Dissertations / Theses on the topic 'Pre-deployment'

Thesis (M.S. in Defense Analysis)--Naval Postgraduate School, June2010.
Thesis Advisor(s): Giordano, Frank R. Second Reader: Wilson, Gregory R. "June 2010." Description based on title screen as viewed on June, 17, 2010. Author(s) subject terms: Pre-Deployment; Training; DMETL; CMETL; Task; Influence; Constraint; Afghanistan. Includes bibliographical references (p. 113-115). Also available in print.

Approved for public release; distribution is unlimited
In this thesis, the author evaluated whether HOMER Micropower Optimization should be used by the Marine Corps as a pre-deployment tool for meeting expeditionary energy demands. The author created two unique experiments to facilitate the evaluation of HOMER's modeling capability. First, a grid-tied-photovoltaic (PV) system at the Naval Postgraduate School was monitored for a one-month period. During this experiment, a HOMER model of the system was created. The actual energy production from the system was compared to the model. Then, the model was calibrated to the particular system to ensure that the model's energy estimate matched that of the actual system. The second experiment involved the use of two different types of PV panels and a small wind turbine. Each system was monitored over a one-month period, and the results were compared to a HOMER model of the systems. The difficulty of modeling wind turbines and the related limitations of HOMER's modeling strategy is discussed in this thesis. The calibration method established in the grid-tied-PV experiment was used to ensure the HOMER models were accurate. Following the calibration, the concept of expeditionary energy density as it pertains to power production was defined and utilized to evaluate each of the systems. The final portion of this thesis shows the advantage of using HOMER as part of the Experimental Forward Operating Base (ExFOB). The ExFOB was conducted by the Marine Corps to evaluate alternative power solutions currently on the market for expeditionary energy purposes. Four distinct power production solutions were chosen by the Marine Corps following the ExFOB. These solutions were then field tested in Morocco and scheduled to be deployed to Afghanistan. This thesis details how the use of HOMER would have benefited the ExFOB process had it been utilized.

Purpose: Describe the perceptions of military registered nurses (RNs) on being prepared to provide nursing care during a deployment. Background: Injuries sustained by war are different from trauma-related injuries occurring within the Unites States. Nurses who provide care during overseas deployment encounter patients with poly-trauma, multiple and highly complex injuries; consequently, this type of nursing requires strong clinical skills beyond what is required in stateside facilities. Additionally, military nurses undergo intense stress related to overseas deployment in a war zone. In fact, military medical providers have one of the highest rates of post-traumatic stress post-deployment. Methods: This project employed a qualitative, case study approach. Semi-structured interviews were conducted with military RNs who worked as nurses during overseas deployment. A script was developed to guide the interview, and further discussion stemmed from participant responses. Recorded interviews were transcribed into text and analyzed for commonalities. Findings: Five military nurses who previously provided nursing care during overseas employment participated in this study. Commonalities included the realization that higher acuity injuries are seen in a deployed setting versus a military stateside hospital. Participants also shared fears that they would not be prepared enough to provide excellent patient care. Another commonality was not knowing how to prepare for an unknown experience. A positive commonality was the sense of pride expressed by the participants regarding their military service and deployment experience. All participants outlined what they had learned and what might have helped them to better prepare for deployment. Unanimously, more clinically relevant training was recommended. The participants shared that they had grown and changed from their deployment experience. Implications: All five participants felt a gap in perceived adequate pre-deployment preparation. They believed they lacked the training to care for the high acuity patients they would encounter during overseas deployment; this shook their self-confidence and caused them to worry about being an effective team member. Military preparedness programs should be expanded to include skills and knowledge relevant to nursing in high acuity, hostile environments. Research looking at the personal preparations, specifically mental and emotional, of military members may be helpful in determining any links between mental resilience and the development of PTSD.

The purpose of this thesis was to evaluate the impact of dynamic message signs (DMS) on the Orlando-Orange County Expressway Authority (OOCEA) toll road network using the Pre-Deployment DMS Survey (henceforth referred to as "pre-deployment survey"). DMS are electronic traffic signs used on roadways to give travelers information about travel times, traffic congestion, accidents, disabled vehicles, AMBER alerts, and special events. The particular DMS referred to in this study are large rectangular signs installed over the travel lanes and these are not the portable trailer mount signs. The OOCEA is currently in the process of adding several fixed DMS on their toll road network. Between January 2007 and February 2008, approximately 30 DMS are planned on their network. It is important to note that there was one DMS sign on the OOCEA network before this study started. Since most of the travelers on OOCEA toll roads are from Orange, Osceola and Seminole counties, this study is limited to these counties. This thesis documents the results of pre-deployment analysis. The instrument used to analyze the travelers' perception of DMS was a survey that utilized computer aided telephone interviews. The pre-deployment survey was conducted during early November of 2006. Questions pertaining to the acknowledgement of DMS on the OOCEA toll roads, satisfaction with travel information provided on the network, formatting of the messages, satisfaction with different types of messages, diversion questions (Revealed and Stated preferences), and classification/socioeconomic questions (such as age, education, most used toll road, and county of residence) were asked to the respondents. The results of the pre-deployment analysis showed that 54.4% of the OOCEA travelers recalled seeing DMS on the network. The respondents commonly agreed that the DMS are helpful for providing information about hazardous conditions, and that the DMS are easy to read. The majority of the travelers preferred DMS formats as a steady message for normal traffic conditions, and use of commonly recognized abbreviations such as I-Drive for International Drive. The results from the binary logit model for "satisfaction with travel information provided on OOCEA toll road network" display the significant variables that explain the likelihood of the traveler being satisfied. The results from the coefficients show that infrequent travelers are more likely to be satisfied with traveler information on OOCEA toll roads. In addition, the provision of hazard warnings, special event information, and accuracy of information on DMS are associated with higher levels of satisfaction with traveler information. The binary logit model for "Revealed Preference (RP)" diversion behavior showed that Seminole County travelers were likely to stay on the toll road, and SR 408 travelers were likely to divert off the toll road. The travelers who acknowledged DMS on the OOCEA network were also likely to divert off the toll road, but those who learned of the congestion by DMS were likely to stay on the toll road. Learning of congestion by DMS could encourage travelers to stay, since when they are on the toll roads, diversion at times could be difficult with no access to exits or little knowledge of alternate routes. But it is also possible that travelers stayed because their perception was that the toll roads are faster, especially when messages on DMS show travel times that confirm the travelers' belief. Travelers who were not satisfied with travel information on the network were more likely to divert off the toll road. The implications for implementation of these results are discussed in this thesis. DMS should be formatted as a steady message for normal traffic conditions. Commonly recognized abbreviations, such as I-Drive for International Drive, must be used for roadway identification when possible. DMS messages should be pertained to information on roadway hazards when necessary because it was found that travelers find it important to be informed on events that are related to their personal safety. Accuracy of information provided on DMS was important for traveler information satisfaction because if the travelers observe inaccurate travel times on DMS, they may not trust the validity of future messages. DMS information that led to the travelers canceling their intended stops led to a higher likelihood of them being dissatisfied with traveler information. It is important to meet the travelers' preferences and concerns for DMS.
M.S.
Department of Civil and Environmental Engineering
Engineering and Computer Science
Civil Engineering MS

The aim of the study is to examine how well MSB (Swedish Civil Contingencies Agency) prepare the Swedish humanitarian aid workers to face stressful situations, looking at the pre- deployment preparation received. The study investigates how effective and relevant the pre- deployment preparation is in terms of building resilience against stress, according to the humanitarian aid workers’ experience. This study has been conducted with a quantitative online survey, combined with a qualitative open- ended survey. The surveys were based on research models on work-related stress and on resilience- building among humanitarian aid workers. The frameworks point out criteria needed to be fulfilled by an organization in order to build sufficient stress- resilience among humanitarian aid workers in the pre- deployment phase. The results reveal that MSB is able to fulfill most of the criteria set in accordance with the theoretical framework. The one and only criterion MSB fails in providing is an open, in- depth discussion about mental health before the aid worker is deployed. This also corresponds to the aid workers experience of not being provided with this. The aid workers’ experience reveals that resilience against stress is important, however, many seem to build resilience independently from MSB.

Initial research on moral injury has shown the construct to be associated with many negative mental health outcomes such as depression, suicidal ideation, reduced sense of belonging, anxiety, anger, and spiritual distress. In addition, moral injury among service members has also been shown to be related to experiences of interpersonal violence occurring pre-deployment. Given that social support has consistently been found to play a key role in moderating the relationship between experiences of interpersonal violence and stress and trauma related symptoms and disorders, this study will be the first to explore the relationship between experiences of interpersonal violence and moral injury, and the moderating role of social support. This study will use stepwise regression analysis to examine secondary data collected from 935 U.S. military personnel primarily from the Army National Guard.
MS

The purpose of this study is to examine the role and impact that uniformed women play in UN peacekeeping operations, and to further establish how appropriate pre-deployment training (PDT) supports the performance of women in operational zones. The research questions posed are ‘whether women make a difference to peacekeeping operations’ and ‘to what extent PDT enables them to do so’. To answer these questions, the thesis takes a two-pronged approach. Firstly, a literature search evaluates the nature of uniformed women’s participation in UN peacekeeping missions, their contribution to effective peacekeeping and the UN policies supporting women’s participation in its missions. The main resources accessed for the literature research are the UN and contributing countries’ official policies, publication and reports. Secondly, primary data were acquired through field research on the training needs of three Southeast Asian countries, namely Indonesia, the Philippines and Country A. Across these sample states, empirical research data was gathered through in-depth semi-structured interviews with 37 female peacekeepers, 17 trainers and seven decision-makers. The literature reveals that women participate in UN peacekeeping missions in two ways, those that form part of a contingent and others that act as individual military experts, observers or staff officers. Women make a difference by allowing a UN mission to have greater reach to the local community, especially to the female population in segregated communities, including the survivors of sexual- and gender-based violence. The field research also reveals that the three Southeast Asian country case studies provide different PDT to their personnel, although the UN provides standardised training materials. Interview data from all three countries indicate that women and men receive combined PDT training, with the majority of the respondents arguing that there is no need for segregated gender training. However, they endorse differentiated training for specialist skills, such as for mentoring teams by same sex members to discuss biological and logistical issues in deployment, including, for instance, the best strategy for ensuring continuity in the supply of women’s sanitary requirements. Not all the three sample countries support uniformed women deployment on par with male peacekeeper deployment, and rarely support women holding leadership positions, due to discrimination in military education access, limitations on human resources and apprehension at putting women into dangerous positions. Such constraints limit the roles that women can play in UN field missions.

The purpose of this thesis was to evaluate the impact of dynamic message signs (DMS) on the Orlando-Orange County Expressway Authority (OOCEA) toll road network using a Pre and Post-Deployment DMS Survey (henceforth referred to as "pre and post-deployment survey") analysis. DMS are electronic traffic signs used on roadways to give travelers information about travel times, traffic congestion, accidents, disabled vehicles, AMBER alerts, and special events. The particular DMS referred to in this study are large rectangular signs installed over the travel lanes and these are not the portable trailer mount signs. The OOCEA have been working over the past two years to add several fixed DMS on their toll road network. At the time of the pre-deployment survey, only one DMS was installed on the OOCEA toll road network. At the time of the post-deployment survey, a total of 30 DMS were up and running on the OOCEA toll road network. Since most of the travelers on the OOCEA toll roads are from Orange, Osceola, and Seminole counties, this study was limited to these counties. This thesis documents the results and comparisons between the pre and post-deployment survey analysis. The instrument used to analyze the travelers' perception of DMS was a survey that utilized computer aided telephone interviews. The pre-deployment survey was conducted during early November of 2006, and the post-deployment survey was conducted during the month of May, 2008. Questions pertaining to the acknowledgement of DMS on the OOCEA toll roads, satisfaction with travel information provided on the network, formatting of the messages, satisfaction with different types of messages, diversion questions (Revealed and Stated preferences), and classification/socioeconomic questions (such as age, education, most traveled toll road, county of residence, and length of residency) were asked to the respondents. The results of both the pre and post-deployment surveys are discussed in this thesis, but it should be noted that the more telling results are those of the post-deployment survey. The results of the post-deployment survey show the complete picture of the impact of DMS on travelers' experience on the OOCEA toll road network. The pre-deployment results are included to show an increase or decrease in certain aspects of travel experience with relation to DMS. The results of the pre-deployment analysis showed that 54.4% of the OOCEA travelers recalled seeing DMS on the network, while a total of 63.93% of the OOCEA travelers recalled seeing DMS during the post-deployment analysis. This showed an increase of almost 10% between the two surveys demonstrating the people are becoming more aware of DMS on the OOCEA toll road network. The respondents commonly agreed that the DMS were helpful for providing information about hazardous conditions, and that the DMS are easy to read. Also, upon further research it was found that between the pre and post-deployment surveys the travelers' satisfaction with special event information provided on DMS and travel time accuracy on DMS increased significantly. With respect to formatting of the DMS, the following methods were preferred by the majority of respondents in both the pre and post-deployment surveys: • Steady Message as a default DMS message format • Flashing Message for abnormal traffic information (94% of respondents would like to be notified of abnormal traffic information) • State road number to show which roadway (for Colonial – SR 50, Semoran – SR 436 and Alafaya – SR 434) • "I-Drive" is a good abbreviation for International Drive • If the distance to the international airport is shown on a DMS it thought to be the distance to the airport exit The results from the binary logit model for "satisfaction with travel information provided on OOCEA toll road network" displayed the significant variables that explained the likelihood of the traveler being satisfied. This satisfaction model was based on respondents who showed a prior knowledge of DMS on OOCEA toll roads. With the use of a pooled model (satisfaction model with a total of 1775 responses – 816 from pre-deployment and 959 from post-deployment), it was shown that there was no statistical change between the pre and post-deployment satisfaction based on variables thought to be theoretically relevant. The results from the comparison between the pre and post-deployment satisfaction models showed that many of the coefficients of the variables showed a significant change. Although some of the variables were statistically insignificant in one of the two survey model results: Either the pre or post-deployment model, it was still shown that every variable was significant in at least one of the two models. The coefficient for the variable corresponding to DMS accuracy showed a significantly lower value in the post-deployment model. The coefficient for the variable "DMS was helpful for providing special event information" showed a significantly higher value in the post-deployment model. The final post-deployment diversion model was based on a total of 732 responses who answered that they had experienced congestion in the past 6 months. Based on this final post-deployment diversion model, travelers who had stated that their most frequently traveled toll road was either SR 408 or SR 417 were more likely to divert. Also, travelers who stated that they would divert in the case of abnormal travel times displayed on DMS or stated that a DMS influenced their response to congestion showed a higher likelihood of diversion. These two variables were added between the pre and post-deployment surveys. It is also beneficial to note that travelers who stated they would divert in a fictitious congestion situation of at least 30 minutes of delay were more likely to divert. This shows that they do not contradict themselves in their responses to Revealed Preference and Stated Preference diversion situations. Based on a comparison between pre and post-deployment models containing similar variables, commuters were more likely to stay on the toll road everything else being equal to the base case. Also, it was shown that in the post-deployment model the respondents traveling on SR 408 and SR 417 were more likely to divert, but in the pre-deployment model only the respondents traveling on SR 408 were more likely to divert. This is an expected result since during the pre-deployment survey only one DMS was located on SR 408, and during the post-deployment survey there were DMS located on all toll roads. Also, an interesting result to be noted is that in the post-deployment survey, commuters who paid tolls with E-pass were more likely to stay on the toll road than commuters who paid tolls with cash. The implications for implementation of these results are discussed in this thesis. DMS should be formatted as a flashing message for abnormal traffic situations and the state road number should be used to identify a roadway. DMS messages should pertain to information on roadway hazards when necessary because it was found that travelers find it important to be informed on events that are related to their personal safety. The travel time accuracy on DMS was shown to be significant for traveler information satisfaction because if the travelers observe inaccurate travel times on DMS, they may not trust the validity of future messages. Finally, it is important to meet the travelers' preferences and concerns for DMS.
M.S.C.E.
Department of Civil and Environmental Engineering
Engineering and Computer Science
Civil Engineering MS

Potential risk factors for development of Post-Traumatic Stress Disorder (PTSD) are still unclear. One potential risk factor for the development of PTSD is an individual’s cardiovascular reactivity and recovery in response to stressor tasks. The current study was conducted with 763 Army National Guard and Army Reserve soldiers. Participants completed a stressful induction along with self-report measures prior to deployment. Post-deployment, self-report measures were completed to assess PTSD symptomatology and experiences related to deployment and combat. Multiple regression was used to determine the ability of blood pressure response to stress to predict PTSD symptoms immediately and one-year after return from deployment. Results indicated that soldiers who had a less reactive systolic blood pressure response to and recovery from stressor tasks reported more PTSD symptomatology immediately after and one year after return from deployment. These results suggest that soldiers who develop PTSD after deployment have less pre-deployment emotion regulation ability.

The automotive industry is repeatedly tasked with improving vehicles structural strength, optimizing active and passive safety features, reducing occupant injury potential, and hitting lean manufacturing goals. The challenge is to find solutions to reduce production and research costs, and to maximize the vehicle's capability in protecting its occupants in the event of an accident. The influence of impact loading on the dynamic responses of vehicle structures and occupants require special consideration in the field of vehicle crashworthiness. The main goal of this study is to address the fundamental aspect on the impact injury biomechanics of vehicle occupants and safety performance of vehicle structures, and the development of various new technologies aimed at enhancing the passive safety of road vehicles in side-impact accidents. Four case studies related to the dissipation of crash energy, impact injury biomechanics, injury prediction model and pre-crash sensing algorithm form the basis of this thesis. The application examples include the investigation of pre-deploying airbags as a potential solution in reducing occupants' injuries at higher speed side-impact crashes; examination of the vehicle structural responses with the inclusion of high-energy absorbing cellular materials within the door panels in side-impact accidents; development of injury prediction model to out-of-position occupants from frontal- and side- airbags using Design-of-Experiment methodologies; and the estimation of the relative driver fatality risks of two colliding vehicles using some quantitative measurements. A detailed methodology is developed for each application, and the results present several new technologies that can be implemented to enhance the safety performance of road vehicles. These goals are achieved through the use of finite element approaches, multi-body dynamic analyses and Design-of-Experiment statistical methods.
Thesis (M.S.)--Wichita State University, College of Engineering, Dept. of Mechanical Engineering

Le Fog Computing décentralise le Cloud en rapprochant les services de calcul, de stockage et de réseau de la périphérie du réseau. Cette approche réduit la latence, l’utilisation de la bande passante et améliore le traitement en temps réel. Cependant, la complexité et l’hétérogénéité des systèmes Fog rendent leur gestion difficile et coûteuse. Identifier les erreurs en phase d’exécution nécessite souvent de revenir aux étapes de conception, ce qui entraîne des coûts élevés en temps et en ressources. La vérification pré-déploiement est donc essentielle pour garantir fiabilité et efficacité. Les systèmes Fog, utilisés dans des domaines variés (santé, automobile, villes intelligentes), ajoutent de la complexité aux processus de vérification et de déploiement. Pour y répondre, nous proposons une approche générique et personnalisable basée sur un processus de vérification en deux étapes, combinant les phases de conception et de pré-déploiement, tout en automatisant les activités de vérification et de déploiement. Notre solution repose sur un langage de modélisation Fog adaptable, la vérification des propriétés non fonctionnelles à la conception (sécurité, énergie), et la génération de configurations de déploiement indépendantes des outils. La vérification pré-déploiement associe outils de déploiement et solutions QoS pour garantir la conformité aux critères définis avant le déploiement final. Validée par trois cas d’usage (campus intelligent, parking intelligent, hôpital intelligent), cette approche réduit les coûts, simplifie la gestion des systèmes Fog et garantit leur QoS. En intégrant des pratiques DevOps, elle répond aux exigences des environnements industriels et académiques modernes
Fog Computing decentralizes the Cloud by bringing computation, storage, and network services closer to the edge. This reduces latency and bandwidth usage while improving real-time processing. However, the complexity and heterogeneity of Fog systems, often comprising diverse entities, make lifecycle management challenging and costly. Runtime error handling frequently requires revisiting earlier phases, which is both timeconsuming and expensive. Ensuring reliability through pre-deployment verification is therefore essential. Fog systems are deployed in domains such as healthcare, automotive, and smart cities, further complicating verification and deployment processes. To address these challenges, we propose a generic and customizable approach based on a two-step verification process. This approach focuses on the design-time and pre-deployment phases, automating key verification and deployment activities. Our solution features a customizable Fog modeling language, design-time verification of non-functional properties (e.g., security, energy), preparation of pre-deployment configurations, and integration with industrial DevOps tools and Quality of Service (QoS) solutions. By combining Model-Based Engineering and DevOps practices, our approach ensures QoS, reduces deployment costs, and enhances automation to tackle the complexity of Fog systems. We validated this approach using three literature-based use cases—smart campus, smart parking, and smart hospital. Results demonstrate its effectiveness in QoS verification, deployment automation, and reducing complexity and costs, highlighting its relevance to state-of-the-art engineering and DevOps practices

Pre-deployment field testing in is the process of testing software to uncover unforeseen problems before it is released in the market. It is commonly conducted by recruiting users to experiment with the software in as natural setting as possible. Information regarding the software is then sent to the developers as logs. Log data helps developers fix bugs and better understand the user behaviors so they can refine functionality to user needs. More importantly, logs contain specific problems as well as call traces that can be used by developers to trace its origins. However, developers focus their analysis on post-deployment data such as bug reports and CVS data to resolve problems, which has the disadvantage of releasing software before it can be optimized. Therefore, more techniques are needed to harness field testing data to reduce post deployment problems. We propose techniques to process log data generated by users in order to resolve problems in the application before its deployment. We introduce a metric system to predict the user perceived quality in software if it were to be released into market in its current state. We also provide visualization techniques which can identify the state of problems and patterns of problem interaction with users that provide insight into solving the problems. The visualization techniques can also be extended to determine the point of origin of a problem, to resolve it more efficiently. Additionally, we devise a method to determine the priority of reported problems. The results generated from the case studies on mobile software applications. The metric results showed a strong ability predict the number of reported bugs in the software after its release. The visualization techniques uncovered problem patterns that provided insight to developers to the relationship between problems and users themselves. Our analysis on the characteristics of problems determined the highest priority problems and their distribution among users.
Thesis (Master, Electrical & Computer Engineering) -- Queen's University, 2009-09-16 17:50:31.094

The severity of cyber threats towards existing and future industrial systems has resulted in an increase of security awareness in the industrial automation domain. Compared to traditional information systems, industrial communication systems have different performance and reliability requirements. The safety and availability requirements can also sometimes conflict with the system security design of plants. For instance, it is not acceptable to create a secure system that may take up additional time to establish security and as a consequence disrupt the production in plants. Similarly, a system that requires authentication and authorization procedures before an emergency action may not be suitable in industrial plants. On the other hand, lack of security can hamper safety of a plant. Therefore, there is a need for improvement of the security workflow in industrial plants, so that the practical realization of security can be achieved. This includes secure device deployment and secure data communication inside the industrial plants. Furthermore, the industrial plant networks are heterogeneous in terms of hardware, software, and protocols. This complicates security assessment of industrial networks.   In this thesis, the focus is on achieving a secured communication infrastructure for heterogeneous industrial networks. The initial trust establishment is the starting point for enabling a secure communication infrastructure. A framework for the initial trust establishment for industrial devices that can support key management using the existing trust of employees in a plant is proposed. With the help of a proof-of-concept implementation and security analysis, it is shown that the proposed framework is feasible to implement and satisfies the security objectives. After establishing initial trust within industrial devices, assessing heterogeneous security properties based on the network architecture is another focus of this thesis. A model to estimate the security assurance of nodes in a heterogeneous network, where all devices are not having the same level of security mechanisms, is given. Along with cyber security requirements of industrial plants, it is also necessary to consider other important requirements of plants in terms of network performance. In this thesis, identification of an optimized path between two systems in a heterogeneous network in terms of the network performance and the network security is explored. The applicability of this balancing approach has been demonstrated in a specific case of smart grid application where security, network capacity and reachability need to be optimal for successful network operation.

The rapid development of clinical artificial intelligence, AI, technologies has outpaced the development of robust regulatory and clinical safety mechanisms. AI systems are cleared for use and deployed in practice relying on pre-clinical performance studies, without evidence of the impact this will have on patient and provider outcomes. This has led to concerns of an, implementation gap, where systems that appear to perform well on pre-clinical testing fail to produce the expected outcomes in practice. While there is an urgent need for direct clinical testing of AI systems and evaluation of the impact of these systems on patient and provider outcomes, it is implausible to expect the clinical evaluation will be performed at the scale necessary to mitigate potential AI harms of the many AI systems already in use and currently under development. In this body of work I look at factors which may contribute to the implementation gap, in particular the effects of low-quality training and testing data, flawed and incomplete study design methodologies, and an over-reliance on explainability methods to address safety. I suggest a series of improvements to how we design, evaluate, and utilise AI systems in clinical practice, with the goal of better estimating the potential harms of AI during the pre-clinical testing phase, and by doing so closing the implementation gap.
Thesis (Ph.D.) -- University of Adelaide, School of Public Health, 2022