A Data Transformation Adapter for Smart Manufacturing Systems with Edge and Cloud Computing Capabilities

Abstract

The manufacturing industry is constantly seeking novel solutions to improve productivity and gain a competitive advantage. Considering the large amount of data that manufacturing operations generate, the capability to make a smart decision is tied to the ability to process plant floor data gaining insight into machine and system level performance. This work aims to bridge the gap between the plant floor operation and “Big Data” analysis solutions to help improve manufacturing productivity, quality, and sustainability. The proposed framework incorporates three main elements: data sourcing, analysis, and visualization. The combination of these aspects lays the groundwork for processing large amounts of data on a multi-layer infrastructure that leverages both edge and cloud computing. The data processing framework was tested using a manufacturing testbed with with machines, robots, conveyors, and different types of sensors to replicate the diverse data sources in a manufacturing plant. The data processing infrastructure was used to monitor machine health, detect anomalies, and evaluate throughput.