A driving simulator reproduces the essential features of a vehicle and provides an interface for direct human operation. It provides a safe and less expensive way of training people how to drive. Against the backdrop of a comprehensive literature survey on driving simulators and their applications, this thesis endeavours to make five unique contributions. Many of the military armoured vehicles have eight wheels, are able to cross trenches of approximately two meters, and can climb steps of as high as one meter. Available research, however, focuses primarily on the vehicle dynamics modelling of commercial four wheel vehicles. In this thesis, a mathematical model is given for simulating the vehicle dynamics of an eight wheel vehicle over rough terrain, taking into account the limitations of real-time driving simulation. A discussion of the model by Janse van Rensburg et al. is contained in a paper which is currently under review by the International Journal of Modern Physics C (IJMPC). To prove the validity of a vehicle model, it is necessary to provide a method of testing the model. Detail about the vehicle dynamics model used is not always available when developed by a third party. This thesis describes a “black box” testing method for the verification of a vehicle dynamics model. An article regarding this matter by Janse van Rensburg et al. has been submitted to the IJMPC and is currently under review. Normally, the focus on driving simulators is on the modelling of realistic vehicle dynamics models. However, the design of a realistic simulation environment is of equal importance. A human driver usually steers one vehicle, but the rest of the vehicles used in the simulation should be managed by a computer program. An automatic driver model is described to be used within the simulation environment. The current presentation is based on the published paper [86] by Janse van Rensburg et al. (IJMPC, 16(6):895-908, 2005). An understanding of three-dimensional coordinate system transformations is one of the most important parts of a flight or driving simulator. Although the procedure of using Euler angles for coordinate system transformations is nothing new, almost no literature is available of how it can be applied on more complex situations. This thesis supplies more information on how a program language such as C++ could be used to apply more complex coordinate transformations in real-life situations. Results appeared in the published paper by Janse van Rensburg et al. (IJMPC, 16(6):909-920, 2005). Finally the use of vocoders is proposed for the modelling of engine sound. For a driving simulator which should be an exact replica of a certain vehicle, an accurate sound model is of extreme importance. By using vocoders, a technique used for the manipulation of voice, a higher level of accuracy and realism can be obtained than with the methods currently discussed in literature. A paper on this matter, compiled by Janse van Rensburg et al. is currently under review by the IJMPC. / Prof. M. A. van Wyk

The purpose of the study was to examine the efficacy of virtual reality exposure therapy (VRET) to treat driving phobia using a multiple baseline across-subjects design. The sequence of events included a pre-treatment assessment, a baseline phase, 8 weekly VRET sessions using a standardized treatment protocol, a post-treatment assessment, and 1- and 3-month follow-up assessments. A sample of seven treatment seeking adults with a primary diagnosis of specific phobia (driving) was recruited. Five completed the treatment and follow-up phases. One individual withdrew after the pre-treatment assessment, and the other, after the first treatment session. It was hypothesized that VRET would reduce driving anxiety and avoidance symptoms between pre- and post-treatment assessments using several outcome measures. Visual and statistical analysis methods were used to assess treatment outcome. Three participants showed clear improvement in driving anxiety and avoidance symptoms between pre- and post-treatment assessments. There was a marginal improvement in these symptoms for one participant. The remaining participant showed very little improvement, and some outcome measures revealed slight deterioration in some of her symptoms. There was negligible change in actual driving frequency in any participant. Some gains were lost at the 1- and 3-month follow-up assessments, but symptoms remained far below pre-treatment results. Possibilities for future research and practice implications are discussed. / Education, Faculty of / Educational and Counselling Psychology, and Special Education (ECPS), Department of / Graduate

Background. Today autonomous vehicles (AV) are a well-debated topic and manystudies try to understand or solve different ethical dilemmas concerning self-driving.There are legal issues and liability concerns that require clarity from a governmentfor any type of AV to operate on the road. The Swedish transport agency has notallowed autonomous cars because they do not fulfill the requirements to be classifiedas a safe vehicle and there are only a few studies in Sweden that focus on this topic.  Objectives. This thesis will investigate previous research in this area to find outwhat are the most common topics that are a concern to people when it comes toAVs. Based on the findings from the literature review, a survey will be distributed,containing both qualitative and quantitative questions. Methods. A literature review has been performed to set up the base work for thesurvey. The questions to the survey were based on the findings from the literaturereview thereafter a pilot survey was distributed. The result from the survey will beanalysed by categorizing themes and keywords from the respondents. Results. The literature review can reveal that the most concerning categories according to people are safety, security, legal aspects & liability, and cost. The respondents to the survey indicate that safety is a high prioritized topic together with anylegal and liability aspects. The result also shows that there is a high expectationfrom the car manufacturers to make the vehicle secure from any attacks. As opposedto other research the Swedish respondents in this survey do not show much concernwhen it comes to costs.  Conclusions. A common theme from the respondents shows that safety must beprioritized before any AVs can be released on the road, with no exceptions. Many ofthe respondents show that they trust the Swedish government and its legal systemwhich indicates that the acceptance of AVs could correlate with an acceptance fromthe Swedish transport agency. Though the respondents in the survey do seem to trustin the companies ability to provide a secure vehicle there are still some concernedabout the possibility of being a victim of a cyber attack. There are no concernsabout the cost of the autonomous technology amongst the respondents but it is notevaluated why and that leaves those results open for interpretation.

TITLE: DELAY DISCOUNTING AND CAMPUS SPEEDING BEHAVIORMAJOR PROFESSOR: Dr. Ryan Redner Speeding, as defined by exceeding the speed limit posted in a designated area, is a problem that has a direct negative effect on a majority of America, college campuses not being exempt. A minimal amount of research has been conducted on specifically dangerous driving such as speeding on university campuses; this paper looks to expand upon the existing research. In addition, this study serves the purpose of attempting to correlate speeding and impulsivity. This was achieved by utilizing a survey in conjunction with delay discounting scenarios embedded within. Participants (n = 89) took a survey that involved questions about demographic information and past driving. The survey also included two hypothetical scenarios that encouraged participants to answer whether they would speed in the presented scenarios and at what point. To complete the survey the MCQ (Monetary Choice Questionnaire) was placed at the end. To assess the results the 89 participants were split into two groups two different times to be analyzed. One group was split into an impulsive group and non-impulsive group while the other group was split into speeders and non-speeders. As expected, both the impulsive and speeders were more likely to discount speed at a faster rate; this shown at higher rates in the penalty hypothetical scenario. The implications of this study include that those with that display higher impulsivity, shown through k-values, may have a higher tendency to engage in dangerous driving such as speeding, as well as understanding that penalty conditions, such as point loss, show a correlation with speeding.

This thesis examined Bill C-46 as a case study of the constructed nature of public policy as inspired by Foucault’s (1982) argument that this construction, over time, can become “truth.” Our main research question was if the Canadian governments' construction of the problem of cannabis-impaired driving, which was used to justify the necessity and the content of new offences in Bill C-46, was contested in whole or in part during the parliamentary debates that resulted in the adoption of this law or, was the construction of cannabis-impaired driving accepted without contest by most of the parliamentarians and witnesses? This thesis used Bacchi and Goodwin’s (2016) What’s the Problem Represented to be (WPR) framework to answer this question. Our two main findings of this thesis were first, that the construction of the problem was not substantially contested by parliamentarians or witnesses during the debate and second, that inequality is embedded in the content of Bill C-46. Our analysis presented the constructed division between “good citizens” who are not punishable by criminal sanctions for impaired driving or problematic driving behaviours such as driving while fatigued, or distracted driving. On the other hand, the “bad citizens”, including drivers who had consumed cannabis, were perceived as deserving of criminal punishment despite not necessarily being impaired. The Liberal Party was concerned more with political viability of cannabis legalization resulting in a problem construction that generates ineffective solutions to the objective of promoting road safety.

Context: In this work a system, the eco service, is developed that incorporates elements from gamification to help drivers adapt to a more energy-efficient driving style. An energy-efficient driving style can help reduce fuel consumption, increase traffic safety and help reduce the emissions made from vehicles. Objectives: The main goal of this work is to explore ways of how gamification can be used in the context of eco-driving. Evaluating different elements and how they work in this context is important to help drivers to continue improving their driving style. Method: The eco service was tested on 16 participants where each participants was asked to drive a predetermined route. During the experiment the participants were given access to the eco service in order to gain feedback on their driving. Lastly interviews were held with each participant on questions regarding the use of gamification and how it can be improved in the context of eco-driving. The research was done in collaboration with a swedish company, Swedspot AB, that works with software solutions for connected vehicles. Results & Conclusions: Positive results were found on the use of gamification. Participants reported that the eco service made them more aware of their driving situation and how to improve. Game elements with positive influence were reward and competitive based and helped motivate the driver to improve.

Neuromorphic computing has many opportunities in future autonomous systems, especially those that will operate at the edge. However, there are relatively few demonstrations of neuromorphic implementations on real-world applications, partly because of the lack of availability of neuromorphic hardware and software, but also because of the lack of availability of an accessible demonstration platform. In this work, we propose utilizing the F1Tenth platform as an evaluation task for neuromorphic computing. F1Tenth is a competition wherein one tenth scale cars compete in an autonomous racing task; there are significant open source resources in both software and hardware for realizing this task. We present a workflow with neuromorphic hardware, software, and training that can be used to develop a spiking neural network for neuromorphic hardware deployment to perform autonomous racing. We present initial results on utilizing this approach for this small-scale, real-world autonomous vehicle task.

PURPOSE: The purpose of this study was to investigate the potential relationship between alcohol and drug prevalence in drunk- and drug-impaired driving cases in New York City (NYC) between January 1, 2015 and December 31, 2017 and to determine how this prevalence has changed over time. The study also investigated the demographic characteristics of drivers to determine if there are certain groups who are consistently involved in alcohol and/or drug abuse while operating a motor vehicle. METHODS: This retrospective study determined the alcohol and drug prevalence in individual drivers represented as cases per year over three consecutive years. A total of 613 cases were included in the study for individuals, age 16 to 75 years old arrested for suspicion of driving while intoxicated (DWI) in NYC. Individual data collected included basic demographic information, time and day of incident, borough in which incident occurred, type of matrix used for toxicological analysis and the presence and absence of alcohol and/or drugs. Drug findings were combined into classes based on their likely effect and included the following categories: alcohol, antidepressants, cannabinoids, narcotic analgesics, sedatives, stimulants and other. RESULTS: Results from the study compared data over three consecutive years from DWI cases (2015 to 2017). In comparing prevalence of drug classes by year, the percent of cases tested positive for cannabinoids, narcotic analgesics and stimulants changed significantly from 2015 to 2017. Delta-9-tetrahydrocannabinol (THC), the active component of marijuana, was the most frequent individual drug identified using a screening method. The prevalence rate of cannabinoids increased significantly in 2017 to 43.0% from 32.5% the previous year and 29.3% in 2015. The narcotic analgesics prevalence rate increased significantly in 2016 to 28.5% from 13.4% in the previous year and slightly decreased to 26.9% in 2017. Comparison of stimulants by year showed a significant increase in 2017, 28.1% versus 19.0% (2016) versus 18.3% (2015). When comparing the 2017 results to the drugs tested for in 2015 and 2016, significantly higher daytime drug prevalence was found between the previous years and 2017. In evaluating race and drug use, white drivers were significantly more likely to test positive for sedatives and stimulants than other races. In Manhattan, there was a significantly higher alcohol detection rate compared to the other boroughs and in Staten Island there was a significantly higher narcotic analgesics detection rate. In comparing the top five individual drugs identified by borough, cannabinoids were the most common drug across all of the boroughs. Alprazolam and cocaine (identified by its metabolite, benzoylecgonine, 98% of the time) were the next most frequently encountered drugs alternating as the top two and three drugs identified in the following four boroughs: Manhattan, Queens, the Bronx, and Staten Island. Phencyclidine (PCP) (“angel dust”) was identified in the top five for Manhattan, the Bronx and Staten Island. A statistically significant negative association was found between cannabinoid-positive and alcohol-positive drivers. The percentage of drivers with a BAC greater than .08 g/dL was significantly lower among cannabinoid-positive drivers than those who tested negative for cannabinoids. Although there were no strong correlations between drug classes, sedatives were associated (according to significant correlations) most to other drugs (correlated to 6 out of 6 categories). CONCLUSIONS: This study summarizes the results of the first OCME FTL of NYC toxicological findings in DWI cases to estimate alcohol and drug-involved driving prevalence. It is important to note that this is a prevalence study and not a study that reports the risks associated with drugged-driving. Since many drugs may be detected long after its impairing effects are gone, the focus of this study was to merely convey the use of particular drugs in the driving population.

Use of cellular phones while driving, and safety implications thereof, has captured public and scientific interest. Previous research has shown that driver reactions and attention are impacted by cellular phone use. Generally, previous research studies have not focused on how visual attention and driver performance may interact. Strayer and colleagues found lower recognition for items present in the driving environment when drivers were using a cellular phone than when not using the phone; however, the tested items were not directly relevant to driving. Relevance to driving may have an impact on attention allocation. The current project used a mediumidelity driving simulator to extend previous research in two ways: 1) how attention is allocated across driving-relevant and -irrelevant items in the environment was investigated, and 2) driving performance measures and eye movement measures were considered together rather than in isolation to better illustrate the impact of cellular phone distraction on driver behavior. Results from driving performance measures replicated previous findings that vehicle control is negatively impacted by driver distraction. Interestingly, there were no interactions of relevance and distraction found, suggesting that participants responded to potential hazards similarly in driving-only and distraction conditions. In contrast to previous research, eye movement patterns (primarily measured by number of gazes) were impacted by distraction. Gaze patterns differed across relevance levels, with hazards receiving the most gazes, and signs receiving the fewest. The relative size of the critical items may have impacted gaze probability in this relatively undemanding driving environment. In contrast to the driving performance measures, the eye movement measures did show an interaction between distraction and relevance; thus, eye movements may be a more direct and more sensitive measure of driver attention. Recognition memory results were consistently near chance performance levels and did not reflect the patterns found in the eye movement or driving performance measures.

Fatigue and drowsy driving in the state of Utah has been a causal factor in thousands of crashes over the years and poses a serious threat to public safety. Consequently, the purpose of this research was to evaluate the impact of drowsy driving in the state, to identify locations where fatigue and drowsy driving may be contributing factors to current crashes, and to identify methods to help mitigate these crashes. A 3-year drowsy driving crash rate spanning the years 2002 – 2004 was used to determine which segments of Utah highway are most prone to drowsy driving crashes. Drowsy driving corridors were located on Interstates 15, 70, 80, and 84 as well as United States Routes 89 and 91. Furthermore, State Route 36 also had two drowsy driving corridors. In order to recommend appropriate drowsy driving countermeasures for the drowsy driving corridors, a review of the existing countermeasures was conducted. The existing countermeasures included cable median barrier, rumble strips, rest areas, and drowsy driving freeway signage. The freeway signage is used to alert drivers of the adverse effects of drowsy driving and was the basis for a before-after study as well as a public survey of drowsy driving along Interstate 80 west of Salt Lake City. The before-after study of the drowsy driving freeway signage concluded that the freeway signage has played a part in reducing the number of crashes by as much as 63 percent in the eastbound direction and by as much as 22 percent in the westbound direction. As indicated, a public survey was conducted at two rest areas to supplement the findings of the before-after analysis. Using the 405 completed surveys, 14 Chi-Square tests were conducted with five of the test yielding statistically significant results. Finally, recommendations were made for the 41 drowsy driving corridors resulting from the 3-year crash rate analysis. Drowsy driving countermeasures recommended include: additional shoulder and centerline rumble strips, cable median barrier, guardrail replacement, and drowsy driving highway signage. Drowsy driving countermeasures not yet implemented but which should be considered by the Utah Department of Transportation are transverse rumble strips, wider longitudinal pavement markings, in-lane pavement markings indicating “AVOID FATIGUE DRIVING,” minimizing edge drop off, flattening slopes in clear zones, and adding a modified rest area.