A Data Fusion Service-Oriented Infrastructure for Production Line Monitoring

The Industry 4.0 paradigm has deeply changed classical manufacturing by introducing data-based analytics and decision-support strategies. At the state of the art, data used for manufacturing monitoring is mostly originated by sensors, that undergo a fusion step to align different data sources. However, this data is only relative to the monitored process, and it does not include the corresponding operating conditions and parameters, that are known by the Manufacturing Execution System (MES). Such information is currently either not included or labeled by hand, thus incurring in errors and limiting the amount of available labeled data. To overcome this issue and go beyond the sole data fusion of sensor data, this paper proposes an infrastructure that automatically label time series generated by sensors with information extracted from the MES, to achieve enhanced monitoring of the production process. The relevance of the proposed solution and the possibilities opened by its application are stressed with the application to a robotic arm.