Search

The advent of Intelligent Transport Systems (ITS) has facilitated a shift towards proactive safety measures in which a crash occurrence is anticipated and prevented before it happens under the proactive safety approach. Given the capability to identify pre-crash conditions, real-time crash prediction has become widely examined with the availability of disaggregated real-time traffic data. In spite of this, real-time crash prediction models have yet to be introduced to the market due to ongoing concerns pertaining to the models' predictability, transferability, and explainability. Existing real-time crash prediction studies examined crash prediction as a classification task. Turbulent traffic dynamics imminently before crashes are used to classify between crash cases and more stable traffic dynamics in normal traffic conditions (i.e., non-crash cases). Matched case-control sampling is the most commonplace methodology employed for real-time crash prediction models. Crash cases are predicted from several non-crash cases sampled for controls. Due to their simplicity in application, various statistical and machine learning models have utilised this methodology in real-time crash prediction which attained satisfactory predictability. Yet, non-crash cases are under-utilised due to matched case-control sampling. Further limitations of this approach are low ecological validity and sampling bias. To overcome the shortcomings from matched case-control sampling, studies have begun to investigate using the full dataset for crash prediction. Using the full dataset supports a data-driven analysis, which is in-line with the advocate from ITS. Furthermore, machine learning and deep learning models leverage the power of Big Data and produce more accurate prediction results. However, crashes are rare occurrence. The compiled dataset for real-time crash prediction is naturally imbalanced with crash cases being the minority. Predicting an imbalanced dataset would give rise to an undesirable 'high accuracy, low sensitivity' circumstance in that non-crash cases are predicted whilst none or very little crash cases can be predicted. Multiple studies have attempted to tackle class imbalance through data sampling, in which a Synthetic Minority Oversampling Technique (SMOTE) has been adopted, to oversample crash cases by generating synthetic crashes from its nearest neighbours. Despite the successful application of SMOTE, it has been criticised with overgeneralisation and prone to overfitting. As such, the emergence of deep generative models, artificial intelligence (AI) models such as Variational Encoder (VAEs) and Generative Adversarial Network (GAN) can generate high quality synthetic data. This thesis has compiled a heavily imbalanced dataset consisting of 257 crash cases and 10 million non-crash cases along the M1 Motorway between Junction 1 and Junction 30 in the United Kingdom for 2017. The dataset is aggregated in 5-minute intervals from minute-level real-time traffic data. 195 variables were calculated containing information about flow, speed, occupancy and headway. To ameliorate class imbalance, this thesis adopted the Wasserstein Generative Adversarial Network (WGAN) to generate high quality synthetic data to oversample crash cases. The class balancing performance is compared with other oversampling methods such as SMOTE, Adaptive Synthetic sampling (ADASYN) and class balancing methods such as cost-sensitive learning and ensemble methods. Deep Neural Network (DNN) has been leveraged for its ability to learn features thoroughly in a non-linear manner. The model is trained on a sample of 500,000 non-crash cases due to computational limitations. The effects of oversampling methods (i.e., WGAN, SMOTE, ADASYN) are compared with crash predictions performed under different crash to non-crash ratios, whereas comparison with other class balancing methods has been made in the balanced dataset. The issue of class imbalance alleviates when additional synthetic data are entered in the training set at ratio of 1:10 onwards. Between class balancing methods, WGAN is proved to outperform the others with superior AUC values. Model predictability of DNN is also compared with Logistic Regression at a balanced data setting. The optimal model attained is trained with DNN with data oversampled by WGAN with the area under the curve (AUC) value of 0.86, sensitivity with 0.63 at a 7% false alarm rate under the ratio of 1:2. Apart from data imbalance, model transferability and explainability are two other major limitations that prohibit the practical deployment of real-time crash prediction models. Transferability of a model refers to its ability to generalise the model from one setting to another temporal, spatial or spatio-temporal settings. Current studies demonstrated limited transferability with models directly tested on other settings under matched case-control sampling approach. In this thesis, five other heavily imbalanced datasets collected along different motorways (i.e., M1, M4 and M6) in 2017 and 2018 were compiled for transferability assessment. Transfer learning is utilised which shared the weights trained from the baseline model developed from the M1 2017 dataset and applied to five other datasets to be transferred. The performance of transfer learning is compared to direct transfer and the standalone models developed from the transferred datasets. The transferability assessment revealed that direct transfer is not possible whereas transfer learning can improve model transferability with the AUC ranging between 0.69 to 0.95. The best transferred model can predict 89% of crashes accurately at a 7% false alarm rate. Model explainability has not been a challenge when statistical models are employed as they are inherently interpretable with parameter estimates provided. However, model explainability become a significant concern when neural networks are employed. These models are described as "black-box" models of which the decision function of the model is not human interpretable. This thesis addresses model explainability through employing Shapley Additive explanation (SHAP) and Counterfactual Explanations (CE) to explain predicted crash cases from the best DNN model developed from the M1 2017 dataset. SHAP is used to explain why a crash is predicted as a crash and CE is used to demonstrate how can the predicted crash be altered as a non-crash. Explanations from CE is compared with SHAP summary plot, consistent results were found that aggregated flow and the standard deviation of flow played an important part in crash prediction. To potentially reduce crash likelihood, the aggregated flow should be reduced with standard deviation of flow increased. These suggestions are also in-line with what literature suggested. Application of SHAP and CE can demystify the "black-box" limitation from the AI models which traffic managers can trust and act up on the model predictions. This PhD thesis has made methodological advancements in current limitations in real-time crash prediction models in handling class imbalance, improving model transferability and explainability. Hence, this study gives rise to a shift from predictive to a pragmatic approach for real-time crash prediction models that can be readily deployed in the future.

Objective: Medicine administration is the last process of the medication cycle. However, errors can happen during this process. Children are at an increased risk from these errors. This has been extensively investigated but evidence is lacking on effective interventions. Therefore, the aim of this research is to propose safety measures to reduce medication administration errors (MAE) in the Paediatric Intensive Care Unit (PICU). Method: The research was carried out over five studies; 1) systematic literature review, 2) national survey of PICU medication error interventions, 3) retrospective analysis of medication error incidents, 4) prospective observation of the administration practice, and 5) survey of PICU healthcare professionals' opinions on MAE contributory factors and safety measures. Results: Hospital MAE in children found in literature accounted for a mean of 50% of all reported medication error reports (n= 12552). It was also identified in a mean of 29% of doses observed (n= 8894). This study found MAE retrospectively in 43% of all medication incidents (n= 412). Additionally, a total of 269 MAEs were observed (32% per dose observation). The characteristics of the interventions used to reduce MAE are diverse but it illustrated that a single approach is not enough. Also for an intervention to be a success it is fundamental to build a safety culture. This is achieved by developing a culture of collaborative learning from errors without assigning blame. Furthermore, MAE contributing factors were found to include; interruptions, inadequate resources, working conditions and no pre-prepared infusions. The following safety measures were proposed to reduce MAE; 1) dose banding, 2) improved lighting conditions, 3) decision support tool with calculation aid, 4) use of pre-prepared infusions, 5) enhance the double-checking process, 6) medicine administration checklist, and 7) an intolerant culture to interruption. Conclusion: This is one of the first comprehensive study of to explore MAE in PICU from different perspectives. The aim and objectives of the research were fulfilled. Future research includes the need to implement the proposed safety measures and evaluate them in practice.

A small percentage of wheelchair users are unable to transfer from their wheelchair to a vehicle during transportation. Reasons for an occupant to remain in the wheelchair during transport may be the inability to safely transfer to a vehicle seat, the occupant's requirement of a specialist postural management wheelchair seating system or reliance on life support equipment attached to the wheelchair. The Rehabilitation Engineering Unit at Rookwood Hospital deal with people who require either a specialist postural support wheelchair seating system, life support equipment or both. To cater for such equipment the wheelchairs have to be modified to some degree and sometimes completely custom made. In performing modifications to the wheelchairs the Rehabilitation Engineering Unit take on the manufactures responsibilities, one of which is to ensure that the wheelchair is safe for use in transport. Standard crash tests for production wheelchairs are destructive so are impractical to use for bespoke wheelchair designs meaning that the Clinical Engineers at the Hospital have to rely on their best engineering judgement as to whether a wheelchair design is crash worthy or not. It was proposed that by using computer crash simulation techniques an informed judgement of the crashworthiness of the bespoke wheelchair designs could be attained. A series of computer models of occupied wheelchairs were created and validated against physical crash data performed on surrogate wheelchairs. These validated wheelchair computer models were then used to examine a series of different crash scenarios that provided the Clinical Engineers at Rookwood hospital with an informed process for virtually assessing the crashworthiness of their wheelchair designs. The validation results showed that the wheelchair crashworthiness could feasibly be predicted by computer simulation. This thesis concluded that attaching equipment to the wheelchair can increase both its horizontal displacement and the forces on the tiedowns securing the wheelchair to the vehicle chassis. Skewed impact simulations also highlighted the poor lateral restraint ability of the 4-point webbing tie-down system and also the importance of sufficient lateral support on the wheelchair for occupant protection.

A deficient safety culture has been implicated in a number of organisational accidents from a range of high hazard industries. Despite its implications for safety, many questions about safety culture remain unanswered. In order to contribute to the literature on this topic, this thesis set out to investigate the role of trust in safety culture. The oil and gas industry was chosen as the context for study due to the hazardous nature of its work, the industry’s focus on continuous improvement in safety performance and the interest shown by oil companies in participating in safety research. Leading models of safety culture have stressed the importance of trust in developing and maintaining patterns of safe behaviours at work. This thesis proposed a new model of safety culture based on dual attitudes about trust. This model states that explicit attitudes about trust are part of safety climate and that implicit attitudes about trust comprise some of the basic underlying assumptions that are the deepest level of safety culture. In order to test this model, this thesis developed a method to measure implicit attitudes about trust in an industrial setting. Using this method, two studies of dual attitudes about trust were conducted at different UK gas plants. In both of these studies, different patterns of results were found for measures of explicit and implicit attitudes about trust for workmates, supervisors and the plant leadership, respectively. These findings support the proposed model suggest that explicit and implicit attitudes about trust are separate constructs that may influence different types of safety behaviours.  Positive relationships were found between measures of explicit attitudes about trust and self-report items about safety behaviours like reporting incidents and challenging unsafe acts.  These findings were taken as a step toward validating the proposed model.

Safety and Health in the workplace has long been the priority work of the Hong Kong administration but the accidentstatistics in Hong Kong tell another story. No matter how sophisticated a safe system is designed, its ultimate success depends very much on the person who carries out the job. Safe behaviour has therefore become the contemporary study of safety and health at work.Since human behaviour is a multidimensional construct, its understanding requires a multiple theory approach. Inspired by this concept, this study explores a Multi-Dimensional Safe Behaviour Model in explaining Construction Workers' Safe Behaviour. The study examines its implication for management when safe behaviours are to be instilled. Nine psychological theories and models, identified under the perspective of the Intrapersonal, Interpersonal and Community Level are examined. A short list of 9 variables of "Social Norm", "Management Commitment", "Safety Knowledge", "Perceived Risk", "Safety Experience", "Self Efficacy"; "Perceived Consequence", "Chance" and Intention to Behave" was constructed. Three hypothetical constructs of "Social Support", "Attitude" and "Expectance" composed of observable indicators from the nine identified variables are also formed. The variables are then put together into a hypothesised Multi-Dimensional Safe Behaviour Model with the casual relationships between variables identified. A Structural Equation Modelling procedure shows that the hypothesised Multi-Dimensional Safe Behaviour Model fits the data reflecting the necessity of adopting a holistic approach in addressing behavioural issues. "Safety Attitude" is found to impose a positive effect on the worker's "Intention to Behave Safely" indirectly via a mediating factor of "Chance Locus". The study conveys practical implications to safety management and researchers. Research limitations and areas for further study are also discussed. Safety behavioural initiatives based on the model testing results in promotion workers' safety behaviours are also addressed.

The repair and maintenance (R&M) of buildings are becoming increasingly important, given the significant growth of existing building stock. The safety performance of R&M works has become alarmingly low. Building R&M works are characterised by aspects such as the need to work in confined workspaces, at a height, in hard-to-access places and in occupied work environments; the changing nature of work tasks and scope; the need to adhere to tight working schedules; and the use of sophisticated technologies, which create a dynamic and complex work arrangement for building R&M workers. Traditional safety management systems (SMSs) have played a key role in reducing safety incidents at workplaces; however, these systems may be insufficient to deal with the varying and unforeseen safety risks emerging because of the inherent complexity associated with building R&M works. The main reason for this problem is that these classical systems stress on minimising uncertainty—as opposed to coping with uncertainty—because they consider safety issues to derive from deviations that must be removed. In contrast, resilience engineering, which is identified as a complementary to the conventional SMS, assists in coping with the uncertainty caused by the complexity in socio-technical systems. Therefore, resilience must be engineered into SMSs to address the safety risks that occur owing to the complex nature of R&M works. To this end, this study aims to develop resilience for the SMSs in building R&M companies. The specific objectives are (1) to identify the dimensions of resilience for SMSs in building R&M companies, (2) to identify the drivers of resilience for SMSs and (3) to examine the impact of resilience for SMSs on the safety performance of building R&M companies. To achieve these research objectives, this study developed 16 research hypotheses based on the existing body of knowledge. A quantitative research approach and a survey research design were employed to test the research hypotheses. Data were collected through a questionnaire survey with building R&M professionals from 145 Australian building R&M contracting companies. Using the analysis results, this study identifies three dimensions of resilience for the SMSs of building R&M companies: people, place and system resilience. This study establishes that resilience for SMSs has a positive impact on the safety performance of building R&M companies.

In order to evaluate the safety of radioactive waste disposal underground it is important to understand the way in which radioactive material is transported through the rock to the surface. If the rock is fractured the usual models may not be applicable. In this work we look at three aspects of fracture networks: connectivity, flow and transport. These are studied numerically by generating fracture networks in a computer and modelling the processes which occur. Connectivity relates to percolation theory, and critical densities for fracture systems are found in two and three dimensions. The permeability of two-dimensional networks is studied. The way that permeability depends on fracture density, network size and spread of fracture length can be predicted using a cut lattice model. Transport through the fracture network by convection through the fractures and mixing at the intersections is studied. The Fickian dispersion equation does not describe the resulting hydrodynamic dispersion. Extensions to the techniques to three dimensions and to include other processes are discussed.

Use of roadside safety barriers has greatly enhanced highway safety and reduced the severity of traffic accidents and injuries. Generally speaking, safety barriers should be sufficiently designed to contain and redirect the vehicle away from its errant travel without causing any severe injuries to vehicle occupants and road pedestrians. A significant body of literature already exists on designing roadside safety barriers to achieve high standards of safety to road users, but the literature, in general, does not provide clear insights on the effects of vehicular impact on the foundation of the barrier systems. Alhough the 2D plane-strain approach is commonly used for the stability design of the retaining structures including barrier systems, none of the design codes in Australia, USA and Europe provide adequate guidance to convert the vehicular impact loading to an equivalent plane-strain loading. The consequence of this is that there is a lot of guesswork without much rational basis currently being applied within the geotechnical community to analyse vehicular impact loading on foundations. This thesis is concerned with a finite element study of a vehicle crash against the safety barrier system and its impact loading on the system foundation. The possible combinations of vehicle-barrier system crash are extremely large, but this thesis is only focused on the combination of a 44t special performance vehicle crashing against the barrier crash of a 3m high reinforced concrete integral barrier-wall system. The integral barrier-wall is a type of system where the barrier is fully integrated with and located at the top of the retaining wall. Moreover, a 44t special performance vehicle is the largest class of vehicle by weight to which the safety barrier is to be designed against based on current design codes. This class of safety barrier is known as the “44t special performance level” barrier and it is the highest class specified in AS5100.2:2017. It has been chosen for this study because of the potential catastrophic effects of such a crash not just on the vehicle and its occupants, but also on the integrity of the foundation of the barrier-system. One of the main reasons numerical modelling is such an important tool in vehicle-barrier crash study is because full-scale physical crashes are very expensive to conduct, and the larger the crash the more costly will be the test. Very few full-scale tests have actually been conducted for the vehicle-barrier crash of the 44t special performance level. The lack of real test data is also compounded by the fact that vehicle-barrier crash tests were mainly concerned with the performance of the safety barriers, and paid little heed to the performance of the barrier system foundation. Therefore, the main goal of this thesis is to develop a 3D finite element model of the 44t special performance vehicle-barrier crash which impacts a 3m high integral barrier-wall system. The model was created and analysed using Abaqus/Standard and Abaqus/Explicit software. It was calibrated against the design impact loading for a 44t special performance vehicle specified in AS5100.2:2017, and the calibrated model was then used to perform further numerical simulations to investigate the foundation responses due to the impact loading. This study has defined foundation responses as: (1) mobilised normal reaction (2) mobilised moment reaction (3) mobilised shear resistance of the foundation at the foundation-soil interface. Based on these analyses, an effective length corresponding to the length to which the impact loading of the vehicle-barrier crash has dispersed along the foundation is established. The effective length is then used to calculate the equivalent 2D plane-strain loading to apply in the stability design of the 3 m integral barrier-wall system. In addition, a sensitivity study was carried out to assess the mobilised effects on the foundation by varying the impact angle and impact velocity. It is found that the effective length is only marginally sensitive to these variations. Hence it may be surmised that the recommended effective length and 2D-plane strain loadings are reasonably robust.

In the global view, among other rail safety issues, highway-railway accidents continue to be a major problem from both public health and socio-economic perspectives. It is noted that many research studies have been conducted in the past in relation to developing appropriate models to assess the road traffic safety through collision prediction, but a considerable amount of work has been carried out only regarding safety at “highway-railway” grade crossings. The primary objective of this study is to provide an improved method for rail safety appraisal at “railway-highway” grade crossings through the development and application of suitable safety risk scores (called ‘Safety Risk Index’) with a combination of both accident frequency and accidental consequences prediction models generated for crossings, and also by using these safety risk scores to identify the worst or most dangerous locations. The Safety Risk Index (SRI) is a simple composite index, which can measure, compare and rank safety levels at different risk situations and locations. These safety risk scores are designed to generate an overall grade crossings safety risk, which is based on the combination of three basic risk elements - namely the exposure of the crossing users, the probability of an accident taking place, and the severity of consequences should an accident occur. This method facilitates the assessment of the safety risks at grade crossings and also ranks, identifies and prioritises the worst performing crossings or the problematic crossings (Black-Spots). This model is very simple and easily understood by those with different levels of knowledge on safety. The SRI index based on quantitative methods and developed in this study seems very promising and has great potential to be a major tool for safety risk assessment at grade crossings in various countries. The secondary objective of this study is to provide an index with a single meaningful value for assessing the risk at grade crossings, through the gathering and analysis of data, information and knowledge (from various data sources) on rail safety. The research study also establishes appropriate statistical methodologies in order to develop and to construct a quantitative model for risk assessment.

Patient safety climate is defined as a holistic snapshot of enacted work environment practices and procedures related to patient safety, derived from shared perceptions of social and environmental work characteristics. While patient safety climate has been touted as a critical factor underlying safe patient care, our understanding of input factors influencing shared climate perceptions and, in turn, the effects of climate as a collective, group-level construct on important outcomes remains underdeveloped, both theoretically and empirically. Therefore, the current study examines (1) the antecedents that impact individual patient safety climate perceptions and (2) the relationships between hospital unit patient safety climate and two important unit level outcomes: patient willingness to recommend a facility to others and patient safety. This study also examines climate strength--the degree to which climate perceptions are shared--as a moderator of these relationships. While climate is conceptualized as a holistic description of the working environment, existing evidence has focused on relationships between the independent dimensions of patient safety climate and patient safety. No study to date has examined the configurations (i.e. patterns or profiles) among the multiple dimensions of patient safety climate or how these configurations are related to important employee and patient outcomes. This gap is redressed in the current study by examining patient safety climate in terms of three profile characteristics: (1) climate elevation (i.e., mean positive or negative valence across all dimensions), (2) climate variability (i.e., variance among dimensions), and (3) climate shape (i.e., the pattern of peaks and valleys among climate dimensions). Evidence from studies of general organizational climate suggests that the shape of the pattern among climate dimensions, the overall mean score across dimensions, and the degree to which dimension scores vary are predictive of employee attitudes, customer satisfaction, and organizational financial performance (Dickson et al., 2006; Joyce & Slocum, 1984; Jackofsky & Slocum, 1988; Gonzalez-Roma, Peiro, & Zornoza, 1999; Litwin & Stringer, 1968; Schulte et al., 2009). The current study, then, tests a theoretical model of patient safety climate examining the configural nature of the construct. An archival dataset collected from seven hospitals located in a metropolitan area of the southeastern United States was utilized to test study hypotheses. Data was collected from 3,149 individuals nested within 84 hospital units using the Hospital Survey on Patient Safety Culture (Sorra & Nieva, 2004). Unit level patient safety and patient willingness to recommend was collected by the hospital risk management and nursing administration departments. Hierarchical linear modeling (HLM7; Raudenbush, Bryk, Cheong, Congdon, & du Toit, 2011) was utilized to test hypotheses regarding antecedents of individual level perceptions of patient safety climate to account for the fact that individuals were nested within hospital units. Traditional multiple regression analyses were utilized to test unit level hypotheses examining the relationships between unit level patient safety climate and patient outcomes. Results indicated that unit membership was significantly related to individual climate perceptions--specifically, individual-level climate profile elevation. In turn, individual climate profile elevation and profile variability were related to employee willingness to recommend their organization to family and friends in need of care. At the unit level of analysis, climate profile variability was significantly related to patient willingness to recommend the organization to others, and climate shape was found to be related to patient safety. Furthermore, these results were not dependent on climate strength. The current study meaningfully contributes to the conceptual understanding of the patient safety climate construct by examining the degree to which configural aspects of the construct are predictive of important outcomes across multiple levels of analysis. In this way, it extends beyond existing studies of climate configurations to examine relationships at multiple levels of analysis and to also examine the moderating effects of climate strength. Practically, results provide insight into how the construct of patient safety climate can be used diagnostically and prescriptively to improve patient care and the working environment for providers. In addition to contributing to the theoretical understanding of the patient safety climate construct, this study also augments the evidence-base available to administrators, front-line providers, and regulators regarding how patient safety climate can be used to guide and align quality improvement efforts for greatest impact.

While good progress has been made in the reduction of fatalities and serious incidents leading to injury in the electricity distribution industry in New South Wales, a further reduction seems difficult to achieve. This thesis investigates the influence that the organisational culture, and in particular the safety culture, has had on efforts to further reduce safety incidents in the electricity distribution industry in New South Wales. This industry is concerned with providing electricity throughout New South Wales through a network of wires carried on poles and also underground cables. Frontline workers construct and maintain the distribution system, operating from various field centres throughout the state in a diverse environment with risks including communication difficulties due to remote locations and adverse weather conditions. In addition to the safety risks associated with manual handling and operating plant and machinery experienced by many people in workplaces, this industry has hazards associated with working with, or physically close to electricity. Historically a number of fatalities and serious injuries have occurred to frontline workers coming into contact with energised conductors and equipment, and falling from heights. The frontline workers have a raft of safety legislation, guidelines and procedures to enable them to work safely in the industry and a strict training and qualification regime. In spite of this there have occurred over the study period a number of serious incidents resulting in injury and fatalities to front line workers. Transcripts of prosecutions in the industry from 1993 to 2007 reveal that the safety incidents were avoidable. The analysis of the transcripts indicate that a number of failures in the management of safety have occurred leading to the injury, or death, of front line workers. The missing factor in the management of safety is revealed as the lack in most cases of a positive safety culture. This is illustrated by the recognition by one of the three distributors (Country Energy) of the importance of organisational safety culture and the efforts by them to improve their safety performance. A study of the research that has been conducted by various authors including Hopkins (2005) and Reason (1997) in the discipline of organisational safety culture indicates that improvement can be made to the safety of front line workers by organisations creating a positive safety culture. The creation of a safety culture will require a greater focus upon total commitment and active involvement at all levels of the organisations concerned, from the Board of Directors to the frontline workers. This would entail a genuine consultation process in the design and implementation of the cultural change program. Attitude changes in management from the ‘blame the victim’ approach evident in the transcripts of prosecutions to that of seeking to identify all causal factors in safety incidents and rectifying hazards identified. In this way further reduction in safety incidents can be achieved in the industry.

Enforcement strategies utilized by the civil aviation safety regulator in Australia have to date, received scant attention from researchers. This thesis is a first attempt to look at such enforcement strategies and apply them to practical situations. It does so in an all embracing manner, taking into account the part played by aviation's historical origins, both international and domestic, that have led to the complex set of aviation safety regulations now in existence. It pays particular attention to the historical importance of Australia's former unique two-airline policy and the constitutional problems facing the early lawmakers dealing with a constitution written before the advent of aviation. It describes and examines regulatory theory as it relates to such enforcement strategies and attempts to bring together such theory with practice by applying it to practical examples. Finally it analyses how certain pressures and influences may have swayed the type of enforcement strategies pursued by the regulator at various times as these are evidenced in parliamentary inquiries and reported cases in the years since economic de-regulation of the aviation industry. The thesis puts forward the hypothesis that strategies of enforcement employed by the civil aviation regulatory authority have a profound effect on aviation safety. It contends that evidence drawn from inquiries and reported cases over a twenty-year period point to a regulator who prefers a compliance strategy of cooperation with industry, rather than a sanctioning or deterrence strategy. The thesis also emphasizes that the evidence drawn from various inquiries and coroner's inquests points to the fact that such a strategy, if it is lacking or seen to be lacking in adequate checks and balances, can pose an increased risk to aviation safety. Emphasis is placed on the need for the aviation safety regulator to be at all times cognizant of the dangers that are inherent in a compliance based enforcement strategy, especially where the regulator puts itself in 'partnership' with industry. A number of topics have been identified for further research. They include the need for a greater comparative analysis of enforcement strategies used by regulators in similarly aligned countries to Australia, the concept of regulatory 'capture' and a comparative inquiry into the sometimes ambiguous dual role of 'educator' and 'policeman' expected of the regulator.

Many types of foods are sterilised and packaged aseptically to maximise their shelf life and ensure that they are safe for consumption. However, if even a few micro-organisms survive these treatments, the quality and/or safety of the food can be seriously compromised. Therefore, packaging materials that can actively participate in the process of inactivating micro-organisms have a potentially important role in protecting the quality and safety of packaged foods. The aim of this project is to explore the effectiveness of the photodynamic effect as a method of inactivating micro-organisms on the surface of packaging materials. Photodynamic action occurs when a photosensitiser molecule absorbs light in the presence of oxygen, leading to the formation of reactive oxygen species such as singlet oxygen, superoxide radicals, or other free radicals; these species are known to inactivate cells. Medical uses of photodynamic action have been studied extensively, usually in applications where the reactive oxygen species are produced intracellularly. However, very little research has been conducted to investigate the efficacy of extra-cellularly generated reactive oxygen species on the viability of micro-organisms, and accordingly this research project investigated the possibility that this approach might be used for inactivating micro-organisms on packaging materials. The results of this study indicate that singlet oxygen may be produced at the surface of polymer films that contain selected photosensitisers, resulting in the oxidation of known singlet-oxygen acceptors to form endoperoxides. This provides evidence that singlet-oxygen mediated reactions can occur at the surface of a sensitising film. It is likely that cell destruction is caused by singlet-oxygen, rather than involving other activated species. It was further shown that the photodynamically generated singlet-oxygen can inactivate micro-organisms (extra-cellularly) on the surface of a polymeric material. The study included examples from each genus of micro-organisms that are of concern to the food and packaging industry, including: Gram-negative bacteria (Escherichia coli); Gram-positive bacteria (Bacillus cereus), both vegetative cells and endospores; yeast (Saccharomyces cerevisiae), and mould (Fusarium oxysporium.) Results indicate that the photodynamic effect causes a substantial reduction in viable cell numbers for vegetative cells and spores (both bacterial and fungal) that have been inoculated onto a plastic surface containing the photosensitiser, anthraquinone,. The results show that an increase in the amount of reactive oxygen species produced by photodynamic action increases the inactivation rate of the micro-organisms. The micro-organisms investigated were susceptible to photodynamic action to varying extents. In conclusion, the results demonstrate that control of microbial populations on the surface of polymeric films (used in food packaging) is achievable using photodynamic action produced from polymers, based on the reported amounts of micro-organisms found on food packaging materials.

Remote area staff performing manual patient handling tasks in the absence of patient lifting hoists available in most health care settings are at an elevated risk of musculoskeletal injuries. The objective of this project was to identify the patient handling methods that have the lowest risk of injury. The patient handling task of lifting a patient from floor to a chair or wheelchair is a common task performed in a remote health care setting. The task was performed utilising three methods, these being: (1) heads/tails lift, (2) use of two Blue MEDesign® slings and (3) use of a drawsheet. The task of the heads/tails lift was broken down into two distinctly separate subtasks: lifting from the (1) head and (2) tail ends of the patient load. These techniques were selected based on criteria including current practice, durability, portability, accessibility, ease of storage and cost to supply. Postural data were obtained using a Vicon 370 three - dimensional motion measurement and analysis system in the Biomechanics & Gait laboratory at the University of New South Wales. Forty reflective markers were placed on the subject to obtain the following joint angles: ankle, knee, hip, torso, shoulder, elbow, and wrist. The raw data were converted into the respective joint angles (Y, X, Z) for further analysis. The postural data was analysed using the University of Michigan’s Three-Dimensional Static Strength Prediction Program (3D SSPP) and the relative risk of injury was based on the following three values: (1) a threshold value of 3,400 N for compression force, (2) a threshold value of 500 N for shear force, and (3) population strength capability data. The effects on changes to the anthropometric data was estimated and analysed using the in-built anthropometric data contained within the 3D SSPP program for 6 separate lifter scenarios, these being male and female 5th, 50th and 95th percentiles. Changes to the patient load were estimated and analysed using the same computer software. Estimated compressive and shear forces were found to be lower with the drawsheet and tail component of the heads/tails lift in comparison to the use of the Blue MEDesign® straps and head component of the heads/tails lift. The results obtained for the strength capability aspect of each of the lifts indicated a higher percentage of the population capable of both the drawsheet and tail end of the heads/tails lift. The relative risk of back injury for the lifters is distributed more evenly with the drawsheet lift as opposed to the heads/tails (tail) lift where risk is disproportionate with the heavier end being lifted. The use of lifter anthropometrics does not appear to be a realistic variable to base assumptions on which group of the population are capable of safely performing this task in a remote setting. This study advocates the use of the drawsheet lift in a remote setting based on the author’s experience and the biomechanical results obtained in this study. The drawsheet lift is both more accessible and provides a more acceptable risk when more than two patient handlers are involved, in comparison to the other lifts utilised lifting patients from floor to a chair.

This study investigates the research on current self escape practices, the need for training on emergency evacuation procedures and the use of virtual reality simulations for the purposes of training. Thus, this project focuses on the advantages of using virtual reality as an educational tool and follows thedevelopment and design aspects of the tool. It includes each phase in development of the tool, including current research on self escape practices, training needs analyses and the actual design phases of the simulation. The design phase focuses on the objects and animations required to be implemented in order to create a realistic virtual mine. The programming logic of the simulation tool is also examined. Moreover, the final simulation is evaluated and proved to be an appropriate form of assessment and efficient learning environment for the trainee. Furthermore, recommendations are made for future developments.

Over the centuries, the assessment of risk has become an integral part of the decision process. Assessment techniques have developed to meet different applications, but all have problems and none is entirely suited to the assessment of risks relating to fire. This dissertation examines the development of risk assessment processes and frameworks, identifying common features and problems and key differences in approach. Despite generically similar approaches, different applications have led to the development of many different paradigms, none of which appears to be entirely suitable for application to building and infrastructure fires. Current fire risk assessment methods which incorporate important advances in fire modelling and Monte Carlo simulation, rely on a fire engineering approach. They tend to consider only the limited range of fire safety systems that are directly involved in construction, failing to address many of the procedural and other activities that can overwhelm traditional controls, and taking insufficient account of interactions between different controls and the factors that influence them. Further, comparative risk levels are generally evaluated against the ill-defined scenario of current practice, as defined in outdated prescriptive regulations. The result is that catastrophic consequences continue to occur, despite the presence of traditionally accepted controls. The problem is to find a framework that evaluates the sensitivity of levels of risk in fire against a defined, uncontrolled state, taking into account the effects of a comprehensive range of factors and controls. A new approach to risk assessment that addresses a comprehensive set of factors and controls and evaluates the event without, and with, controls, is considered. The framework, together with the steps for its implementation, appears to provide a versatile and flexible method of risk assessment. It is likely that the framework can be applied to all risk assessment situations. A study is undertaken to investigate the impact of factors and interactions that are not commonly taken into account in fire risk assessment. The chosen situation is a fire in the driver’s cab of a train. Current driver procedures are examined, and fire growth rates for specified materials are considered. Using the fire spread model CFAST, conditions in the cab for a range of ventilation conditions and fire growth rates are calculated. Threshold levels that determine response times for engineered and human controls and tenability, and common factors that influence consequences, can play a critical role in modelling the decision process. A driver’s decision model is proposed that determines the impact of the driver’s decisions to adjust ventilation by opening or closing windows and doors, and to extinguish the fire. The model takes into account time to respond and time to perform the necessary activities. The study shows that, even with a limited choice of actions, the decisions of the driver can have a critical effect on the outcome of a fire in the driver’s cab, altering the situation from a controlled to an uncontrolled state. Recommendations are made for further development of the new risk assessment framework, and for the development of fire modelling for risk assessment purposes. Finally, recommendations are made for the continuation of the development of the train driver response model that would result in the generation of driver decision support software.

Safe, high-quality health care requires an open, transparent, and just culture where people are willing and have the opportunity to discuss errors and system problems and to do something about them. There is paucity of research in relation to safety issues in health care. Objective: To identify the components of a safe culture and measure staff’s attitudes to those components in an area health service. Method: A mixed mode method comprising qualitative and quantitative measures was used. A 60-item survey comprising a likert response scale and measuring safety attitudes and values was administrated to 3,200 staff in an Area Health Service (AHS) in Western Sydney. Focus groups were conducted with various professional groups to provide greater depth to salient issues identified in the survey. Results: The response rate was 26%. Teamwork, communication, leadership, and performance shaping factors were identified as components of a safe working health service and that staff were committed to safety. The results show that there is a positive safety and teamwork culture. However, the results about organisational culture were not positive. Results also showed that there is a high level of stress factors that influence safety in the organisation. There were a number of attitudes and patterns that warranted improvement and further investigation. However several areas of discontent among staff in terms of leadership and communication were acknowledged. In regard to attitudes suggesting invulnerability to the effects of stress and fatigue, the majority of respondents acknowledged this belief. These findings are more favourable than those earlier studies using the Operating Room Management Attitudes Questionnaire (ORMAQ), although these were conducted in other countries and the respondents were surgeons, anaesthetists, and Intensive Care Unit staff. A number of factors have emerged both from the questionnaire survey and the follow up focus groups that should be considered. The awareness of the need to encourage teamwork, communication, leadership prevailed. Conclusion: Health care organisations can benefit from the valuable information and an understanding of staff’s attitudes to patient safety. It confirms the findings of other studies that attitudes to leadership, teamwork and performance shaping factors such as fatigue and stress have an impact on an organisation’s capacity to develop a culture of safety. Health care organisations need to acquire this kind of information, and build on it if they are considering developing and implementing patient safety programs.

In this thesis, the author has drawn on a large body of international research and his own surveys, interviews and experience to examine the interrelated issues of Fatigue and Stress in the long distance sector of the Australian road transport industry. Two major Australian studies are studied in some detail: House of Representatives Standing Committee on Communications, Transport and the Arts: Beyond the Midnight Oil, Report into an Inquiry into Managing Fatigue in Transport, Parliament of the Commonwealth of Australia October 2000 (The Neville Report) and Quinlan, Michael, Report of an Inquiry into Safety in the Long Haul Sector of the Road Trucking Industry, Motor Accident Authority of NSW, Nov 2001 and the United States of America: Belzer, Michael H., Sweatshop on Wheels: Winners and Losers in Trucking Deregulation, Oxford University Press, New York, 2000, has direct relevance to this study, especially as it relates to the underlying causes of much of the fatigue and stress suffered by long distance drivers. An extensive literature review was undertaken to help form an understanding of the issues and developments relating to workplace fatigue and stress, in several different settings, including the military, aviation and general industry. The author has surveyed a number of long distance drivers and interviewed managers and others linked to this industry, analysed their responses and reported on the findings. Since the commencement of this study, several significant changes have occurred in the relevant legislation and its application, and where possible these have been incorporated into the study

Recent concern on increased levels of accidents to children falling from play equipment and the seriousness of these accidents, has led to the need for more research into the area of resilient undersurfacing materials available in New Zealand for playground surfacing. The objective of this dissertation is to install a range of surface materials available here in New Zealand. Evaluations will then be carried out on factors such as - ease of installation, ease of availability and cost. Unfortunately it will not be possible to cover factors such as resilience and durability in this dissertation. These will be included in evaluations to be carried out at a later date.

This thesis "System Safety in Rail Transportation," is addressed to an individual having a basic technical background but little or no experience in this field. The thesis discusses the need for and the benefits to be obtained by using system safety techniques and principles in the railroad industry. Examples of typical railroad accidents are reviewed, and it is pointed out that analysis of the hazards in the railroad industry prior to the accidents would have identified problems which eventually resulted in the accidents. The system safety approach, which was developed for use in the aerospace and aviation fields, has proved to be extremely effective and is now being adapted to many other areas. The surface modes of transportation have the greatest need for these techniques. The techniques covered in this thesis include Hazard Analysis, Fault or Logic Tree Analysis, Failure Modes and Effects Analysis, and Probabilistic Cost Analysis. The thesis also describes a hypothetical model for organizing and implementing system safety approaches in an existing railroad company.

A digital computer model has been developed to simulate the kinetic response of a thermal pressurized water reactor power plant. The program is capable of predicting core power and plant temperature variations which result from disturbances due to routine demand changes, control rod movement, and fission product poison transients as well as from transients during plausible accident situations. The development of the differential equations describing the influence of various primary plant components on the dynamics of the reactor have been heretofore documented in the literature. Finite difference equations facilitating the employment of a digital computer solution are fully derived. Confidence in program validity is supported by the simulation of previously studied accidents and comparison with the safety analyses of a licensed nuclear power generating plant. A source listing of the computer program is provided in the Appendix.

This thesis, "A Hypothetical Model for System Safety in Air Transportation," is addressed to an individual having a basic technical background and some familiarity with the air transportation system in the United States. The thesis identifies the need for the benefits to be obtained from a system-side program of safety activities. The organizational framework for implementation of a system safety program is presented as well as techniques of safety analysis.