iba AG uses visual information for process diagnosis

The automated control using camera-based solutions helps detecting error sources in industrial production processes in a very reliable way and hence lays the basis for an interactive correction. For automatic-visual process analysis, iba AG presents the image processing system ibaMachineVision at the SPS IPC Drives 2013 in Nuremberg. This module extends the functionality of ibaCapture-CAM, which is a complete system consisting of hardware and software for recording visual information. “With ibaMachineVision, we offer a system for process analysis and Condition Monitoring, that gives the user the opportunity to integrate also visual information in online analysis and process diagnosis”, explains Dr. Ulrich Lettau, CEO of iba AG. The image processing system can be used in all areas of industrial development and manufacturing.

Using ibaCapture-CAM in combination with the process data acquisition system ibaPDA-V6, synchronized recording of video images and measurement signals of technical processes becomes possible. Video sequences and process data are recorded in a time synchronous manner and can be played back also time synchronously in order to analyse the process behaviour. By means of ibaMachineVision, the process data acquisition system gets additional values from the image processing module. Numerical measurement values like e.g. dimensional values of objects or information concerning the production like e.g. size and location of defects are extracted from the video stream. These data are processed by the “MachineVision-Engine”, using algorithms from image processing for object recognition and information extraction becoming visual measurement values.

Dr. Andreas Quick, head of product management at iba AG, explains the concept: “The MachineVision module gets the video stream recorded by ibaCapture-CAM and generates numerical values – the so-called visual measurement values. These measurement values can be fed in the process data acquisition system to be recorded and displayed as long-term trends.” On the HMI of the ibaMachineVision system, the user watches in parallel the live video stream, the result of the image processing and the long-term trend of the visual measurement values. Based on this long-term trend, the process behaviour can be analyzed and diagnosed. Hence, production errors can be detected early in the process and signalled to the user – this method can be considered as “camera-based Condition Monitoring“.

A typical application for ibaMachineVision is monitoring a continuous production plant like e.g. a continuous bloom caster. The image processing system recognizes the corner points on the front of a square-shaped steel strand and calculates the difference of the diagonals. The process data acquisition system generates a long-term trend of this characteristic value and alerts the user in case of a significant trend using a virtual traffic light on the user surface. The user can immediately intervene in the process and initiate corresponding corrective actions. The production of deficient products is reduced, the process efficiency is enhanced and the product quality ensured.

No Stand Alone system

“The advantage of ibaMachineVision is, that it can be easily integrated in an existing iba system“, Dr. Lettau explains. Due to their connectivity, the manufacturer independent iba products can be integrated in automation systems and bus technologies of all well-established automation manufacturers and generations of devices.

Because of the diversity of demands and applications, at the beginning of each project a feasibility study is being conducted. In this study, the realization of ibaMachineVision concerning the location of the camera installation, lighting conditions, ambient temperatures and mechanical influences is checked. This way, the system can be exactly adapted, configured and optimized to the applications and customer requirements.

Leave a Comment