The dynamic behavior of machines can be observed with mobile robotic systems as well. Alignment measurements of rolling mills, for instance, are usually conducted during downtime. However, during operation there are dynamic forces applied on the mill which generate micro movements of its components. Due to hazards for workers when approaching a rolling mill during operation and high cost for fixed installed condition monitoring systems with vibration sensors, a mobile robot will be used for this task combined with a newly developed video microscope that analyses an image sequence (movie) and amplifies small sub-pixel motions in the images. The result is an image sequence with exaggerated motions detectable for both the human eye and image processing algorithms. 

The resulting video stream reveals unexpected dynamic behavior due to wear, fatigue, bearing clearance, suspension and bend of the inspected machine, hence providing new knowledge on dynamics and rigidity of the processing equipment. This improves the settings and, in a compact manner, provides important information for process models and maintenance planning. Figure 5 shows the roughing mill at SWERIM pilot plant. 

The video microscope software and according algorithms will be tested and evaluated at the industrial plant of Kanthal, Sweden, focusing on the blooming mill. Moreover, the video microscope system integrated to the UAV will be tested at SWERIM pilot plant, focusing on the roughing mill. In the pilot plant, there are opportunities to more freely alter the rolling process and the surrounding equipment compared to an industrial production site.