Arabian Journal of Geosciences, Год журнала: 2024, Номер 17(11)
Опубликована: Окт. 8, 2024
Язык: Английский
Safety Science, Год журнала: 2024, Номер 177, С. 106596 - 106596
Опубликована: Июнь 17, 2024
Язык: Английский
Процитировано
2
Опубликована: Сен. 14, 2023
This research paper presents a novel decision tree-based method for predicting health hazards based on multilevel Internet of Things (IoT). study's primary objective is to employ machine learning and deep techniques the field medical science in an effort make physicians' jobs easier have positive effect humanity. dataset consists 132 parameters from which 42 distinct disease types can be predicted. The data collected by (IoT) devices, are also used validation purposes. train tree classifier, then integrated into IoT-based device real-time risk prediction. Using classification metrics, accuracy model evaluated, feature importances analysed determine most significant risks. In addition, process selection employed eradicate less parameters, resulting refined model. multi-level IoT data, proposed demonstrates promising results with high hazards. contribute development intelligent healthcare systems facilitate early detection prevention.
Язык: Английский
Процитировано
4
Engineering Construction & Architectural Management, Год журнала: 2024, Номер unknown
Опубликована: Дек. 26, 2024
Purpose Equipment failure is a critical factor in construction accidents, often leading to severe consequences. Therefore, this study addresses two significant gaps safety research: (1) effectively using historical data investigate equipment and (2) understanding the classification of according Occupational Safety Health Administration (OSHA) standards. Design/methodology/approach Our research utilized multi-stage methodology. We curated from OSHA database, distinguishing accidents involving failures. Then we developed framework generative artificial intelligence (AI) large language models (LLMs) minimize manual processing. This employed two-step prompting strategy: classifying narratives that describe failures analyzing these cases extract specific details (e.g. names, types, categories). To ensure accuracy, conducted analysis subset reports establish ground truth tested different LLMs within our approach, comparing their performance against truth. Findings The demonstrated 95% accuracy determining if 73% extracting enabling automated categorical identifications. These findings highlight LLMs’ promising identification compared methods. Research limitations/implications research’s focus on not only validates but also highlights its potential for broader application across various accident categories beyond construction, extending into any domain with accessible narratives. Given such are essential regulatory bodies like OSHA, framework’s adoption could significantly enhance reporting, contributing more robust protocols industry-wide. Practical implications Using enables us use narratives, reliable source data, analysis. It provides deeper insights than traditional detailed at an unprecedented level. enhanced can inform improve worker training, education policies, applications safety-critical domains. Originality/value presents novel approach AI LLMs, reducing processing time while maintaining high accuracy. By identifying efficiently, work lays groundwork developing targeted protocols, overall improvements practices advancing data-driven processes.
Язык: Английский
Процитировано
1
Safety Science, Год журнала: 2024, Номер 174, С. 106468 - 106468
Опубликована: Фев. 23, 2024
Язык: Английский
Процитировано
1
Proceedings of the Institution of Civil Engineers - Management Procurement and Law, Год журнала: 2024, Номер 177(4), С. 193 - 206
Опубликована: Июль 23, 2024
Accurate forecasting of work accidents is paramount importance in promoting workplace safety and improving risk-management strategies. This study proposes a novel approach based on neural network fitted with the Levenberg–Marquardt algorithm to predict future accident numbers 22 different occupational groups within Turkish construction industry. By utilising historical official data spanning years 2014–2022, aim provide insights into potential rates that may arise job categories. The constructed model consists two hidden layers. Leveraging powerful capabilities algorithm, trained capture effectively complex dynamics underlying findings demonstrate effectiveness proposed high degree precision. successfully leverages temporal trends factors present data. employing an advanced framework this offers robust methodology for predicting across diverse results obtained from can guide development targeted preventive measures, tailored training programmes efficient resource allocation
Язык: Английский
Процитировано
1
Safety and Health For Medical Workers, Год журнала: 2024, Номер 1(2), С. 78 - 94
Опубликована: Июль 10, 2024
Objective: This study focuses on utilizing Machine Learning (ML) approaches to improve Occupational Safety and Health (OSH) performance, involving the prediction prevention of risks based data.Methods: Analysis a dataset 550 OSH incident reports from Metax Cancer Hospital (2019–2023) was conducted using descriptive inferential statistics. algorithms including decision trees, random forests, support vector machines were used for evaluation results. The models evaluated various performance metrics such as accuracy, precision, recall, AUC.Findings: analysis made key observations both workplace environmental factors, safety protocols, occurrence. ML demonstrated high with forests achieving best accuracy in terms correct classification events. These findings highlight promise hospitals.Novelty: We propose an original contribution integration process towards improvement hospital ecosystem also characterized complex challenges which predictive analytics can yield substantial risk mitigation.Research Implications: proposes spillover framework establishing intelligence systems that combines data-driven techniques traditional management structures. It highlights role real-time improving outcomes. demonstrates ability facilitate assessment safety.
Язык: Английский
Процитировано
1
Buildings, Год журнала: 2024, Номер 14(6), С. 1797 - 1797
Опубликована: Июнь 13, 2024
Before starting any construction work, providing workers with awareness about past similar accident cases is effective in preventing mishaps. Based on reports, this study developed two models to identify accidents at sites site information. The information includes 16 parameters, such as type of accident, the work which occurred, weather conditions, contract etc. first model, classification uses named entity recognition tasks classify information, extracted from reports. second similar-site retrieval finds most that occurred input a semantic textual similarity task match classified it. A total 17,707 reports South Korean were found; these trained use Language Understanding Evaluation–Bidirectional Encoder Representations Transformers (KLUE-BERT) for processing. model achieved an average accuracy 0.928, and was precisely matched, mean cosine score exceeding 0.90. These could provide accidents, enabling proactive safety measures, site-specific hazard identification worker education, thereby allowing risks before work. By integrating historical data, offer approach improving safety.
Язык: Английский
Процитировано
0
International Journal of Civil Engineering, Год журнала: 2024, Номер 11(11), С. 58 - 66
Опубликована: Ноя. 30, 2024
The construction industry in the United States has a significant representation of Hispanic workers, who comprise about one-third total workforce. Every year, 300 workers die workplace. Worse still, within small companies, this population disproportionately higher rates fatalities and injuries. Current safety training methods have been insufficient addressing specific needs due to language barriers, cultural differences, immigration concerns, operational issues. To address these issues, paper summarizes research on enhancing for companies through innovative AI-based systems. This first identifies designs programs enhance culture awareness. Next, it investigates how AI information technology (IT) can improve effectiveness by focusing communication, establishment, workers' rights awareness, safety. Finally, will establish systems tailored companies. proposed intelligent provide platform connect worker communities generate comprehensive suited explicitly needs. By critical issues faced demographic, develops an infrastructure create safer healthier workplace workers.
Язык: Английский
Процитировано
0
2020 4th International Conference on Electronics, Communication and Aerospace Technology (ICECA), Год журнала: 2023, Номер unknown, С. 1197 - 1203
Опубликована: Ноя. 22, 2023
The alarming increase in road accidents recent years has elevated them to a significant global issue., making the ninth leading cause of death worldwide. It is unfortunate and completely unacceptable that there have been fatalities because these accidents. As result., it critical this problem be addressed thoroughly. With aid cutting-edge machine learning techniques., proposed study will analyze traffic incidents detail. main goal pinpoint major causes collisions offer insightful suggestions for reducing issue. divide accident severity into three categories: fatal injury., serious minor injury. To do this., makes use variety supervised such as Decision Trees., Support Vector Machines (SVM)., Multinomial Naive Bayes., K-Nearest Neighbors (KNN)., Random Forests., XGBoost., MLPClassifiers., AdaBoost. Notably., Forest model excels with outstanding performance., obtaining stunning 90% accuracy rate successfully forecasting severity.
Язык: Английский
Процитировано
1
Journal of Engineering Design and Technology, Год журнала: 2024, Номер unknown
Опубликована: Фев. 19, 2024
Purpose The construction business is widely recognised for its inherent complexity and dynamic nature, which stems from the nature of job involved. industry often regarded as one most challenging industries globally in terms implementing environmental, health safety (EHS) practices. However, absence EHS, cannot be considered sustainable. Therefore, this study aims to identify trends, knowledge gaps implications EHS research enhance activities knowledge. Design/methodology/approach adopted a science mapping approach involving bibliometric scientometric analysis 407 publications Scopus database with VOSviewer software. based on journal articles without restriction any time range. Findings main focus identified includes sustainability-related studies, risk-related, environmental issues, management, integrated management systems related process. Some emerging areas also include productivity, design, culture, social sustainability machine learning. influential productive publication sources, countries/regions highest impact were determined. Research limitations/implications Documents published because wider coverage database. Journal written English language represent inclusion criteria, whereas other documents excluded analysis. limited search engineering subject area. Practical findings will enlighten stakeholders practitioners focal domain, are vital enhancing industry. Originality/value To best authors’ knowledge, review-based first attempt internationally conduct extant literature domain through assessments.
Язык: Английский
Процитировано
0