposes some generalized considerations for the development of non-visual interfaces for driving, using the instructional interfaces developed for the Blind Driver Challenge® as a case study. A model is suggested for the function of blind driving as an open-loop control system, wherein the human is an input/output device. Further, a discussion is presented on the relationship between the bandwidth of information communicated to the driver, the amount of human decision-making involved in blind driving, and the cultivation of driver independence. The considerations proposed here are intended to apply generally to the process of non-visual interface development for driving, enabling efficient concept generation and evaluation. / Master of Science

This thesis explores the possibility of identifying risk behaviour patterns among forklift drivers through the analysis of telemetry data using unsupervised clustering algorithms. The objective is to predict whether certain behaviour patterns increase the risk of accidents. With the increasing accessibility of Internet of Things technology, data from forklifts has become more available, allowing for the study of driver behaviour. The telemetry data utilised is sourced from Toyota Material Handling Manufacturer Sweden’s internal database, collected from Data Handling Units that are installed on forklifts across Europe. This data, referred to as shock data, is triggered when a force is applied to the forklift, such as a collision. The thesis investigates combinations of various clustering algorithms and dataset modifications. The evaluation of the results is conducted using several quantitative measures and visualisation, along with analysis of time distribution, geographical placement, comparison of forklift models, and comparison with "no-shock" data. The evaluation yields K-Prototypes and K-Means as the best performing algorithms, while indicating that soft clustering and density-based clustering are not well-suited for the data. The identified best performing algorithms reveal two recurring driver behaviour patterns: the first one being driving forward at high speed with the lift motor idle, and the second pattern being driving backward at low speed while lowering the forks. Furthermore, a majority of the data points remain unclassified into specific behaviour patterns, suggesting that the dataset or methods used may not be sufficient enough. The inclusion of additional featuers, such as steering angle and forklift height, should be considered for exploration in future work. The thesis demonstrates the feasibility of identifying risk behaviour patterns, with potential for future research expanding on the findings to further contribute to the prevention of workplace accidents involving forklifts.

The aging process is often accompanied by declines in one or more physical, vision, and/or cognitive abilities that may impact driving safety. As older drivers become more self-aware of these functional deficits, they have the tendency to engage in self-regulation practices, such as less driving and avoiding challenging driving situations. This tendency may gradually evolve to give up driving altogether. Advanced Driver Assistance Systems (ADAS) holds promise for improving older drivers' safety on the road as well as maintaining their mobility by compensating for declines in visual, cognitive, and physical capabilities. However, the perception of these technologies can influence the realization of these expected benefits. The overarching goal of this research is to understand and enhance the safety and mobility of older adults by examining the impact of ADAS. The dissertation addresses this goal by investigating mobility, perception, safety measures, and safety. Study 1 employed structure equation modeling (SEM) on the data from the Second Strategic Highway Research Program (SHRP 2) on driving habits with respect to age, gender, living status, health, and functioning capabilities. The results illustrate that older drivers' health is a reliable predictor of driving exposure, and cognitive and physical declines are predictive of their intention to reduce exposure and actual driving in challenging situations. These findings highlight that the aging population requires support for their mobility and likely road safety given their age-related impairments. Study 2 employed structure topic modeling on a focus group of older adults driving vehicles equipped with ADAS for six weeks was conducted to reveal five key issues to older drivers (in the order of prevalence): (1) safety, (2) confidence concerning ADAS, (3) ADAS functionality, (4) user interface/usability, and (5) non-ADAS related features. The findings point to a need for holistic ADAS design that not only must consider safety concerns but also user interfaces accommodating older adults' preferences and limitations as well as in-depth training programs to operate ADAS given the technology limitations. Study 3 employed correlation analysis and logistic regression on SHRP 2 data to reveal that the longitudinal deceleration events at greater than 0.60g and lateral acceleration events at greater than 0.40g appear most associated with older adults' driving risk and are predictive of near future crash and near-crashes (CNCs) occurrence and high-risk older drivers with acceptable accuracy. These findings indicate that high g-force events can be used to assess risk for older drivers, and the selection of thresholds should consider the characteristics of drivers. Study 4 compared high g-force events between two naturalistic driving studies to reveal that drivers who drove vehicles equipped with ADAS had lower longitudinal declaration rates, indicating the benefits of ADAS presence on older drivers' safety. When lane keeping assist (LKA) was engaged, lower high longitudinal deceleration was observed than when LKA was not engaged, indicating that older drivers tended to apply less aggressive braking when using LKA. Over several weeks of exposure to vehicles with ADAS presence, older drivers showed decreasing longitudinal deceleration but increasing lateral acceleration events. In other words, the potential of ADAS for positive safety-related impacts exists but some refinement in the design to reduce lateral events might be necessary. / Doctor of Philosophy / As people grow older, they may experience declines in their physical, vision, and cognitive abilities, which can affect their ability to drive safely. Many older drivers become more aware of these limitations and tend to drive less or avoid challenging situations, gradually some eventually stop driving altogether. Advanced Driver Assistance Systems (ADAS) hold the potential to enhance the safety and mobility of older drivers by compensating for these declines in vision, cognition, and physical capabilities. However, the way older adults perceive and accept these technologies can influence their effectiveness. This research focuses on understanding and improving the safety and mobility of older adults by examining the impact of ADAS on them through four studies. These studies fill gaps in research and provide insights into the potential of ADAS to enhance both the safety and mobility of older drivers. This research is vital for improving the quality of life for older adults and making our roads safer for all.

Esta pesquisa caracterizou o comportamento de motoristas em interseções semaforizadas sob três aspectos: (1) reação frente à mudança do verde para o amarelo; (2) comportamento durante a desaceleração para parar; e (3) comportamento durante a saída do cruzamento semaforizado. Os dados foram coletados em uma interseção localizada em pista de testes no Virginia Tech Transportation Institute, nos EUA. A amostra foi composta por 60 motoristas voluntários igualmente divididos em função do gênero; dos quais 32 tinham idade inferior a 65 anos (\"jovens\"). Foram investigados efeitos da idade, do gênero e da declividade da via sobre as seguintes situações: tomada de decisão entre parar ou prosseguir no amarelo; posição de parada em relação à faixa de retenção; tempo de percepção e reação (TPR) para frenagem e partida do cruzamento; efeito de zonas de opção e de dilema; taxa de desaceleração para parada na interseção; e taxa de aceleração para partida da interseção. As análises indicaram que: (1) os motoristas mais jovens invadiram mais a faixa de retenção que os idosos; (2) mulheres apresentam maiores TPR para decidir partir da interseção; e (3) o TPR é menor no declive tanto para a decisão de frear quanto para a partida do cruzamento. As taxas de desaceleração não apresentaram influência dos fatores avaliados. Por outro lado, constatou-se que a aceleração foi afetada pelo fator declividade. Como resultado final da pesquisa, foram propostos modelos, em função do tempo, que exprimem a desaceleração/aceleração usada pelos motoristas ao frear/acelerar. Foram propostos modelos para o motorista médio e para motoristas desagregados em três grupos em função da agressividade. / The objective of this research was to characterize driver behavior at signalized intersections according to three aspects: (1) reaction at the onset of the amber phase; (2) behavior during the deceleration to stop at the signal; and (3) behavior during the acceleration to leave the intersection at the onset of the green. The data were collected at a signalized intersection on a private highway, at the Virginia Tech Transportation Institute, in the USA. The sample consisted of 60 volunteer drivers, equally divided by gender. The sample was divided into two age groups: younger drivers (age was less than 65) and older drivers. Effects of gender, age group and roadway grade were investigated for the following aspects: decision making at the onset of amber; final stopping position with relation to the stop line; perception/reaction times (PRT) at the onset of the amber and the green lights; effects of dilemma and option zones; and deceleration and acceleration rates used by the drivers. The analyses suggest that: (1) younger drivers tend to stop farther past the stop line, compared to older drivers; (2) women have longer PRT at the onset of the green; and (3) PRT are shorter on downgrade at the onset of both amber and green lights. The observed deceleration rates were not affected by gender, age group or roadway grade. Acceleration rates were found to be influenced by the grade. A set of models that express the acceleration/deceleration rates as a function of time were proposed to represent the average behavior observed for drivers in the sample. Specific models were also proposed for aggressive, non-aggressive and intermediate drivers.