Application of Process Mining to Production Lines Using Industrial Internet of Things

Process mining has emerged as a powerful solution for analyzing production processes in the manufacturing sector, yet its integration with the Internet of Things (IoT) remains a challenge. Drawing from IoT ontologies and business process context models, we introduced a practical methodology for constructing event logs from industrially gathered IoT data. The study was conducted in a controlled laboratory environment to emulate a discrete production model through an industrial communication and database setup. Employing a systematic process mining framework encompassing data preparation, pre-processing, production process analysis, and interpretation steps, we used industrial controllers and Industrial Internet of Things (IIoT) tools to simulate a sauce production process. The generated data was analyzed using ProM software, fortified by visualization techniques such as the 'log as log skeleton' method, for enhancing process flow comprehension and pinpointing ways for optimization. The study results unveiled a systematic linear structure by employing the Alpha Miner method for process model construction. Notably, the analysis exposes extended duration in the production process, prompting further investigation and refinement. The resultant model provides valuable insights into the actual performance of the process and emphasizes the range of potential process scenarios, making it easier to pinpoint deviations from the ideal model. Through the demonstration of the IIoT data integration, the result significantly contributes to the growing enthusiasm for process-oriented methodologies within the manufacturing sector as well as providing an example approach to transform captured IoT data into event logs via contextualization. Furthermore, the study result provides the groundwork for future research and improvement endeavors to enhance manufacturing practices and outcomes.