With the developments in science and technology, automation systems have become an important part of manufacturing plant and production lines. Especially with the integrated systems with robots, machines, and vision systems, the production has become efficient. Vision systems are used for inspection task, quality monitoring and vision-guided robotic applications. Vision-based line tracking is a new trend in vision applications. The objects can be tracked or identified while moving and the coordinates of the objects can be sent to the robot for pick and place operations. By using the line tracking in conveyors time to stop the object can be saved which contributes to efficiency. Along with line tracking product inspection also can be made which ensures the quality of the products.

Cell Impact AB is a company which produces bipolar flow plates for hydrogen fuel cells. Currently, the company has a manufacturing plant in Karlskoga with manual operations. The company is planning to improve the production line by adding robots, conveyors and a vision system. The thesis aims at designing a vision-based line tracking system for a product passing through a conveyor. The research is carried out to investigate requirements to build a vision-based line tracking system.

The project is carried out to design a vision-based line tracking system with a focus on several factors such as light conditions, type of image acquisition and type of lens. An experiment is carried out with an ABB camera and an ABB industrial robot for defining light conditions and image processing of sheet metal products. The sheet metal products are processed with different vision tools for building aneffective system. The data files of sheet metal products are tested with an Omron and LMI vision simulators to test the tools for image processing. The simulation of the plant is carried out in Visual Components and human machine interface is created in TwinCAT PLC for visualization of the simulation.

Two different layouts are proposed for a robot pick and place operation. A theoretical study is made on 2D/3D vision systems, acquisition methods and camera placement from websites, scientific journals and papers. The result comprises all the work done in the project. The experiment provided optimum light conditions, different software's provided image processing tools for sheet metal parts and the simulation provided a 3D view of the plant.