| Summary: | In shipping, condition monitoring (CM) has the capacity for big data but also very high communications costs. Thus, the use of continuous condition monitoring in the shipping industry is not as prevalent as in others. It is found that trust in technology, data security/ownership, the capital cost of investment, cost of training, operational cost and direct benefit association are some of the most important inhibitors. To reduce the volume of data, edge processing is a new paradigm of computing. Its goal is to address the issues generated through the increasing flow of recorded data to central locations for big data analytics. aThe existing solutions are adopted by 12% of the global fleet corresponding to the newbuilt ships, while sensors and monitoring infrastructure exists in the majority. Solutions targeting newbuilt ships have a requirement of extensive refitting and training overheads. Solutions for existing vessels are mostly hand-held equipment, do not support continuous monitoring and do not display the information in business relevant terms. A wireless ship CM reduces capital investment costs but has high operational costs due to the centralised data processing software. The proposed novel system is edge processing wireless ship CM data under constrains. The system's traffic reduction is achieved through feature and event extraction on the data acquisition devices. Also a data management strategy is implemented along with decision support which provides direct benefit association with maintenance actions. The multi-constrain multi-parameter approach identifies the best maintenance action to be taken onboard the ship. Finally, minimal satellite data transmission provides visibility of condition to shore. The system was successfully applied in case studies. According to the evaluation results, the system is reliable and suitable for the application, is able to identify and suggest appropriate maintenance actions and offers several benefits against other maintenance and condition monitoring approaches.
|
|---|
