17628 research outputs found
Sort by
Hierarchical ensemble deep learning for data-driven lead time prediction
Get PDFThis paper focuses on data-driven prediction of lead times for product orders based on the real-time production state captured at the arrival instants of orders in make-to-order production environments. In particular, we consider a sophisticated manufacturing system where a large number of measurements about the production state are available (e.g. sensor data). In response to this complex prediction challenge, we present a novel ensemble hierarchical deep learning algorithm comprised of three deep neural networks. One of these networks acts as a generalist, while the other two function as specialists for different products. Hierarchical ensemble methods have previously been successfully utilised in addressing various multi-class classification problems. In this paper, we extend this approach to encompass the regression task of lead time prediction. We demonstrate the suitability of our algorithm in two separate case studies. The first case study uses one of the largest manufacturing datasets available, the Bosch production line dataset. The second case study uses synthetic datasets generated from a reliability-based model of a multi-product, make-to-order production system, inspired by the Bosch production line. In both case studies, we demonstrate that our algorithm provides high-accuracy predictions and significantly outperforms selected benchmarks including the single deep neural network. Moreover, we find that prediction accuracy is significantly higher in the synthetic dataset, which suggests that there is complexity (i.e. subtle interactions) in industrial manufacturing processes that are not easily reproduced in artificial model
Do non-contact injuries occur during high-speed running in elite football? Preliminary results from a new GPS and video-based method
Get PDFObjectivesUnderstanding how injuries occur (inciting circumstances) is useful for developing etiological hypotheses and prevention strategies that can be tested. The aims of this study were 1) to evaluate the feasibility of a method combining video and GPS data to estimate the speed and acceleration of activities leading to injuries; 2) to use this method to analyse the inciting circumstances leading to non-contact injuries in football players.DesignRetrospective descriptive study.MethodsData collected from 46 elite players over three seasons are included. Training and matches were video recorded and external load measures were collected through Catapult Vector S7 GPS. Injury nciting circumstances were analysed through GPS measures and visual inspection.ResultsIn total 34 non-contact injuries were analysed. Sixteen out of the seventeen hamstring injuries occurred when players were running for (median and IQR) 16.75 m (8.42–26.65 m), achieved a peak speed of 29.28 km·h−1 (26.61–31.13 km·h−1) which corresponded to 87.55% of players' maximal speed (78.5% - 89.75%). Of the three adductor injuries, one occurred while the player was decelerating without the ball, one injury occurred while the player was accelerating and controlling the ball at knee level, and one injury occurred while the player was performing an instep kick. Two quadriceps injuries occurred while the players were kicking either while walking or while running.ConclusionsFrom the preliminary results reported in this study most hamstring injuries occurred when players ran > 25 km·h−1 and above 80% of their maximal speed. This study suggests that this novel approach can allow a detailed and standardised analysis of injury inciting circumstances
Federated Learning for IoT Intrusion Detection
Get PDFThe number of Internet of Things (IoT) devices has increased considerably in the past few years, resulting in a large growth of cyber attacks on IoT infrastructure. As part of a defense in depth approach to cybersecurity, intrusion detection systems (IDSs) have acquired a key role in attempting to detect malicious activities efficiently. Most modern approaches to IDS in IoT are based on machine learning (ML) techniques. The majority of these are centralized, which implies the sharing of data from source devices to a central server for classification. This presents potentially crucial issues related to privacy of user data as well as challenges in data transfers due to their volumes. In this article, we evaluate the use of federated learning (FL) as a method to implement intrusion detection in IoT environments. FL is an alternative, distributed method to centralized ML models, which has seen a surge of interest in IoT intrusion detection recently. In our implementation, we evaluate FL using a shallow artificial neural network (ANN) as the shared model and federated averaging (FedAvg) as the aggregation algorithm. The experiments are completed on the ToN_IoT and CICIDS2017 datasets in binary and multiclass classification. Classification is performed by the distributed devices using their own data. No sharing of data occurs among participants, maintaining data privacy. When compared against a centralized approach, results have shown that a collaborative FL IDS can be an efficient alternative, in terms of accuracy, precision, recall and F1-score, making it a viable option as an IoT IDS. Additionally, with these results as baseline, we have evaluated alternative aggregation algorithms, namely FedAvgM, FedAdam and FedAdagrad, in the same setting by using the Flower FL framework. The results from the evaluation show that, in our scenario, FedAvg and FedAvgM tend to perform better compared to the two adaptive algorithms, FedAdam and FedAdagrad
Promoting human rights in mental healthcare: beyond the ‘Geneva impasse’
Get PDFThe past decade has seen a significant growth in attention to the human rights of persons with disabilities, taken to include mental health conditions. Consequently, challenges to important areas of current psychiatric practice have emerged, with which the profession has, in general, shown limited engagement
Sludge-derived biochar: Physicochemical characteristics for environmental remediation
No full textThe global production of fecal wastes is envisioned to reach a very high tonnage by 2030. Perilous handling and consequential exposition of human and animal fecal matter are inextricably linked with stunted growth, enteric diseases, inadequate cognitive skills, and zoonoses. Sludge treatment from sewage and water treatment processes accounts for a very high proportion of overall operational expenditure. Straightforward carbonization of sludges to generate biochar adsorbents or catalysts fosters a circular economy, curtailing sludge processing outlay. Biochars, carbonaceous substances synthesized via the thermochemical transformation of biomass, possess very high porosity, cation exchange capacity, specific surface area, and active functional sorption sites making them very effective as multifaceted adsorbents, promoting a negative carbon emission technology. By customizing the processing parameters and biomass feedstock, engineered biochars possess discrete physicochemical characteristics that engender greater efficaciousness for adsorbing various contaminants. This review provides explicit insight into the characteristics, environmental impact considerations, and SWOT analysis of different sludges (drinking water, fecal, and raw sewage sludge) and the contemporary biochar production, modification, characterization techniques, and physicochemical characteristics, factors influencing the properties of biochars derived from the aforestated sludges, along with the designing of chemical reactors involved in biochar production. This paper also manifests a state-of-the-art discussion of the utilization of sludge-derived biochars for the eviction of toxic metal ions, organic compounds, microplastics, toxic gases, vermicomposting approaches, and soil amelioration with an emphasis on biochar recyclability, reutilization, and toxicity. The practicability of scaling up biochar generation with multifaceted, application-accustomed functionalities should be explored to aggrandize socio-economic merits
‘Our bodies are not strong anymore’: a focus group study of health risk perception of ambient air pollution near a petrochemical industry
Get PDFBackground. Ambient air pollution has persisted in less-endowed communities, resulting in exposure to unhealthy pollutants. Epidemiological studies on air pollution have been mainly quantitative, with a dearth of information on community health risk perception, a key component of risk management. Objectives. The aim of this focus group study was to illuminate the health risk perception of ambient air pollution among persons residing near a petrochemical industry. In addition, determine their perception of existing control measures and ideas for more effective control. Methods. Participants were purposively selected based on age, sex, long-term residence near a petroleum refinery and occupation. Three 90-min face-to-face focus groups and one individual interview were conducted. The moderator guided discussions using a pre-formed topic guide. Discussions were audio-recorded, transcribed manually and coded using NVivo software. Data analysis was conducted using reflective thematic analysis. Results. Six themes were generated namely, Negative perception of the environment, the refinery is to blame, Air pollution is seen or smelt, Air pollution is associated with health and non-health risks, Poor response to air pollution: everyone is to blame, and Government is primarily responsible for healthy air quality. The participants were not aware of the extent of air pollution health risks. Suggestions for air pollution control included regulating gas flaring, environmental health education, and incentives for community members. Conclusions. Participants perceived that their ambient air was unhealthy; however, concern about the health risks was shaped by contextual factors. The key barriers to effective mitigation were poor environmental health literacy and political factors
Experimental investigation on thermomechanical properties and micro-machinability of carbon nanofibre reinforced epoxy nanocomposites
Get PDFA comprehensive experimental investigation on thermomechanical properties and micro-machinability of carbon nanofibre reinforced epoxy nanocomposites (EP/CNF) is presented in this study. The machinability indicators including cutting force and surface roughness have been investigated. Tensile properties, morphology of tensile fracture surfaces, glass transition temperature, machined chip morphology, and machined surface morphology were also characterised. To investigate the effect of both workpiece material properties and operating conditions on the machinability of EP/CNF, three controlled quantitative factors were selected at different levels, namely CNF loading, cutting speed and feed per tooth (FPT). Micromilling experiments were performed on an ultra-precision desktop micro-machine tool using titanium‑carbon-nitride (TiCN) coated micro-end mills. Among all compositions with CNF concentration ranging from 0.3 to 1 wt%, EP/1 wt% CNF exhibited the best machinability among other nanocomposites with its lowest cutting force of approximately 0.5 N and surface roughness of 0.18 μm. Size effect appeared at FPT below minimum uncut chip thickness (MUCT) indicated by the strong deterioration of surface quality owing to the dominant ploughing effect
Moral distress in midwifery practice: A Delphi study
No full textBackgroundMoral distress is a psychological concept that describes the harm associated with actions or inactions that oppose an individuals’ moral beliefs. Moral distress is linked to moral compromise in the workplace that may negatively impact mental wellbeing. Current tools available to assess moral distress are not specific for the Australian health care system or midwifery practice.AimThe aim of this study was to develop a list of situational and outcome statements associated with moral compromise and levels of moral distress in midwifery to inform the development of a tool to measure levels of moral distress in midwives.MethodsThis e-Delphi study was the third stage of a sequential exploratory mixed-methods study. Using an online strategy, three iterative rounds of e-Delphi were collected and analysed for consensus on situations leading to moral distress and the associated psychological outcomes.FindingsTwenty participants contributed across the three rounds. Consensus was met in 40 morally compromising situation statements. The highest level of consensus related to excessive workloads and the associated negative impact of this on women and families. Consensus on outcomes following exposure to morally distressing situations led to the development of a continuum scale from moral frustration to moral injury.Discussion/conclusionThis is the first study to use a consensus method to establish different levels of moral compromise, frustration, distress, and injury in midwifery practice. The findings of this study contribute to a growing body of literature that supports the concept of moral distress occurring across a continuum
The satisfaction of elderly people with elderly caring social organizations and its relationship with social support and anxiety during the COVID-19 pandemic: a cross-sectional study
Get PDFBackground: With the deepening of China’s aging population, higher demands have been placed on the supply of elderly care services. As one of the main sources of providing elderly care services, the quality of service provided by elderly caring social organizations (SOs) directly affects the quality of life of the elderly. In recent years, mental health issues among the elderly have become increasingly prominent, especially with the onset of the COVID-19 pandemic. Necessitating the need to pay much more attention to the social support and mental health of this population. This study, therefore, explores the mediating role of institutional satisfaction between the social support and anxiety levels of elderly people in Chongqing’s elderly caring SOs. Method: This study employed a multi-stage stratified random sampling method to survey 1004 service recipients in elderly caring social organizations from July to August 2022. The self-made sociodemographic questionnaire, institutional satisfaction questionnaire, MSPSS, and GAD-7 were used to collect data on sociodemographic characteristics, institutional satisfaction, social support, and anxiety levels of older adults. Exploratory Factor Analysis and Cronbach’s alpha were used to test construct validity and scale reliability, respectively. Data features were described with One-Way Analysis of Variance, while Multiple Linear Regression and Structural Equation Modeling were used to evaluate relationships between social support, institutional satisfaction, and anxiety levels. Results: The average institutional satisfaction score for elderly people in elderly caring SOs was 48.14 ± 6.75. Specifically, the satisfaction score for environmental quality and the satisfaction score for service quality were 16.63 ± 2.56 and 31.52 ± 4.76, respectively. In terms of socio-demographic variables, the presence of visits from relatives, personal annual average income, and self-rated health status all have significant effects on anxiety. Elders who receive visits from relatives have lower levels of anxiety compared to those who do not. Personal annual average income and self-rated health status are negatively correlated with anxiety levels. Social support had significant positive effect on institutional satisfaction, while institutional satisfaction had significant negative effect on anxiety. Institutional satisfaction partially mediated the relationship between social support and anxiety. Conclusions: Our research demonstrates that improving the quality of organizational services in elderly caring SOs and increasing institutional satisfaction among the elders has significant potential for reducing anxiety levels among the elderly. Additionally, the social support by visits from family members cannot be overlooked. We encourage increasing the frequency of family visits through various means to enhance the support provided to elderly individuals
“Accompanying the series”: Early British television cookbooks 1946-1976
Get PDFThis paper provides a historical analysis to demonstrate the connections and developmental links which emerged between cookbooks and television in Britain after World War II, focused on television broadcasts in the period 1946 and 1976. In this paper, I discuss how early presenters of British television cookery programmes, and their publishers, had vision and marketing skills which enabled links between visual and printed media, and established a pattern of connected cookbook and television production which is taken for granted today. I examine the connected television and publishing careers of three early British Broadcasting Corporation (BBC) television cooking pioneers: Marguerite Patten, Philip Harben and Fanny Cradock, who collectively dominated on-screen cooking programmes from the late 1940s until the mid-1970s. By analyzing their cookbooks, particularly their jackets and promotional materials, and interpreting archival research conducted in the BBC Written Archives and other documentary archives, their contributions will be discussed alongside the development of the television-connected cookbook in Britain. I conclude that these television cooks and presenters made a significant contribution on and off our screens during that period which established the connection between television cooking programmes and cookbooks in Britain
Data sources:
Access Repository Dashboard
Do you manage Repository@Napier? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!