Using NoSQL Databases in the Context of Manufacturing Execution Systems

Due to globalization, rapid growth in new markets and increased competition, the companies are forced to find new challenges to continue their success. Efficiency in production is one of the important points that should be considered by companies. A production schedule, that is as accurate as possible, requires current information on the actual state of the production. This can be achieved by a deep integration of planning mechanisms with the implementation level. Manufacturing Execution Systems (MES) provide the opportunities to optimize production activities in a manufacturing facility with the focus on quick response to changing conditions. Production activities produce a large amount of data and they should be stored and managed efficiently. In some situations, the traditional databases exceed to their limits and cannot achieve this. As an alternative, NoSQL databases are offered. With features like better scalability, performance, and schema-free structure, they promise a lot. Focus of this work is using NoSQL databases in the context of MES. First, the relational database PostgreSQL and three NoSQL databases Apache Cassandra, MongoDB and Redis are analyzed. Then, using the developed test cases, these databases are compared to each other. After that, MES is introduced and some of its features are explained. Based on this, different application scenarios are implemented in context of MES, to determine how often the database is accessed. Finally, based on these scenarios and test results, suggestions are made, which database is suitable for which MES scenario.