Challenges from Data-Driven Predictive Maintenance in Brownfield Industrial Settings

In the last years many companies made substantial investments in digitization of production and started collecting a lot of data. However, the big question arises how to make sense of all these data and to create competitive advantage? In this regard maintenance is an ever-urged topic and seems to be a low hanging fruit to realize benefits from analyzing large amounts of sensor data now available. This is however, very challenging in typical industrial environments where we can find a mixture of old and new production infrastructure, called brownfield environment. In this work in progress paper we want to investigate this context and identify challenges for the introduction of Big Data approaches for predictive maintenance. For this purpose, we conducted a case study with a world reputed electronic components company. We found that making sense out of sensor data and finding the right level of detail for the analysis is very challenging. We developed a feedback app to incorporate the employees’ domain knowledge in the sense making process.