Hierarchical Federated Transfer learning and Digital Twin Enhanced Secure Cooperative Smart Farming DOI

Lopamudra Praharaj,

Maanak Gupta, Deepti Gupta

et al.

2021 IEEE International Conference on Big Data (Big Data), Journal Year: 2023, Volume and Issue: unknown, P. 3304 - 3313

Published: Dec. 15, 2023

The agriculture industry is extensive utilizing AI and data-driven systems for efficiency automation, with the goal to meet rising food demand. Individual farm owners can leverage agricultural cooperatives consolidate resources, exchange data, share domain knowledge. These enable generation of AI-supported insights their member farmers. However, this collaborative approach has raised concerns among individual smart regarding cybersecurity threats, privacy. A breach not only endangers attacked but also risks entire network farms members within cooperative. In research, we emphasize security challenges cooperative farming introduce a multi-layered architecture incorporating Digital Twins (DT). Further, hierarchical federated transfer learning framework designed address mitigate threats in farming. Our leverages Federated Learning (FL) based Anomaly Detection (AD), which operate on edge servers, enabling execution AD models locally without exposing farm's data. This localization excellent generalization ability, highly improve detection unknown cyber attacks. We employ FL structure that supports aggregation at various levels, fostering multi-party collaboration. Furthermore, have devised an integrates Convolutional Neural Networks (CNN) Long Short-Term Memory (LSTM) models, complemented by learning. objective expedite training duration while upholding high accuracy levels. To illustrate our proposed architecture, present use case demonstrate model's capabilities. proof-of-concept implementation Amazon Web Services (AWS) environment, reflecting real-world feasibility.

Language: Английский

How digital twin technology may improve safety management: A multi-industry perspective DOI
Patrick X.W. Zou, Songling Ma

Safety Science, Journal Year: 2025, Volume and Issue: 189, P. 106837 - 106837

Published: May 2, 2025

Language: Английский

Citations

0
General purpose digital twin framework using digital shadow and distributed system concepts DOI
Ayman AboElHassan, Ahmed H. Sakr, Soumaya Yacout

et al.

Computers & Industrial Engineering, Journal Year: 2023, Volume and Issue: 183, P. 109534 - 109534

Published: Aug. 14, 2023

Language: Английский

Citations

10
Industrial Insights on Digital Twins in Manufacturing: Application Landscape, Current Practices, and Future Needs DOI Creative Commons
Rosario Davide D’Amico, Sri Addepalli, John Ahmet Erkoyuncu

et al.

Big Data and Cognitive Computing, Journal Year: 2023, Volume and Issue: 7(3), P. 126 - 126

Published: June 29, 2023

The digital twin (DT) research field is experiencing rapid expansion; yet, the on industrial practices in this area remains poorly understood. This paper aims to address knowledge gap by sharing feedback and future requirements from manufacturing industry. methodology employed study involves an examination of a survey that received 99 responses interviews with 14 experts 10 prominent UK organisations, most which are involved defence industry UK. explored topics such as DT design, return investment, drivers, inhibitors, directions for development manufacturing. study’s findings indicate DTs should possess characteristics adaptability, scalability, interoperability, ability support assets throughout their entire life cycle. On average, completed projects reach breakeven point less than two years. primary motivators behind were identified be autonomy, customer satisfaction, safety, awareness, optimisation, sustainability. Meanwhile, main obstacles include lack expertise, funding, interoperability. concludes federation twins paradigm shift thinking essential components development.

Language: Английский

Citations

9
An IoT Architecture Leveraging Digital Twins: Compromised Node Detection Scenario DOI
Khaled Alanezi, Shivakant Mishra

IEEE Systems Journal, Journal Year: 2024, Volume and Issue: 18(2), P. 1224 - 1235

Published: June 1, 2024

Language: Английский

Citations

2
A Review of Digital Twinning for Rotating Machinery DOI Creative Commons
Vamsi Inturi, Bidisha Ghosh,

G. R. Sabareesh

et al.

Sensors, Journal Year: 2024, Volume and Issue: 24(15), P. 5002 - 5002

Published: Aug. 2, 2024

This review focuses on the definitions, modalities, applications, and performance of various aspects digital twins (DTs) in context transmission industrial machinery. In this regard, around Industry 4.0 even aspirations for 5.0 are discussed. The many definitions interpretations DTs domain first summarized. Subsequently, their adoption levels rotating machineries manufacturing lifetime observed, along with type validations that available. A significant focus integrating fundamental operations system scenarios over lifetime, sensors advanced machine or deep learning, other statistical data-driven methods highlighted. summarizes how individual extremely helpful design, manufacturing, decision making when a DT can remain incomplete limited.

Language: Английский

Citations

2
Possibilities of information technologies in the processes of designing efficient processes for the production of aircraft structures DOI Creative Commons

Vyacheslav Bekhmetiev,

B. B. Safoklov,

Pavel Gusev

et al.

E3S Web of Conferences, Journal Year: 2023, Volume and Issue: 376, P. 01092 - 01092

Published: Jan. 1, 2023

The work presents the process of studying CAD- Systems and systems technological automated design elements aviation structures in serial production equipment. It has been established that most important goal laying CAD is to create a single information space, which involves rejection direct interaction data transfer between all participants product life cycle, implemented system "Vertical".

Language: Английский

Citations

6
The use of the digital twin in the design of a prefabricated product DOI Creative Commons

Dmitry Golovin,

Andrey N. Smolyaninov,

Dmitry Degtev

et al.

E3S Web of Conferences, Journal Year: 2022, Volume and Issue: 363, P. 04001 - 04001

Published: Jan. 1, 2022

This paper discusses the process of creating a digital twin product, which is virtual model mechanical connection. The modeling was carried out using Pro/ENGINEER software, allows building three-dimensional product assembly process, set electronic models equipment and tools. include mathematical description geometric, physical-mechanical technical parameters objects under consideration. It shown that it formation triad models: product-man-equipment in considered area computer-aided design technological processes for implementation connections with necessary accuracy adequacy.

Language: Английский

Citations

9
A Novel Method of Digital Twin-Based Manufacturing Process State Modeling and Incremental Anomaly Detection DOI Creative Commons
Qinglei Zhang, Zhen Liu, Jianguo Duan

et al.

Machines, Journal Year: 2023, Volume and Issue: 11(2), P. 151 - 151

Published: Jan. 22, 2023

In the manufacturing process, digital twin technology can provide real-time mapping, prediction, and optimization of physical process in information world. order to realize complete expression accurate identification changes state a framework incremental learning driven by stream data is proposed. Additionally, novel method data-driven equipment operation modeling anomaly detection proposed based on twin. Firstly, hierarchical finite machine (HFSM) for was completely express state. Secondly, used detect job data, so as change status real time. Furthermore, F1 value time consumption algorithm were compared analyzed using general dataset. Finally, applied practical case development welding manufacturer’s system. The flexibility model calculated quantitative method. results show that help system mapping quickly, effectively, flexibly.

Language: Английский

Citations

5
Digital Twins Temporal Dependencies-Based on Time Series Using Multivariate Long Short-Term Memory DOI Open Access
Abubakar Isah, Hyeju Shin, Seungmin Oh

et al.

Electronics, Journal Year: 2023, Volume and Issue: 12(19), P. 4187 - 4187

Published: Oct. 9, 2023

Digital Twins, which are virtual representations of physical systems mirroring their behavior, enable real-time monitoring, analysis, and optimization. Understanding identifying the temporal dependencies included in multivariate time series data that characterize behavior system crucial for improving effectiveness Twins. Long Short-Term Memory (LSTM) networks have been used to represent complex identify long-term links Industrial Internet Things (IIoT). This paper proposed a Twin dependency technique using LSTM capture IIoT data, estimate lag between input intended output, handle missing data. Autocorrelation analysis showed lagged variables, aiding discovery dependencies. The evaluated model by providing it with set previous observations asking forecast value at future steps. We conducted comparison our six baseline models, utilizing both Smart Water Treatment (SWaT) Building Automation Transaction (BATADAL) datasets. Our model’s capturing was assessed through Function (ACF) Partial (PACF). results experiments demonstrate enhanced achieved better prediction performance.

Language: Английский

Citations

5
Deep learning anomaly detection in AI-powered intelligent power distribution systems DOI Creative Commons

Jing Duan

Frontiers in Energy Research, Journal Year: 2024, Volume and Issue: 12

Published: March 15, 2024

Introduction: Intelligent power distribution systems are vital in the modern industry, tasked with managing efficiently. These systems, however, encounter challenges anomaly detection, hampered by complexity of data and limitations model generalization. Methods: This study developed a Transformer-GAN that combines Transformer architectures GAN technology, efficiently processing complex enhancing detection. model’s self-attention generative capabilities allow for superior adaptability robustness against dynamic patterns unknown anomalies. Results: The demonstrated remarkable efficacy across multiple datasets, significantly outperforming traditional detection methods. Key highlights include achieving up to 95.18% accuracy notably high recall F1 scores diverse scenarios. Its exceptional performance is further underscored highest AUC 96.64%, evidencing its ability discern between normal anomalous patterns, thereby reinforcing advantage security stability smart systems. Discussion: success not only boosts but also finds potential applications industrial automation Internet Things. research signifies pivotal step integrating artificial intelligence into sector, promising advance reliability intelligent evolution future

Language: Английский

Citations

1