Performance Monitoring of Rotating Fluid Machines Using Cloud Computing and Digitized Characteristic Plots
Hirvonen, Jukka (2024)
2024
Automaatiotekniikan DI-ohjelma - Master's Programme in Automation Engineering
Tekniikan ja luonnontieteiden tiedekunta - Faculty of Engineering and Natural Sciences
This publication is copyrighted. Only for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2024-08-23
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202408208201
https://urn.fi/URN:NBN:fi:tuni-202408208201
Tiivistelmä
The rise of industrial internet, Big Data and cloud computing ensure that more and more data from industrial processes is available for analysis. Processing this massive flow of information into actionable insights presents a great challenge, but the potential benefits are equally great. Among the leading solutions to this challenge are modern, cloud-based enterprise data warehouses, featuring information models for process data and equipment data.
In performance monitoring of industrial machines, characteristic data on the machine provides crucial context for decision making. One type of such data are characteristic curves, which represent the expected performance of the machine. These curves are often not available in numerical form, and their information must be digitized before use in computer calculations. The machines in the focus of this study were rotating pumps and fans. They are critical pieces of equipment used across many industries. Performance monitoring of these machines is important for safety and reliability reasons, as well as for reducing the energy consumption and climate impact of their operation.
The goal of this study was to design and test a way of combining cloud-borne industrial process data and digitized characteristic curves to utilize them in performance monitoring of industrial machinery. The resulting prototype system utilizes an existing cloud environment, drawing process and equipment data from a live enterprise data warehouse and visualizing it in a web-based monitoring interface. The combined visual information enables the user to see the operating history of the machine immediately, enriched with the context of expected performance and limits of safe operation.
The study also established a suitable way to digitize the curves for integration in the system by creating a configuration application. The graphical user interface of the application provides structure and assistance to the digitizing process with the intent to enable the user to input the required curves and attributes quickly and reliably.
The designs of both solutions were based on a review of existing literature and state-of-the-art and they were proven to function by deploying them in a real-world monitoring case. Both solutions were tested for speed and accuracy and found feasible. Additionally, the user experience of the visualization prototype was tested through survey. The surveyed System Usability Scale result was high and overall experience rated positive.
The solution architecture composes of widely used technologies and services. Therefore, similar solutions could be implemented in other cloud environments and applied to a variety of monitoring cases in the future.
In performance monitoring of industrial machines, characteristic data on the machine provides crucial context for decision making. One type of such data are characteristic curves, which represent the expected performance of the machine. These curves are often not available in numerical form, and their information must be digitized before use in computer calculations. The machines in the focus of this study were rotating pumps and fans. They are critical pieces of equipment used across many industries. Performance monitoring of these machines is important for safety and reliability reasons, as well as for reducing the energy consumption and climate impact of their operation.
The goal of this study was to design and test a way of combining cloud-borne industrial process data and digitized characteristic curves to utilize them in performance monitoring of industrial machinery. The resulting prototype system utilizes an existing cloud environment, drawing process and equipment data from a live enterprise data warehouse and visualizing it in a web-based monitoring interface. The combined visual information enables the user to see the operating history of the machine immediately, enriched with the context of expected performance and limits of safe operation.
The study also established a suitable way to digitize the curves for integration in the system by creating a configuration application. The graphical user interface of the application provides structure and assistance to the digitizing process with the intent to enable the user to input the required curves and attributes quickly and reliably.
The designs of both solutions were based on a review of existing literature and state-of-the-art and they were proven to function by deploying them in a real-world monitoring case. Both solutions were tested for speed and accuracy and found feasible. Additionally, the user experience of the visualization prototype was tested through survey. The surveyed System Usability Scale result was high and overall experience rated positive.
The solution architecture composes of widely used technologies and services. Therefore, similar solutions could be implemented in other cloud environments and applied to a variety of monitoring cases in the future.