Event-Based Data Acquisition for Production Process Analyses: A Systematic Mapping Study

The aspiration on modern manufacturing systems is raising. One reason is the goal of interconnected systems and a striving for a high degree of automation. Data should be available everywhere to make production as transparent and efficient as possible. A process can only be controlled if its behavior can be measured. Intuition gives a direction, data contain knowledge. The term Industry 4.0 covers the contents of the first sentences, but the recognizable benefits have their risks. The increase in technical complexity in combination with already existing process inconsistencies requires mechanisms which identify the causes for deviations between the production program plan and the actual values. Due to the fact that manufacturing systems are optimized continuously, the analyses of the systems with current available data volumes is used comparatively little. Our research project claims on this gap. The idea is to analyse operating states of machines with regard to the period a machine is in a certain state. This can be realised via automated monitoring. As an example, if the required processing time fluctuates regularly and does not relate to the schedule, there is room for improvement. Same for other machine states. In first stage, we focus on the machine runtime by dividing it into production-relevant parts to detect variations in times. The detection of variation contains potential for improvement. In this paper a systematic mapping study is carried out. The focus is to give an overview of current research on - especially event-based - data acquisition and analyses of initially discrete production machines. In addition, the purposes of data acquisition, the analyses of machine runtime data and the way of implementation are considered. By reviewing and extracting information from current publications, the research questions from this paper will be answered and further research gaps identified.