Pregled bibliografske jedinice broj: 1278783
Methods for Monitoring and Detecting Faults in IoT DosimetrymInstrumentation Based on Machine Learning on Edge Computing Devices
Methods for Monitoring and Detecting Faults in IoT DosimetrymInstrumentation Based on Machine Learning on Edge Computing Devices // Proceedings of the 13th International Conference of the Croatian Nuclear Society, (2022), 142; 1-9 (međunarodna recenzija, članak, znanstveni)
CROSBI ID: 1278783 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Methods for Monitoring and Detecting Faults in IoT DosimetrymInstrumentation Based on Machine Learning on Edge Computing Devices
Autori
Pavelić, Dora ; Pavelić, Luka ; Petrinec, Branko ; Prlić, Ivica
Izvornik
Proceedings of the 13th International Conference of the Croatian Nuclear Society (9789-5348)
(2022), 142;
1-9
Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni
Ključne riječi
Radiation Dosimetry, Machine Learning, Fault Detection
Sažetak
Due to the importance of safety, reliability and efficiency of dosimetry instrumentation, as well as increasing complexity of the technologies, we are proposing a method for early failure detection that could enable the necessary prompt response. IoT dosimetry sensors are usually required to operate for several years on a single battery and they are often installed in large numbers which place high energy and cost constraints. Therefore, the analysis and prediction itself is increasingly performed on devices that are close to the sensors. The concept of bringing analytical computational resources closer to the sensors themselves is called edge computing. In this work we will consider the application of machine learning for the purpose of fault detection in IoT dosimetry instrumentation as well as the various approaches with which these detections are realized with the help of edge computing devices.
Izvorni jezik
Engleski