MyAdvantech Registration

MyAdvantech is a personalized portal for Advantech customers. By becoming an Advantech member, you can receive latest product news, webinar invitations and special eStore offers.

Sign up today to get 24/7 quick access to your account information.

How to Help Manufacturers Get Started with Equipment Preventive Maintenance

Date: 1/10/2019 12:00:00 AM

By Liao Pei-jun with images provided by Advantech
Interview with Advantech Embedded IoT Group Product Manager Alan Kou

In recent years, IoT and Industry 4.0 has moved away from the application planning stage and entered the phase of implementation. In the practical process of implementation, equipment network connectivity is an essential foundation underpinning any IoT solution. Only when equipment and devices are fully connected will a system be able to acquire and integrate data to enable a fully automated smart industrial solution. Preventive maintenance is one of the most important features of Industry 4.0, which helps to reduce machine downtime and lowers maintenance cost. This is especially important for the monitoring of motors, which are the operational core of most machinery and should be closely watched.

As Advantech Embedded-IoT Group Product Manager Alan Kou pointed out, “Conventionally, there are two methods for monitoring motor status. One of them is to install sensors to collect data on voltage, current, rotation speeds, and temperature. However, these four data categories will not be able to precisely reflect the entire health condition of any motor, because minor conditions of the motor’s core components, including motor shaft, washers, and bearings, will not immediately manifest themselves in the data until the motor fails and it becomes so serious that operations need to cease for examination and repairs.”

The second method for monitoring is to use a frequency spectrum analyzer and vibration sensor to measure the vibration frequency of the motor on a regular basis. This lets machinery engineers judge whether repairs or maintenance are needed. Because a motor is in constant rotation, the measurement of vibration needs to be more accurate to reflect the true health status of the motor. However, the interpretation of the spectrum analysis results requires a professional individual. The location of a vibration sensor will affect the accuracy of measurement results so if an engineer is not experienced enough and if the sensor is put in the wrong place, it will not be easy to discover vibration abnormalities.

So we can see that both of these two traditional methods have their blind spots. To address them, Advantech worked hand-in-hand with its co-creation partner AnCAD and developed a hardware-and-software integrated equipment vibration monitoring solution called SRP-E2i120, which uses a high-performanced edge intelligence server to acquire massive data from vibration sensors, makes preliminary computations, and uploads pre-processed data to a backend cloud server, thereby breaking the limits of conventional methods which cannot collect and transmit huge volumes of data in real time.         

High integration from the field to the cloud from hardware to software

Alan Kao said that equipment vibration monitoring solution SRP-E2i120 is a complete solution covering all requirements from the bottom to the top layer of a typical IoT architecture by integrating a vibration -sensor, edge intelligence server, and cloud platform. Vibration monitoring software enables acquisition of motor vibration data, real-time monitoring, data analysis, and remote equipment management via a visual dashboard. Vibration monitoring software performs two things: first, it provides multiple math algorithms for analyzing sensory data, such as Fast Fourier Transform (FFT), Root Mean Square (RMS), and Short Term Fourier Transform (STFT); secondly, it uses an equipment management dashboard to visualize vibration data and allow administrators to see the status of their monitored equipment and motors, be able to set thresholds for alerts and alarms, manage events, and even implement predictive maintenance.

Equipment VibrationMonitoring Solution System Diagram


High-end edge server satisfying mass computing requirements

Because it is not feasible to directly send large complicated volumes of data acquired by vibration sensors to the cloud platform, it’s better to implement FFT first to transform the correlation of data between “time and vibrations” into “frequency and vibrations” so as to enable frequency spectrum analysis to plot the regularity of vibrations. To meet this end, Advantech selected a high-end performanced computer running on an Intel Core i5 processor as the edge computing device. Each edge server is able to support an average of four vibration sensors, sufficiently satisfying the distributive computing needs for accurate vibration data. Compared with competitors in the current market, the biggest difference of Advantech’s equipment vibration monitoring solution SRP E2i120 lies in its comprehensiveness. SRP E2i120 is a complete system for monitoring motor vibration and comprises of sensors, edge intelligence server, cloud platform, and vibration-monitoring software with an application dashboard. Most IoT suppliers provide only one of two product categories, unlike this Advantech solution which covers everything from the field to the cloud. With its application dashboard, users can visualize vibration data and status monitoring to reduce overall equipment maintenance cost and provide predictive maintenance, ensuring a better ROI for business owners.

Enabled Semiconductor manufacturer reduces equipment downtime and cost

For manufacturers, using this complete solution means they will avoid conventional problems resulting from buying products separately, such as compatibility and integration problems as well as data security issues. Moreover, the solution provides a system diagram for customers to see clearly what components should be added to fulfill a complete system and avoid wasted repetitive investment—an investment in Industry 4.0 at a reasonable cost. Alan Kao took the example of a semiconductor manufacturing facility, who initially used third party brand vibration sensors. Though the third party supplier also provided data for frequency spectrum analysis and time-frequency analysis, the data could not be uploaded to the backend database for integration; the data could only be viewed on a single machine, providing limited benefit. Taking all this into account, the semiconductor manufacturer decided to cooperate with Advantech instead, replacing their original sensors with high-end sensors to improve the quality of acquired data. The semiconductor manufacturer also accepted Advantech’s suggestion to continually send acquired data to a backend database after preliminary processing at the edge, in order to avoid data congestion at the sensors.

According to the statistics from the semiconductor manufacturer, the introduction of Advantech’s equipment vibration monitoring solution into their facility has brought three major benefits: first, the ability to detect machinery problems at an earlier stage, which has increased maintenance efficiency and normal operational time; second, the establishment of a predictive maintenance model, which has reduced machine downtime for maintenance from 8 hours to 4-5 hours and lowered maintenance cost accordingly; third, improvement of equipment availability rates from 98.43% to 98.85%.












                                                                                             Semiconductor Application Highlights

Advantech equipment vibration monitoring solution SRP-E2i120 is suitable not only for manufacturers of high value products such as semiconductors, but also for production lines of high priced products such as auto parts. Because the solution is able to prevent motor faults and reduce the cost of high-priced components, manufacturers of heavy-duty machines with higher safety concerns can also apply this solution to conduct final checks to ensure reliable operation and transform data into valuable information.