| Paper | Title | Page |
|---|---|---|
WEPP31 |
Web-Based Software Development for Failure Prediction in Commercial Cyclotrons Used in Medical Applications | |
|
||
| Radioisotopes used for PET and SPECT diagnostics are commonly produced using commercial cyclotrons. These accelerators are equipped with online monitoring instrumentation that allows the machine to be operated within the safety margins defined by the manufacturer. However, the probabilities of cyclotron failure cannot be estimated simply by online data logging but must be supplemented by a long-term data-driven analysis and storage system. The low level of development of the analysis and storage systems makes it difficult to optimize the preventive maintenance schedule and makes it almost impossible to prevent failures. Beta Pharma Technologies (BPT) has developed a web-based big data software, which in combination with a database, bridges the gap in storage and analytics capabilities of the current cyclotron software. This software analyzes the evolution of the cyclotron in the medium-long term, extracting statistical information and correlation between the different parameters of the cyclotron. To validate the method, BPT has used the software intensively to analyze the big data generated by various commercial cyclotrons. This article shows that by analyzing logged data using analysis tools the cyclotron failure can be identified before the actual machine failure. Additionally, the data analysis carried out using this methodology allows optimizing the material resources and the workers’ exposure time. | ||
| Export • | reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml) | |