Monitor the image dynamics of the machine's tiny vibrations to ensure the normal operation of the machine!

As we all know, any machine will generate vibrations when it is running. Some vibrations represent normal operation, while some vibrations represent the initial signal of a fault. In the field of predictive maintenance, detecting vibration characteristics is a key factor in the diagnostic process. Vibration detection can identify and alleviate problems to prevent more serious incidents. By sensing, amplifying and measuring subtle movements, FLIR enables users to see the vibration characteristics on factory assets (such as machines) to ensure the normal operation of the machines.

Traditional testing is time-consuming and labor-intensive, with high downtime costs

The traditional method of detecting machine vibration is to deploy wired sensors (such as contact accelerometers) on the machine to monitor the vibrations that occur. After acquiring data from the sensors, the data is subjected to operating vibration shape analysis to present an animated model of the machine's motion, thereby visualizing the vibration mode.

But according to Jeff Hay, Ph.D., founder and CEO of RDI Technologies (Knoxville, Tennessee, U.S.), this technology not only takes time to collect measurements from multiple points, but also requires access to the machine. This is often not possible when the machine is difficult or impossible to access (there are heavy obstacles or glass in front of the machine). In addition, traditional contact measurement often requires the machine to be stopped when installing the accelerometer, resulting in costly downtime.

To this end, engineers at RDI Technologies developed a revolutionary non-contact video processing system called Iris M, which uses FLIR machine vision cameras to sense, amplify, and measure subtle machine-induced vibrations, eliminating the inherent drawbacks of using earlier technologies.

The Iris M system uses a FLIR 2.3Mpixel Grasshopper3 camera mounted on a Vanguard tripod, which acquires monochrome image data at a default resolution of 1920x1050 at 120 frames per second.

Data acquired from the camera is transferred via a USB 3.0 interface to a tablet computer, where it is analyzed using the company’s proprietary software, allowing users to see the vibration signature on factory assets such as machinery.

Reasons to choose FLIR machine vision cameras

In the Iris M system, the FLIR machine vision camera is the data acquisition device that collects video images and then extracts and analyzes motion from them. The FLIR 2.3Mpixel Grasshoppper3 GS3-U3-23S6M-C camera is mounted on a Vanguard tripod and acquires monochrome image data at a default resolution of 1920x1050 and a speed of 120 frames per second. The acquired data is transferred from the camera to a tablet such as Getac F110 or Microsoft Surface Book via a USB 3.0 interface via a non-drop cable.

The Iris M system is able to measure absolute displacement of plant assets that can be used to quantify motion directly from image points, rather than interpreting the point measurements to determine the type of motion and the errors present.

"The machine vibrations are then visualized using a proprietary video processing algorithm in the PC software, Motion Amplification. It analyses each image pixel by pixel, frame by frame, to determine which parts of the scene are moving. Next, it amplifies the magnitude changes in the amplitude of all motions in the scene to a level visible to the naked eye, thereby enhancing the understanding of the relationships between components causing any vibrations," said Dr. Hay. Using a graphical user interface running on a PC, the user can select a portion of the image for further analysis. The system software displays intensity data associated or time-dependent with those areas. Various mathematical functions, such as the Fast Fourier Transform, or FFT, can then be used to convert the time-dependent intensity data set into frequency-dependent intensity data. The user is then presented with the absolute unamplified amplitude and vibration phase for different frequencies for the selected portion of the scene.

Iris M system has high sensitivity and can be applied to more industries

Since its launch in Q3 2016, the Iris M system has changed the way industry professionals view vibrations using machine monitoring. Not only is the system easy to use, it also provides users with visible, easy-to-understand video images to better understand equipment operating conditions.

According to Dr. Hay, the choice of the FLIR Grasshopper camera has been a key reason for the system’s success. The camera’s 12-bit dynamic range allows it to capture subtle differences in pixel intensity between brightly lit and dark areas in an image, allowing the system’s software to extract more detailed changes than would otherwise be possible.

The Grasshopper camera series combines the latest CCD and CMOS technology with Point Grey's expertise to achieve high performance, high quality imaging.

But just as important is the Motion Amplification algorithm itself, “Thanks to this unique algorithm, Iris M is approximately 100 times more sensitive at measuring displacement than traditional image-based measurement tools. Additionally, when necessary, Iris M is able to measure directly from the image point, which can be used to quantify the absolute displacement of the motion, rather than having to interpret that point measurement to determine the type of motion and the errors present,” he said. Another great benefit of this technology is the speed of data feedback and the level of detail in the data. Unlike traditional contact measurement systems, it is also scalable because all asset vibrations within the camera’s field of view can be measured simultaneously. It also lends itself to being a great communication tool between technical and non-technical users, as the source of any problem with any asset can be seen directly in the video.

The new system has been deployed in a variety of practical applications. In addition to performing condition monitoring of industrial assets such as machinery, the Iris M system can also be used to analyze the structural integrity of bridges, buildings and similar structures. It can also be used in biomedical monitoring applications to assess an individual's respiratory function.