Safety is an increasingly important aspect of vehicle design. Legislation requires minimum levels of safety through full scale tests. Customers are provided with information regarding the safety performance of vehicles so that they can make an informed buying decision. Vehicle crashes were responsible for 40000 fatalities and 5.2 million non fatally injured patients in the US during 1994. The direct and direct cost of head injuries in the US is estimated at $25 billion per year. Injury criteria that can predict the severity of head injuries are important engineering tools for improving vehicle safety. At present the injury that the human head is subjected to is predicted by the Head Injury Criterion (HIC). This criterion is inadequate as it is not based upon a thorough understanding of the underlying head injury mechanisms. The important blunt or non-contact head injury mechanisms are diffuse axonal injury, bridging vein disruption and surface contact contusions. The severity of these injury mechanisms is hypothesised to be related to the level of motion of the brain with respect to the skull. Finite element modelling is used to analyse these head injury mechanisms. Models are developed which include all the relevant anatomical entities and detail. Accurate material property information and boundary conditions are used in the modelling to ensure that the head injury mechanisms can be accurately simulated. Tissue failure criteria are developed to link the various field parameters monitored during the simulations with injury severity. The models are then comprehensively validated with information obtained from pathological observations, cadaver experiments, accident reconstructions and volunteer data. These models are then used to determine the biomechanics of head injury and to develop improved head injury tolerance curves. The simulations demonstrate that head injury severity is dependent upon the magnitude, pulse duration and direction of the applied translational and rotational acceleration pulses.

This research contributes to the existing body of knowledge relating to crash compatibility (the minimisation of injury risk faced by all participants involved in a collision in traffic). The research focuses on the topic of structural interaction in collisions involving passenger vehicles, a phenomenon describing the efficiency of energy dissipation within existing deformation-zones of a passenger vehicle during a collision. A new definition for structural interaction was developed and several metrics to evaluate structural interaction and compatibility in car-to-car collisions were proposed, based on the commonly known Equivalent Energy Speed (EES) metric. The new EES metrics describe equivalent closing velocities for a given collision based on the energy dissipated within the front-ends (EESFF) and the entire structure (EESVV) of both vehicles involved in a head-on collision. These metrics form the basis of the new knowledge generated by this research. Additionally, a new method was developed to measure the amount of energy dissipated through structural deformation in a collision, based on accelerometer readings. This method was applied to several experimental and simulationbased car-to-car collisions and the validity of the method was proven. Based on the energy dissipation which occurred in the car-to-car collisions analysed, the degree of compatibility reached and the level of structural interaction which occurred in each collision was evaluated by applying the newly developed EESFF and EESVV metrics. Thie research also investigates the assessment of vehicles' structures in a standardized procedure with a view to improving structural interaction in the real-world. Several fixed barrier crash tests have been proposed in different configurations and with different assessment criteria. All assessments aim to evaluate the geometrical characteristics of the front-ends of passenger vehicles. A set of factors required from a compatilibility assessment focused on assessing vehicle geometry were identified. The proposed compatibility assessment procedures were evaluated based on their ability to predict the potential for structural interaction offered by passenger vehicles.

The automotive industry is actively supporting research and innovation to meet manufacturers' requirements related to safety issues, performance and environment. The Green ITS project is among the efforts in that regard. Safety is a major customer and manufacturer concern. Therefore, much effort have been directed to developing cutting-edge technologies able to assess driver status in term of alertness and suitability. In that regard, we aim to create with this thesis a framework for a context-aware driver status assessment system. Context-aware means that the machine uses background information about the driver and environmental conditions to better ascertain and understand driver status. The system also relies on multiple sensors, mainly video and audio. Using context and multi-sensor data, we need to perform multi-modal analysis and data fusion in order to infer as much knowledge as possible about the driver. Last, the project is to be continued by other students, so the system should be modular and well-documented. With this in mind, a driving simulator integrating multiple sensors was built. This simulator is a starting point for experimentation related to driver status assessment, and a prototype of software for real-time driver status assessment is integrated to the platform. To make the system context-aware, we designed a driver identification module based on audio-visual data fusion. Thus, at the beginning of driving sessions, the users are identified and background knowledge about them is loaded to better understand and analyze their behavior. A driver status assessment system was then constructed based on two different modules. The first one is for driver fatigue detection, based on an infrared camera. Fatigue is inferred via percentage of eye closure, which is the best indicator of fatigue for vision systems. The second one is a driver distraction recognition system, based on a Kinect sensor. Using body, head, and facial expressions, a fusion strategy is employed to deduce the type of distraction a driver is subject to. Of course, fatigue and distraction are only a fraction of all possible drivers' states, but these two aspects have been studied here primarily because of their dramatic impact on traffic safety. Through experimental results, we show that our system is efficient for driver identification and driver inattention detection tasks. Nevertheless, it is also very modular and could be further complemented by driver status analysis, context or additional sensor acquisition.

The automotive industry is actively supporting research and innovation to meet manufacturers' requirements related to safety issues, performance and environment. The Green ITS project is among the efforts in that regard. Safety is a major customer and manufacturer concern. Therefore, much effort have been directed to developing cutting-edge technologies able to assess driver status in term of alertness and suitability. In that regard, we aim to create with this thesis a framework for a context-aware driver status assessment system. Context-aware means that the machine uses background information about the driver and environmental conditions to better ascertain and understand driver status. The system also relies on multiple sensors, mainly video and audio. Using context and multi-sensor data, we need to perform multi-modal analysis and data fusion in order to infer as much knowledge as possible about the driver. Last, the project is to be continued by other students, so the system should be modular and well-documented. With this in mind, a driving simulator integrating multiple sensors was built. This simulator is a starting point for experimentation related to driver status assessment, and a prototype of software for real-time driver status assessment is integrated to the platform. To make the system context-aware, we designed a driver identification module based on audio-visual data fusion. Thus, at the beginning of driving sessions, the users are identified and background knowledge about them is loaded to better understand and analyze their behavior. A driver status assessment system was then constructed based on two different modules. The first one is for driver fatigue detection, based on an infrared camera. Fatigue is inferred via percentage of eye closure, which is the best indicator of fatigue for vision systems. The second one is a driver distraction recognition system, based on a Kinect sensor. Using body, head, and facial expressions, a fusion strategy is employed to deduce the type of distraction a driver is subject to. Of course, fatigue and distraction are only a fraction of all possible drivers' states, but these two aspects have been studied here primarily because of their dramatic impact on traffic safety. Through experimental results, we show that our system is efficient for driver identification and driver inattention detection tasks. Nevertheless, it is also very modular and could be further complemented by driver status analysis, context or additional sensor acquisition.

In order to prevent or mitigate the negative consequences of traffic crashes, automakers are developing active safety systems, which aim to prevent or mitigate collisions.  These systems are expensive to develop and as a result automakers and regulators are motivated to forecast the potential benefits of a proposed safety system before it is widely deployed in the vehicle fleet. The objective of this dissertation was to develop a methodology for predicting fleet-wide benefits for emerging crash avoidance systems as if all vehicles were equipped with a system.  Forward Collision Avoidance Systems (FCAS) were used as an example application of this methodology. The methodology developed for this research includes the following components: 1) identification of the target population, 2) development and validation of a driver model, 3) development of injury risk functions, 4) development of a crash severity reduction model, and 5) computation of fleet-wide benefits.  This dissertation presents a general methodology for each of these components that could be used for any active safety system.  Then a specific model is constructed for FCAS. FCAS could potentially be applicable to 31% of all collisions, 6% of serious injury crashes, and 7% of fatal crashes.  Annually, this accounts for 3.3 million collisions and 18,367 fatal crashes.  We developed a model of driver braking in response to a forward collision warning. Next we used logistic regression to develop injury risk functions that predicted the probability of injury given the crash severity ("V) and occupant characteristics.  Finally, we simulated 2,459 real-world rear-end collisions as if the driver had an FCAS with combinations of warnings, brake assist, and autonomous braking.  We found that between 3.4% and 7.2% of crashes could be prevented and that many more could be mitigated in severity.  These systems reduced the number of injured (MAIS2+) drivers in rear-end collisions between 32% and 55%.  In total, the systems could prevent between $184 and $338 million in economic costs associated with crashes per year. / Ph. D.

Road departure crashes are among the deadliest crash modes in the U.S. each year. In response, automakers have been developing lane departure active safety systems to alert drivers to impending departures. These lane departure warning (LDW) and lane departure prevention (LDP) systems have great potential to reduce the frequency and mitigate the severity of serious lane and road departure crashes. The objective of this thesis was to characterize lane and road departures to better understand the effect of systems such as LDW and LDP on single vehicle road departure crashes. The research includes the following: 1) a characterization of lane departures through analysis of normal lane keeping behavior, 2) a characterization of road departure crashes through the development and validation of a real-world crash database of road departures (NCHRP 17-43 Lite), and 3) develop enhancements to the Virginia Tech LDW U.S. fleetwide benefits model. Normal lane keeping behavior was found to vary with road characteristics such as lane width and road curvature. Consideration of the dynamic driving behaviors observed in the naturalistic driving study (NDS) data is important to avoid LDW false alarms and driver annoyance. Departure characteristics computed in normal driving were much less severe than the departure parameters measured in real-world road departure crashes. The real-world crash data collected in NCHRP 17-43 Lite database was essential in developing enhancements to the existing Virginia Tech LDW fleetwide benefits model. Replacement of regression model predictions with measured crash data and improvement of the injury criteria resulted in an 11-16% effectiveness for road departure crashes, and an 11-15% reduction in seriously injured drivers. / Master of Science / Road departure crashes account for nearly one-third of the roughly 30,000 automobile traffic fatalities in the U.S. each year. Lane departure warning (LDW) and lane departure prevention (LDP) systems are two safety systems developed to reduce the large number of fatalities resulting from road departures. The safety systems warn drivers if the vehicle begins to drift out of the intended lane of travel, and automatically steer the vehicle back into the lane of travel if it continues to drift. While LDW and LDP systems have potential to lower the number of fatal lane and road departure crashes, the technology is not yet a standard feature in production vehicles. There has been a lower than expected acceptance rate, and real-world benefits of the systems have not been published. The research objective for this thesis was to characterize lane and road departures to investigate the effect of these safety systems on road departure crashes. The first section of this thesis analyzed large amounts of time series data recorded from people in normal driving scenarios to model lane keeping behavior in non-crash, drift out of lane departures. We found driving behavior varied with road characteristics such as lane width and road curvature. These dynamic driving behaviors may lead to LDW false alarms and contribute to driver annoyance with the systems. The second portion of this research involved the development and validation of a real-world road departure crash database. The database included key departure parameters such as angle, speed, and road curvature. These parameters were used in the third section of the thesis to enhance the Virginia Tech LDW U.S. fleetwide benefits model, which is a mathematical trajectory simulation model that determines whether or not these road departure crashes could have been prevented if every vehicle in the U.S. was equipped with LDW. We found an effectiveness of 11-16% prevention for road departure crashes, and an 11-15% reduction in serious driver injury.

Motor vehicle accidents are statistically the major cause of accidental severe injuries for pregnant women and fetuses fatality. Volunteers, post mortem human surrogates, anthropomorphic crash test devices and computational occupant models are used to improve human safety in motor vehicle accidents. However, due to the ethical issues, pregnant women and their fetuses cannot be used as volunteers or post mortem human surrogates to investigate the effects of crashes on them. The only anthropomorphic test device representing pregnant women is very limited in design and lacks a fetus. There is no computational pregnant occupant model with a fetus other than 'Expecting'. This thesis focuses on understanding the risk of placental abruption for pregnant drivers involved in road accidents, hence assessing the risk to fetus fatality. An extensive review of existing models in general and pregnant women models in particular is reported. The time line of successive development of crash test dummies and their positive effect on automotive passive safety design are examined. 'Expecting', the computational pregnant occupant model with a finite element uterus and a multibody fetus, is used in this research to determine the strain levels in the uteroplacental interface. External factors, such as the effect of restraint systems and crash speeds are considered. Internal factors, such as the effect of placental location in the uterus, and the inclusion and exclusion of a fetus are investigated. The head of the multibody fetus is replaced with a deformable head model to investigate the effects of a deformable fetus head on strain levels. The computational pregnant driver model with a fetus offers a more realistic representation of the response to crash impact hence provides a useful tool to investigate fetus safety in motor vehicle accidents. Seat belt, airbag and steering wheel interact directly with the pregnant abdomen and play an important role on fetus safety in motor vehicle accidents. The results prove that the use of a three-point seat belt with the airbag offer the greatest protection to the fetus for frontal crash impacts. The model without a fetus underestimates the strain levels. The outcome of this research should assist automobile manufacturers to address the potential safety issues at the design level.

This thesis looked at traffic crashes in the emirate of Dubai in the United Arab Emirates (UAE) to establish the current situation in road safety and ways of improving it. A global overview of road safety literature revealed that standards of road safety vary widely by region. Key indicators like fatality rate and risk (Jacobs et al, 2000) were found to be higher in most neighbouring Gulf Cooperative Council (GCC) countries (10-25 fatalities/100,000 pop., 3-5 fatalities/10,000 motor vehicles) than in the best-performing Western countries (6 fatalities/100,000 pop., 1 fatality/10,000 motor vehicles). Interventions and countermeasures to tackle specific road safety issues were reviewed from international studies. Countermeasures were chosen with consideration for the local situation in Dubai within the categories of Human, Environmental and Vehicle factors. Examples of selected measures include offending driver punishment (Human), Electronic Stability Control (Vehicle) and central barriers (Environment). These measures were mostly studied in different environments to those in Dubai so the aspect of knowledge transfer between areas of different cultural and environmental conditions was discussed. Data from real world injury crashes (as collected by Dubai Police and the Roads & Transport Authority) over twelve years (1995 2006) were subject to macroanalysis in SPSS to identify the main issues over the past decade. 18,142 crashes involving 30,942 casualties and 48,960 vehicles were analysed at the outset. The following issues were among the main concerns: - High proportion of fatal crashes out of all injury crashes (13.5% compared to 1.4% in the UK); - Most fatal crashes involved a single vehicle hitting a pedestrian; - Most injury crashes involved a single vehicle; - Inconsiderate driving was the most common crash cause cited by the police. Countermeasures found in the literature to counteract these problems were then suggested for application and the estimated savings from applying them were calculated. Savings were quantified as either reductions in casualties or injury crashes. Furthermore, cost savings for the calculated reductions were estimated using existing UK crash costs due to the scarcity of UAE crash cost estimates. Calculation of the estimated improvement in safety if these countermeasures were applied retrospectively meant a reduction of 4,634 injury crashes and 1,555 casualties over the 12-year period with an estimated cost saving of approximately £368 million or 2.7 billion Dirhams. To refine this method more detailed data on crashes were required and collected from the dedicated crash investigation team files in Dubai Police for 2006 and part of 2007. This new dataset (300 crashes) was put into a purpose-built database with over 140 fields and subject to microanalysis to more accurately match the problems and interventions. Six interventions were matched to individual cases in the database where they would have positively altered the outcome. This process was verified by independent crash experts and investigators. The benefits from these six countermeasures were then weighted to calculate the benefits for the whole crash population over a year. Examples of specific interventions included guardrails along the roadside; grade-separated crossing facilities for pedestrians; Electronic Stability Control and speed cameras. The estimated total reduction in crashes was 2,412 annually with calculated savings of £40 million or 280 million Dirhams. This was the first time this geographical area was studied in such depth and detail to allow the calculation of benefits from interventions matched to known road safety issues. Various limitations were encountered such as the unavailability of GIS basemaps and the continuously changing infrastructure and population of Dubai. Numerous areas of further work were identified. Such work areas include hospital studies for collecting injury data to compare with police data; changing vehicle standards so that they are better suited to local crash types; the calculation of crash and injury costs based on local figures; vehicle fleet analysis for comparing different vehicle segments and exposure; and improved data collection and storage methods.

Existe necessidade de se desenvolver modelos teóricos e testes experimentais, que nos permitam ter plenas condições de melhor avaliar e concluir sobre o comportamento dinâmico dos ônibus, ao trafegar sobre diferentes pistas e realizar diversos tipos de manobras. O objetivo do trabalho é avaliar e otimizar simultaneamente o comportamento dinâmico lateral e vertical de um ônibus modelado como um sistema multicorpo. A metodologia utilizada no trabalho é dividida em duas partes. A primeira parte consiste na programação de um modelo multicorpo de ônibus que possa ser utilizado para fins de otimização do seu comportamento de dinâmica lateral via programação matemática; o desenvolvimento de uma manobra do tipo mudança dupla de faixa - DLC (Double Lane Change), adaptada da combinação da norma ISO 3888-1:1999 que envolve mudança dupla de faixa para carros de passeio e a norma ISO 14791:2000 que envolve mudança simples de faixa para veículos comerciais, na ausência de normas específicas; e finalmente a validação de resultados através de testes experimentais e simulações computacionais. A segunda parte consiste na programação de um modelo multicorpo de ônibus para fins de otimização do seu comportamento de dinâmica vertical via programação matemática, neste caso sujeito a uma pista da classe C segundo classificação da norma ISO 8608:1995. Os resultados específicos da programação das manobras laterais do modelo de ónibus foram validados experimentalmente, bem como comparados através da simulação das manobras num modelo virtual implementado num software multicorpo comercial. O conjunto das soluções atingidas mostraram boa correlação, possibilitando a posterior otimização dos parâmetros concentrados da suspensão do modelo multicorpo de ônibus, através da técnica de algoritmos genéticos. A função objetivo implementada consiste da composição penalizada do valor RMS do ângulo de rolagem da manobra lateral quanto ao handling, e de parâmetros associados ao conforto e segurança, como o valor RMS da aceleração vertical, do deslocamento máximo da suspensão, e da deflexão máxima do pneu de forma a garantir aderência continua à pista. Os resultados otimizados dos parâmetros concentrados conseguem uma negociação dos objetivos conflitantes. / There is a need for theoretical models and experimental tests to be developed that allow for better assessments and conclusions about the dynamic behavior of buses driving on different lanes and performing various types of maneuvers. The purpose of this work is to evaluate and optimize both the lateral and the vertical dynamic behavior of a bus modeled as a multibody system. The methodology employed comprises two parts. The first part consists in programming a bus multibody model that can be used to optimize the lateral dynamic behavior of buses via mathematical programming; developing a type of maneuver known as Double Lane Change (DLC), adapted from a combination of the ISO 3888-1:1999 standard, which involves double lane changes for passenger cars, and the ISO 14791:2000 standard, which involves single lane changes for commercial vehicles, in the absence of specific standards; and lastly, validating the results by means of experimental tests and computational simulations. The second part consists in programming a bus multibody model to optimize the vertical dynamic behavior via mathematical programming, in this case for a class C road, according to the classification of the ISO 8608:1995 standard. The specific results of the programming of the lateral maneuvers of the bus model were validated experimentally and then compared with simulations of the maneuvers by a virtual model developed using commercial multibody software. The results showed a good correlation, enabling subsequent optimization of the lumped parameters of the suspension of the bus multibody model using the genetic algorithm optimization technique. The objective function consists of the penalized composition of some terms, including the RMS value of the roll angle of the lateral handling maneuver and of parameters associated with comfort and safety, such as the RMS value of vertical acceleration, the maximum suspension working space, and the maximum tire deflection to ensure continuous adherence on the road surface. The optimized results of the lumped parameters of the suspension enable an alignment of the conflicting goals.

Over 300,000 individuals are injured every year by vehicle related accidents in the United Kingdom alone. Government and the vehicle manufacturers are not only bringing new legislation but are also investing in vehicle safety research to bring this figure down. A private self-driven car is an important factor in maintaining the independence and quality of life of the third age individuals. However, since older people brings deterioration of cognitive, physical and visual abilities, resulting in slower reaction times and lapses while driving. The third age individuals are involved in more vehicle related accidents than middle aged individuals. This scenario is corrected by the fact that the number of third age individuals is increasing, especially in developed countries. It is expected that the percentage of third age individuals in the United Kingdom will increase to 20% of the total population by 2010. Several safety systems have been developed by the automotive industry including intelligent airbags, Electronic Stability Control (ESC) and pre-tensioned seat belts, but nothing has been specifically developed for the third age related problems. This thesis proposes a driver posture identification system using low resolution infrared imaging. The use of a low resolution thermal imager provides a reliable non-contact based posture identification system at a relatively low cost and is shown to provide robust performance over a wide range of conditions. The low resolution also protects the privacy of the driver. In order to develop the proposed safety system an Artificial Intelligent Thermal Imaging algorithm (AITl) is created in MatLAB. Experimentation is conducted in real and simulated environment, with human subjects, to evaluate the results of the algorithm. The result shows that the safety system is able to identify eighteen different driving postures. The system also provides other valuable information about the driver such as driver physical built, fatigue, smoking, mobile phone usage, eye-height and trunk stability. It is clear that in incorporating this safety system in the overall automotive central strategy, better safety for third age individual can be achieved. This thesis provides various contributions to knowledge including a novel neural network design, a safety system using low resolution infrared imager and an algorithm that can identify driver posture.