Companies strive to optimize their processes to increase productivity, efficiency, and quality in the production system. The complexity of different operations in a production system can vary, the level of complexity impacts the automation approach. The requirements of an automated solution increases with the level of complexity. A vision system is a type of automation solution which can improve the quality inspection process in many companies where the inspection process is performed manually. A challenge with implementing a vision system in different industries is that there are no general solutions. Depending on the requirements of an industry, the implementation of a vision system needs to be tailored to suit that industry. In this thesis, a case study is conducted in a foundry setting where empirical data is gathered through observations and interviews conducted with several departments. Nine parameters are identified which can affect the performance of a vision system in a metal casting environment. The identified parameters are analysed using relevant literature to find possible solutions. The result of the analysis is a set of guidelines which can be used when implementing vision systems in the metal casting industry and aims to mitigate possible negative effects of each parameter. A conceptual layout of a vision cell used to inspect cylinder heads based on the guidelines is also presented.