Browsing by Author "Unal, Perin"

Now showing 1 - 6 of 6

Citation - WoS: 1
Citation - Scopus: 1
Autonomous Structural Health Monitoring and Remaining Useful Life Estimation of Floating Offshore Wind Turbine Cables: Part I
(Johnson Matthey Publ Ltd Co, 2025) Berker, Metehan; Unal, Perin; Deveci, Bilgin U.; Unal, Aras Firat; Avenoglu, Bilgin; Ozbayoglu, A. Murat
Floating offshore wind (FOW) farms are key in meeting Europe's renewable energy targets, harnessing wind energy from waters 60 m or deeper, where bottom-fixed farms are unfeasible. Additionally, floating structures allow for the installation of larger turbines than stationary farms, which in turn leads to a greater energy output. However, cable failures dramatically impact the energy transmission from the farms and cause most of the financial losses. Monitoring and maintenance tasks are challenging due to the harsh ocean conditions. The FLoating Offshore Wind turbine CAble Monitoring (FLOW-CAM) project, supported by European Union's HORIZON 2020 programme, studies the structural health monitoring (SHM) of defects in the power cables of the FOW farms which encompass inspection and detection applications. An SHM system integrated with a remotely operated vehicle (ROV) was developed for underwater inspection and maintenance, supporting collection and presentation of essential data through an advanced interface. Part I details the technologies and methods used in this research.
Citation - WoS: 1
Citation - Scopus: 1
Autonomous Structural Health Monitoring and Remaining Useful Life Estimation of Floating Offshore Wind Turbine Cables: Part II
(Johnson Matthey Publ Ltd Co, 2025) Berker, Metehan; Unal, Perin; Deveci, Bilgin U.; Unal, Aras Firat; Avenoglu, Bilgin; Ozbayoglu, A. Murat
Part II reports on a new structural health monitoring (SHM) system integrated with a remotely operated vehicle (ROV) developed for underwater inspection and maintenance, part of the FLoating Offshore Wind turbine CAble Monitoring (FLOW-CAM) project, supported by European Union's HORIZON 2020 programme. Image data from underwater systems are analysed using computer vision techniques. Investigations into cable defect detection and the estimation of corrosion and remaining useful life (RUL) have been held to monitor cable health, achieving results close to reality. FLOW-CAM's collective works establish a basis for advancing underwater inspection and maintenance, concentrating on the development of practical and effective tools and strategies to optimise the functionality and reliability of floating offshore wind (FOW) farms.
Citation - WoS: 2
Citation - Scopus: 7
A Comparison of Deep Transfer Learning Methods on Bearing Fault Detection
(IEEE, 2021) Deveci, Bilgin Umut; Celtikoglu, Mert; Alp, Tilbe; Albayrak, Ozlem; Unal, Perin; Kirci, Pinar
In rotating machinery, bearings are widely used as universal components. Bearings are placed in critical positions, therefore, in predictive maintenance, it is crucial to diagnose bearing faults accurately and in a timely manner. In this paper, three diverse pre-trained networks on bearing fault diagnosis are discussed. A generic intelligent bearing fault diagnosis system based on AlexNet, GoogLeNet and ResNet-50 with transfer learning is proposed to distinguish and classify different bearing faults. Three bearing faults at all various loads and speeds selected from the Case Western Reserve University (CWRU) bearing dataset were converted to time-frequency images, in order to improve the performance of the proposed networks. Results showed that when compared to previous methods, the proposed method achieved outstanding execution, with overall classification training accuracy of 100%, validation accuracy of 99.27%.
Citation - WoS: 6
Citation - Scopus: 10
A Review: Sensors Used in Tool Wear Monitoring and Prediction
(Springer International Publishing Ag, 2022) Unal, Perin; Deveci, Bilgin Umut; Ozbayoglu, Ahmet Murat
Tool wear prediction/monitoring of CNCs is crucial for improving manufacturing efficiency, guaranteeing product quality, and minimizing tool costs. As a computer-aided application, it has a significant role in the future and development of Industry 4.0. Sensors are the key piece of hardware used by data-driven enterprises to predict/monitor tool wear. The purpose of this study is to inform about the predominant types of sensors used for tool wear monitoring/prediction. This study serves as a resource for researchers and manufacturers by providing the recent trends in sensors for tool wear monitoring. Thus, it may help reduce the time spent on sensor selection.
Citation - WoS: 5
Citation - Scopus: 7
A Thorough Analysis and Comparison of Data Communication Protocols Used in Industry 4.0: The Case of Smart-CNC
(IEEE, 2022) Deveci, Bilgin Umut; Bas, Hilal; Ummak, Emre; Albayrak, Ozlem; Unal, Perin
Several key methods have been developed to transfer data accurately and completely. In this regard, a communication protocol specifies the rules for transmitting blocks of data between network nodes. This study examines in detail the scope of the Qu4lity project, the data communication protocols used by the pilots, the protocols used by TEKNOPAR, and the benefits, drawbacks, and application areas of different data communication protocols. Protocols were analysed against a number of technical criteria aimed at identifying the most appropriate procedure. Communication protocols were categorized considering the following features: technical qualifications such as transport protocol, coding format, message system, architecture, security, and scalability.
A Tool Condition Monitoring Study To Support Circular Economy
(IEEE Computer Soc, 2024) Kaleli, Inci Sila; Unal, Perin; Deveci, Bilgin Umut; Albayrak, Ozlem
CNC (Computer Numerical Control) machines are vital for precision and efficiency in manufacturing but are prone to tool wear, causing disruptions and sustainability challenges. This study introduces a project aimed at sustainable CNC tool management within the circular economy framework, focusing on extending tool lifespan through predictive analytics. Real-time monitoring predicts optimal tool replacement times, promoting reuse, repair, and recycling. The methodology includes data collection, preprocessing, anomaly detection, real-time analysis, and machine learning model selection, with the Random Forest model proving most effective. Unique contributions include the integration of advanced sensor data with AI-driven predictive maintenance, and the application of circular economy principles to CNC tool management. The results highlight significant accuracy in tool condition categorization, contributing to waste reduction and sustainable practices in manufacturing.