Quality inspection of a production line with a smart camera

Abstract

The technologies for continuous monitoring, diagnostics, prognostics, and control of assets have been developing tremendously in recent years. The new technologies provide tools to achieve greater predictability of plant behaviour and visibility, reduced safety risks, enhanced security and cost efficiency. eSonia project is researching a possibility to create an asset-aware and self recovery plant. In this thesis is implemented quality inspection part of eSonia project. The implementation includes choosing the camera and the components, building the inspection and the user interfaces, and also creating robust communication between the camera and the plant by using mixture of old industrial standards (like Modbus) and new technologies (like web services). The roadmap for building a machine vision application was tailored to suit smart cameras, and all the steps for building the inspection has been presented in detail. The implementation was done by using National Instrument’s NI1774C smart camera, and National Instruments Vision Builder AI software, and the web service was build on Inico’s remote terminal unit S1000. The hardware and software composition proved to be suitable to perform all the tasks, and to be well suited to the asset-aware factory environment. In this thesis is also presented the state of arts of smart cameras, for off-the-shelf solutions as well as for research projects. The market for smart cameras is increasing rapidly, and the market situation is presented together with the direction for future technological development.