Concept for a virtual process data linkage of assembly stations and a dynamic envelope curve for process monitoring

Abstract

Industrial manufacturing and assembly aim to realize a wide range of product variance at high quality standards. [7] The fabrication processes are commonly organized as workshop production or chained production systems, besides standalone machines. [3][4] A lot of process data is generated by every single machine, yet it is hardly used for process optimization. Depending on the manufacturing IT, process data of series production is stored within databases optimized for traceability, whereas standalone machines and machines within workshop production are usually not connected to a common database. The required process data is either stored on the assembly machine itself or inside a local database. [9] The identification of interdependences of each single assembly process and the quality of the finished good is necessary for advanced optimization. Due to the decentralized process data storage, data mining analysis is taking a huge amount of time to find and prepare the process and quality data, especially in workshop production. To enable process monitoring and holistic optimization based on data mining methods in workshop production, a methodology is required to extract, transform and store process data like pressing curves and quality data. Therefore, this paper provides a concept for a virtual process data linkage of assembly stations to enable data mining inside workshop production, which is also able to cope with chained production systems and standalone machines. For further analysis of interdependencies of assembly presses, a dynamic envelope curve is developed for advanced monitoring and optimization as novel methodology.